docs/master

annotate discussion.roff @ 170:97461b0b34d2

find changesets by author, revision, files, or words in the commit message
style: Changed indent width; changed formating of lists. This follows Boris's suggestion.
author markus schnalke <meillo@marmaro.de>
date Tue, 10 Jul 2012 15:58:46 +0200
parents 277eeb5ba223
children 346ff7e201f5
rev   line source
meillo@58 1 .H0 "Discussion
meillo@0 2 .P
meillo@169 3 This main chapter discusses the practical work accomplished in the
meillo@169 4 mmh project.
meillo@169 5 It is structured along the goals set for the project.
meillo@169 6 The concrete work undertaken
meillo@58 7 is described in the examples of how the general goals were achieved.
meillo@87 8 The discussion compares the current version of mmh with the state of
meillo@87 9 nmh just before the mmh project started, i.e. Fall 2011.
meillo@87 10 Current changes of nmh will be mentioned only as side notes.
meillo@87 11 .\" XXX where do I discuss the parallel development of nmh?
meillo@58 12
meillo@58 13
meillo@58 14
meillo@133 15 .\" --------------------------------------------------------------
meillo@125 16 .H1 "Streamlining
meillo@58 17
meillo@0 18 .P
meillo@169 19 MH once provided anything necessary for email handling.
meillo@169 20 The community around nmh has the similar understanding that nmh should
meillo@169 21 provide a complete email system.
meillo@169 22 In fundamental contrast, mmh shall be a MUA only.
meillo@87 23 I believe that the development of all-in-one mail systems is obsolete.
meillo@87 24 Today, email is too complex to be fully covered by single projects.
meillo@87 25 Such a project won't be able to excel in all aspects.
meillo@159 26 Instead, the aspects of email should be covered by multiple projects,
meillo@87 27 which then can be combined to form a complete system.
meillo@169 28 Excellent implementations for the various aspects of email already exist.
meillo@87 29 Just to name three examples: Postfix is a specialized MTA,
meillo@159 30 .\" XXX homepages verlinken
meillo@87 31 Procmail is a specialized MDA, and Fetchmail is a specialized MRA.
meillo@89 32 I believe that it is best to use such specialized tools instead of
meillo@87 33 providing the same function again as a side-component in the project.
meillo@164 34 .\" XXX mail agent picture here
meillo@58 35 .P
meillo@169 36 Doing something well requires focusing on a small set of specific aspects.
meillo@169 37 Under the assumption that development focussed on a particular area
meillo@169 38 produces better results there, specialized projects will be superior
meillo@87 39 in their field of focus.
meillo@87 40 Hence, all-in-one mail system projects \(en no matter if monolithic
meillo@87 41 or modular \(en will never be the best choice in any of the fields.
meillo@169 42 Even in providing the best consistent all-in-one system, they are likely
meillo@87 43 to be beaten by projects that focus only on integrating existing mail
meillo@169 44 components to create a homogeneous system.
meillo@87 45 .P
meillo@169 46 The limiting resource in the community development of Free Software
meillo@87 47 is usually man power.
meillo@87 48 If the development power is spread over a large development area,
meillo@87 49 it becomes even more difficult to compete with the specialists in the
meillo@87 50 various fields.
meillo@87 51 The concrete situation for MH-based mail systems is even tougher,
meillo@169 52 given their small and aged community, concerning both developers and users.
meillo@87 53 .P
meillo@87 54 In consequence, I believe that the available development resources
meillo@100 55 should focus on the point where MH is most unique.
meillo@87 56 This is clearly the user interface \(en the MUA.
meillo@125 57 Peripheral parts should be removed to streamline mmh for the MUA task.
meillo@60 58
meillo@60 59
meillo@100 60 .H2 "Mail Transfer Facilities
meillo@154 61 .Id mail-transfer-facilities
meillo@60 62 .P
meillo@60 63 In contrast to nmh, which also provides mail submission and mail retrieval
meillo@87 64 agents, mmh is a MUA only.
meillo@100 65 This general difference initiated the development of mmh.
meillo@169 66 The removal of the mail transfer facilities was the first work task
meillo@76 67 in the mmh project.
meillo@60 68 .P
meillo@169 69 Focusing on one mail agent role only, is motivated by Eric Allman's
meillo@105 70 experience with Sendmail.
meillo@159 71 He identified the limitation of Sendmail to the MTA task as one reason for
meillo@105 72 its success:
meillo@105 73 .[ [
meillo@105 74 costales sendmail
meillo@105 75 .], p. xviii]
meillo@105 76 .QS
meillo@105 77 Second, I limited myself to the routing function \(en
meillo@110 78 I wouldn't write user agents or delivery back-ends.
meillo@105 79 This was a departure of the dominant through of the time,
meillo@105 80 in which routing logic, local delivery, and often the network code
meillo@105 81 were incorporated directly into the user agents.
meillo@105 82 .QE
meillo@105 83 .P
meillo@159 84 In nmh, the Mail Submission Agent (MSA) is called
meillo@105 85 \fIMessage Transfer Service\fP (MTS).
meillo@105 86 This facility, implemented by the
meillo@105 87 .Pn post
meillo@105 88 command, established network connections and spoke SMTP to submit
meillo@159 89 messages to be relayed to the outside world.
meillo@169 90 The changes in email demanded changes in this part of nmh as well.
meillo@89 91 Encryption and authentication for network connections
meillo@87 92 needed to be supported, hence TLS and SASL were introduced into nmh.
meillo@87 93 This added complexity to nmh without improving it in its core functions.
meillo@87 94 Also, keeping up with recent developments in the field of
meillo@87 95 mail transfer requires development power and specialists.
meillo@169 96 In mmh, this whole facility was simply cut off.
meillo@76 97 .Ci f6aa95b724fd8c791164abe7ee5468bf5c34f226
meillo@76 98 .Ci fecd5d34f65597a4dfa16aeabea7d74b191532c3
meillo@76 99 .Ci 156d35f6425bea4c1ed3c4c79783dc613379c65b
meillo@87 100 Instead, mmh depends on an external MSA.
meillo@60 101 The only outgoing interface available to mmh is the
meillo@60 102 .Pn sendmail
meillo@87 103 command, which almost any MSA provides.
meillo@87 104 If not, a wrapper program can be written.
meillo@87 105 It must read the message from the standard input, extract the
meillo@87 106 recipient addresses from the message header, and hand the message
meillo@87 107 over to the MSA.
meillo@87 108 For example, a wrapper script for qmail would be:
meillo@87 109 .VS
meillo@87 110 #!/bin/sh
meillo@138 111 exec qmail-inject # ignore command line arguments
meillo@87 112 VE
meillo@87 113 The requirement to parse the recipient addresses out of the message header
meillo@87 114 is likely to be removed in the future.
meillo@169 115 Then mmh would pass the recipient addresses as command line arguments.
meillo@100 116 This appears to be the better interface.
meillo@87 117 .\" XXX implement it
meillo@60 118 .P
meillo@60 119 To retrieve mail, the
meillo@60 120 .Pn inc
meillo@169 121 command acted as an Mail Retrieval Agent (MRA).
meillo@100 122 It established network connections
meillo@76 123 and spoke POP3 to retrieve mail from remote servers.
meillo@76 124 As with mail submission, the network connections required encryption and
meillo@87 125 authentication, thus TLS and SASL were added.
meillo@169 126 Support for message retrieval through IMAP will soon become necessary
meillo@169 127 additions, too, and likewise for any other changes in mail transfer.
meillo@100 128 Not so for mmh because it has dropped the support for retrieving mail
meillo@100 129 from remote locations.
meillo@76 130 .Ci ab7b48411962d26439f92f35ed084d3d6275459c
meillo@76 131 Instead, it depends on an external tool to cover this task.
meillo@169 132 Mmh has two paths for messages to enter mmh's mail storage:
meillo@100 133 (1) Mail can be incorporated with
meillo@60 134 .Pn inc
meillo@87 135 from the system maildrop, or (2) with
meillo@60 136 .Pn rcvstore
meillo@87 137 by reading them, one at a time, from the standard input.
meillo@60 138 .P
meillo@60 139 With the removal of the MSA and MRA, mmh converted from an all-in-one
meillo@87 140 mail system to being a MUA only.
meillo@60 141 Now, of course, mmh depends on third-party software.
meillo@87 142 An external MSA is required to transfer mail to the outside world;
meillo@60 143 an external MRA is required to retrieve mail from remote machines.
meillo@169 144 Excellent implementations of such software exist,
meillo@159 145 which likely are superior than the internal version.
meillo@159 146 Additionally, the best suiting programs can be freely chosen.
meillo@60 147 .P
meillo@60 148 As it had already been possible to use an external MSA or MRA,
meillo@60 149 why not keep the internal version for convenience?
meillo@159 150 .\" XXX ueberleitung
meillo@76 151 The question whether there is sense in having a fall-back pager in all
meillo@76 152 the command line tools, for the cases when
meillo@60 153 .Pn more
meillo@60 154 or
meillo@60 155 .Pn less
meillo@76 156 aren't available, appears to be ridiculous.
meillo@100 157 Of course, MSAs and MRAs are more complex than text pagers
meillo@87 158 and not necessarily available but still the concept of orthogonal
meillo@87 159 design holds: ``Write programs that do one thing and do it well.''
meillo@87 160 .[
meillo@87 161 mcilroy unix phil
meillo@87 162 p. 53
meillo@87 163 .]
meillo@87 164 .[
meillo@87 165 mcilroy bstj foreword
meillo@87 166 .]
meillo@87 167 Here, this part of the Unix philosophy was applied not only
meillo@87 168 to the programs but to the project itself.
meillo@87 169 In other words:
meillo@164 170 Develop projects that focus on one thing and do it well.
meillo@169 171 Projects which have grown complex should be split, for the same
meillo@169 172 reasons that programs which have grown complex should be split.
meillo@100 173 If it is conceptionally more elegant to have the MSA and MRA as
meillo@87 174 separate projects then they should be separated.
meillo@169 175 In my opinion, this is the case here.
meillo@87 176 The RFCs propose this separation by clearly distinguishing the different
meillo@87 177 mail handling tasks.
meillo@87 178 .[
meillo@87 179 rfc 821
meillo@87 180 .]
meillo@87 181 The small interfaces between the mail agents support the separation.
meillo@76 182 .P
meillo@169 183 Email once had been small and simple.
meillo@100 184 At that time,
meillo@60 185 .Pn /bin/mail
meillo@169 186 had covered everything there was to email and still was small and simple.
meillo@100 187 Later, the essential complexity of email increased.
meillo@87 188 (Essential complexity is the complexity defined by the problem itself.\0
meillo@87 189 .[[
meillo@87 190 brooks no silver bullet
meillo@87 191 .]])
meillo@87 192 Email systems reacted to this change: They grew.
meillo@100 193 RFCs started to introduce the concept of mail agents to separate the
meillo@100 194 various tasks because they became more extensive and new tasks appeared.
meillo@100 195 As the mail systems grew even more, parts were split off.
meillo@169 196 For instance, a POP server was included in the original MH;
meillo@169 197 it was removed in nmh.
meillo@164 198 Now is the time to go one step further and split off the MSA and MRA, too.
meillo@87 199 Not only does this decrease the code size of the project,
meillo@169 200 more importantly, it unburdens mmh of the whole field of
meillo@87 201 message transfer with all its implications for the project.
meillo@169 202 There is no more need for concern with changes in network transfer.
meillo@169 203 This independence is gained by depending on an external program
meillo@76 204 that covers the field.
meillo@76 205 Today, this is a reasonable exchange.
meillo@60 206 .P
meillo@159 207 .\" XXX ueberleitung ???
meillo@100 208 Functionality can be added in three different ways:
meillo@87 209 .BU
meillo@169 210 Implementing the function in the project itself.
meillo@87 211 .BU
meillo@87 212 Depending on a library that provides the function.
meillo@87 213 .BU
meillo@87 214 Depending on a program that provides the function.
meillo@87 215 .P
meillo@159 216 .\" XXX Rework sentence
meillo@169 217 While implementing the function in the project itself leads to the
meillo@169 218 largest increase in code size and requires the most maintenance
meillo@169 219 and development work,
meillo@169 220 it increases the project's independence of other software the most.
meillo@169 221 Using libraries or external programs requires less maintenance work
meillo@87 222 but introduces dependencies on external software.
meillo@169 223 Programs have the smallest interfaces and provide the best separation,
meillo@87 224 but possibly limit the information exchange.
meillo@169 225 External libraries are more strongly connected than external programs,
meillo@169 226 thus information can be exchanged in a more flexible manner.
meillo@87 227 Adding code to a project increases maintenance work.
meillo@87 228 .\" XXX ref
meillo@159 229 Implementing complex functions in the project itself adds
meillo@87 230 a lot of code.
meillo@87 231 This should be avoided if possible.
meillo@169 232 Hence, the dependencies only change in their character,
meillo@169 233 not in their existence.
meillo@66 234 In mmh, library dependencies on
meillo@66 235 .Pn libsasl2
meillo@66 236 and
meillo@66 237 .Pn libcrypto /\c
meillo@66 238 .Pn libssl
meillo@159 239 were traded against program dependencies on an MSA and an MRA.
meillo@159 240 This also meant trading build-time dependencies against run-time
meillo@87 241 dependencies.
meillo@169 242 Besides providing stronger separation and greater flexibility,
meillo@169 243 program dependencies also allowed
meillo@66 244 over 6\|000 lines of code to be removed from mmh.
meillo@66 245 This made mmh's code base about 12\|% smaller.
meillo@87 246 Reducing the project's code size by such an amount without actually
meillo@87 247 losing functionality is a convincing argument.
meillo@87 248 Actually, as external MSAs and MRAs are likely superior to the
meillo@87 249 project's internal versions, the common user even gains functionality.
meillo@66 250 .P
meillo@169 251 Users of MH should not have problems setting up an external MSA and MRA.
meillo@60 252 Also, the popular MSAs and MRAs have large communities and a lot
meillo@169 253 of available documentation.
meillo@169 254 Choices for MSAs range from full-featured MTAs such as
meillo@159 255 .\" XXX refs
meillo@169 256 .I Postfix ,
meillo@169 257 over mid-size MTAs such as
meillo@60 258 .I masqmail
meillo@60 259 and
meillo@169 260 .I dma ,
meillo@169 261 to small forwarders such as
meillo@60 262 .I ssmtp
meillo@60 263 and
meillo@60 264 .I nullmailer .
meillo@60 265 Choices for MRAs include
meillo@60 266 .I fetchmail ,
meillo@60 267 .I getmail ,
meillo@60 268 .I mpop
meillo@60 269 and
meillo@60 270 .I fdm .
meillo@60 271
meillo@60 272
meillo@100 273 .H2 "Non-MUA Tools
meillo@60 274 .P
meillo@87 275 One goal of mmh is to remove the tools that are not part of the MUA's task.
meillo@169 276 Further more, any tools that don't significantly improve the MUA's job
meillo@87 277 should be removed.
meillo@87 278 Loosely related and rarely used tools distract from the lean appearance.
meillo@87 279 They require maintenance work without adding much to the core task.
meillo@125 280 By removing these tools, the project shall become more streamlined
meillo@87 281 and focused.
meillo@169 282 In mmh, the following tools are not available anymore:
meillo@62 283 .BU
meillo@58 284 .Pn conflict
meillo@87 285 was removed
meillo@76 286 .Ci 8b235097cbd11d728c07b966cf131aa7133ce5a9
meillo@87 287 because it is a mail system maintenance tool that is not MUA-related.
meillo@87 288 It even checked
meillo@58 289 .Fn /etc/passwd
meillo@58 290 and
meillo@58 291 .Fn /etc/group
meillo@87 292 for consistency, which is completely unrelated to email.
meillo@87 293 A tool like
meillo@87 294 .Pn conflict
meillo@87 295 is surely useful, but it should not be shipped with mmh.
meillo@76 296 .\" XXX historic reasons?
meillo@62 297 .BU
meillo@58 298 .Pn rcvtty
meillo@87 299 was removed
meillo@87 300 .Ci 14767c94b3827be7c867196467ed7aea5f6f49b0
meillo@89 301 because its use case of writing to the user's terminal
meillo@159 302 on receival of mail is obsolete.
meillo@87 303 If users like to be informed of new mail, the shell's
meillo@58 304 .Ev MAILPATH
meillo@87 305 variable or graphical notifications are technically more appealing.
meillo@169 306 Writing directly to terminals is hardly ever desired today.
meillo@169 307 If, though, one prefers this approach, the standard tool
meillo@58 308 .Pn write
meillo@58 309 can be used in a way similar to:
meillo@82 310 .VS
meillo@58 311 scan -file - | write `id -un`
meillo@82 312 VE
meillo@62 313 .BU
meillo@58 314 .Pn viamail
meillo@159 315 .\" XXX was macht viamail
meillo@87 316 was removed
meillo@87 317 .Ci eda72d6a7a7c20ff123043fb7f19c509ea01f932
meillo@87 318 when the new attachment system was activated, because
meillo@58 319 .Pn forw
meillo@76 320 could then cover the task itself.
meillo@62 321 The program
meillo@58 322 .Pn sendfiles
meillo@62 323 was rewritten as a shell script wrapper around
meillo@58 324 .Pn forw .
meillo@76 325 .Ci 0e82199cf3c991a173e0ac8aa776efdb3ded61e6
meillo@62 326 .BU
meillo@58 327 .Pn msgchk
meillo@159 328 .\" XXX was macht msgchk
meillo@87 329 was removed
meillo@87 330 .Ci bb9360ead7eb7a3fedcce2eeedfc660014e41dbe ,
meillo@87 331 because it lost its use case when POP support was removed.
meillo@76 332 A call to
meillo@58 333 .Pn msgchk
meillo@87 334 provided hardly more information than:
meillo@82 335 .VS
meillo@58 336 ls -l /var/mail/meillo
meillo@82 337 VE
meillo@100 338 It did distinguish between old and new mail, but
meillo@169 339 these details can be retrieved with
meillo@76 340 .Pn stat (1),
meillo@62 341 too.
meillo@100 342 A small shell script could be written to print the information
meillo@76 343 in a similar way, if truly necessary.
meillo@76 344 As mmh's
meillo@76 345 .Pn inc
meillo@87 346 only incorporates mail from the user's local maildrop,
meillo@62 347 and thus no data transfers over slow networks are involved,
meillo@169 348 there is hardly any need to check for new mail before incorporating it.
meillo@62 349 .BU
meillo@58 350 .Pn msh
meillo@87 351 was removed
meillo@76 352 .Ci 916690191222433a6923a4be54b0d8f6ac01bd02
meillo@87 353 because the tool was in conflict with the philosophy of MH.
meillo@76 354 It provided an interactive shell to access the features of MH,
meillo@159 355 but it wasn't just a shell tailored to the needs of mail handling.
meillo@169 356 Instead, it was one large program that had several MH tools built in.
meillo@76 357 This conflicts with the major feature of MH of being a tool chest.
meillo@76 358 .Pn msh 's
meillo@159 359 main use case had been accessing Bulletin Boards, which have ceased to
meillo@62 360 be popular.
meillo@62 361 .P
meillo@62 362 Removing
meillo@169 363 .Pn msh
meillo@76 364 together with the truly archaic code relicts
meillo@58 365 .Pn vmh
meillo@58 366 and
meillo@169 367 .Pn wmh
meillo@62 368 saved more than 7\|000 lines of C code \(en
meillo@66 369 about 15\|% of the project's original source code amount.
meillo@100 370 Having less code \(en with equal readability, of course \(en
meillo@76 371 for the same functionality is an advantage.
meillo@63 372 Less code means less bugs and less maintenance work.
meillo@76 373 As
meillo@63 374 .Pn rcvtty
meillo@63 375 and
meillo@63 376 .Pn msgchk
meillo@87 377 are assumed to be rarely used and can be implemented in different ways,
meillo@87 378 why should one keep them?
meillo@125 379 Removing them streamlines mmh.
meillo@63 380 .Pn viamail 's
meillo@63 381 use case is now partly obsolete and partly covered by
meillo@63 382 .Pn forw ,
meillo@76 383 hence there's no reason to still maintain it.
meillo@63 384 .Pn conflict
meillo@76 385 is not related to the mail client, and
meillo@63 386 .Pn msh
meillo@63 387 conflicts with the basic concept of MH.
meillo@169 388 These two tools might still be useful, but they should not be part of mmh.
meillo@63 389 .P
meillo@169 390 Finally, there is
meillo@76 391 .Pn slocal .
meillo@76 392 .Pn slocal
meillo@76 393 is an MDA and thus not directly MUA-related.
meillo@100 394 It should be removed from mmh, because including it conflicts with
meillo@100 395 the idea that mmh is a MUA only.
meillo@100 396 .Pn slocal
meillo@100 397 should rather become a separate project.
meillo@87 398 However,
meillo@76 399 .Pn slocal
meillo@76 400 provides rule-based processing of messages, like filing them into
meillo@76 401 different folders, which is otherwise not available in mmh.
meillo@87 402 Although
meillo@76 403 .Pn slocal
meillo@169 404 neither pulls in dependencies, nor does it include a separate
meillo@154 405 technical area (cf. Sec.
meillo@154 406 .Cf mail-transfer-facilities ),
meillo@169 407 it still accounts for about 1\|000 lines of code that need to be maintained.
meillo@76 408 As
meillo@76 409 .Pn slocal
meillo@76 410 is almost self-standing, it should be split off into a separate project.
meillo@76 411 This would cut the strong connection between the MUA mmh and the MDA
meillo@76 412 .Pn slocal .
meillo@87 413 For anyone not using MH,
meillo@87 414 .Pn slocal
meillo@87 415 would become yet another independent MDA, like
meillo@87 416 .I procmail .
meillo@100 417 Then
meillo@87 418 .Pn slocal
meillo@100 419 could be installed without the complete MH system.
meillo@76 420 Likewise, mmh users could decide to use
meillo@76 421 .I procmail
meillo@87 422 without having a second, unused MDA,
meillo@87 423 .Pn slocal ,
meillo@76 424 installed.
meillo@100 425 That appears to be conceptionally the best solution.
meillo@76 426 Yet,
meillo@76 427 .Pn slocal
meillo@87 428 is not split off.
meillo@100 429 I defer the decision over
meillo@78 430 .Pn slocal
meillo@169 431 out of a need for deeper investigation.
meillo@100 432 In the meanwhile, it remains part of mmh.
meillo@159 433 However, its continued existence is not significant because
meillo@100 434 .Pn slocal
meillo@100 435 is unrelated to the rest of the project.
meillo@0 436
meillo@58 437
meillo@133 438
meillo@134 439 .H2 "Displaying Messages
meillo@155 440 .Id mhshow
meillo@131 441 .P
meillo@133 442 Since the very beginning, already in the first concept paper,
meillo@159 443 .\" XXX ref!!!
meillo@58 444 .Pn show
meillo@62 445 had been MH's message display program.
meillo@58 446 .Pn show
meillo@76 447 mapped message numbers and sequences to files and invoked
meillo@58 448 .Pn mhl
meillo@89 449 to have the files formatted.
meillo@88 450 With MIME, this approach wasn't sufficient anymore.
meillo@100 451 MIME messages can consist of multiple parts. Some parts are not
meillo@100 452 directly displayable and text content might be encoded in
meillo@58 453 foreign charsets.
meillo@58 454 .Pn show 's
meillo@76 455 understanding of messages and
meillo@58 456 .Pn mhl 's
meillo@88 457 display capabilities couldn't cope with the task any longer.
meillo@62 458 .P
meillo@88 459 Instead of extending these tools, additional tools were written from
meillo@88 460 scratch and added to the MH tool chest.
meillo@88 461 Doing so is encouraged by the tool chest approach.
meillo@88 462 Modular design is a great advantage for extending a system,
meillo@88 463 as new tools can be added without interfering with existing ones.
meillo@62 464 First, the new MIME features were added in form of the single program
meillo@58 465 .Pn mhn .
meillo@58 466 The command
meillo@82 467 .Cl "mhn -show 42
meillo@58 468 would show the MIME message numbered 42.
meillo@58 469 With the 1.0 release of nmh in February 1999, Richard Coleman finished
meillo@58 470 the split of
meillo@58 471 .Pn mhn
meillo@88 472 into a set of specialized tools, which together covered the
meillo@88 473 multiple aspects of MIME.
meillo@88 474 One of them was
meillo@69 475 .Pn mhshow ,
meillo@88 476 which replaced
meillo@88 477 .Cl "mhn -show" .
meillo@88 478 It was capable of displaying MIME messages appropriately.
meillo@62 479 .P
meillo@88 480 From then on, two message display tools were part of nmh,
meillo@76 481 .Pn show
meillo@76 482 and
meillo@76 483 .Pn mhshow .
meillo@88 484 To ease the life of users,
meillo@69 485 .Pn show
meillo@69 486 was extended to automatically hand the job over to
meillo@69 487 .Pn mhshow
meillo@69 488 if displaying the message would be beyond
meillo@69 489 .Pn show 's
meillo@69 490 abilities.
meillo@88 491 In consequence, the user would simply invoke
meillo@69 492 .Pn show
meillo@69 493 (possibly through
meillo@69 494 .Pn next
meillo@69 495 or
meillo@69 496 .Pn prev )
meillo@69 497 and get the message printed with either
meillo@69 498 .Pn show
meillo@69 499 or
meillo@69 500 .Pn mhshow ,
meillo@69 501 whatever was more appropriate.
meillo@69 502 .P
meillo@69 503 Having two similar tools for essentially the same task is redundant.
meillo@88 504 Usually,
meillo@88 505 users wouldn't distinguish between
meillo@88 506 .Pn show
meillo@88 507 and
meillo@88 508 .Pn mhshow
meillo@88 509 in their daily mail reading.
meillo@88 510 Having two separate display programs was therefore mainly unnecessary
meillo@88 511 from a user's point of view.
meillo@88 512 Besides, the development of both programs needed to be in sync,
meillo@76 513 to ensure that the programs behaved in a similar way,
meillo@76 514 because they were used like a single tool.
meillo@76 515 Different behavior would have surprised the user.
meillo@69 516 .P
meillo@69 517 Today, non-MIME messages are rather seen to be a special case of
meillo@100 518 MIME messages, although it is the other way round.
meillo@69 519 As
meillo@69 520 .Pn mhshow
meillo@159 521 had already been able to display non-MIME messages, it appeared natural
meillo@69 522 to drop
meillo@69 523 .Pn show
meillo@69 524 in favor of using
meillo@69 525 .Pn mhshow
meillo@69 526 exclusively.
meillo@88 527 .Ci 4c1efddfd499300c7e74263e57d8aa137e84c853
meillo@88 528 Removing
meillo@88 529 .Pn show
meillo@88 530 is no loss in function, because functionally
meillo@88 531 .Pn mhshow
meillo@88 532 covers it completely.
meillo@88 533 The old behavior of
meillo@88 534 .Pn show
meillo@88 535 can still be emulated with the simple command line:
meillo@88 536 .VS
meillo@88 537 mhl `mhpath c`
meillo@88 538 VE
meillo@88 539 .P
meillo@76 540 For convenience,
meillo@76 541 .Pn mhshow
meillo@88 542 was renamed to
meillo@88 543 .Pn show
meillo@88 544 after
meillo@88 545 .Pn show
meillo@88 546 was gone.
meillo@88 547 It is clear that such a rename may confuse future developers when
meillo@88 548 trying to understand the history.
meillo@88 549 Nevertheless, I consider the convenience on the user's side,
meillo@88 550 to call
meillo@88 551 .Pn show
meillo@88 552 when they want a message to be displayed, to outweigh the inconvenience
meillo@88 553 on the developer's side when understanding the project history.
meillo@69 554 .P
meillo@88 555 To prepare for the transition,
meillo@69 556 .Pn mhshow
meillo@69 557 was reworked to behave more like
meillo@69 558 .Pn show
meillo@69 559 first.
meillo@154 560 (cf. Sec.
meillo@154 561 .Cf mhshow )
meillo@164 562 .\" XXX code commits?
meillo@88 563 Once the tools behaved more alike, the replacing appeared to be
meillo@88 564 even more natural.
meillo@88 565 Today, mmh's new
meillo@69 566 .Pn show
meillo@159 567 has become the one single message display program once more,
meillo@159 568 with the difference
meillo@88 569 that today it handles MIME messages as well as non-MIME messages.
meillo@88 570 The outcome of the transition is one program less to maintain,
meillo@88 571 no second display program for users to deal with,
meillo@88 572 and less system complexity.
meillo@69 573 .P
meillo@88 574 Still, removing the old
meillo@69 575 .Pn show
meillo@88 576 hurts in one regard: It had been such a simple program.
meillo@159 577 Its lean elegance is missing from the new
meillo@159 578 .Pn show ,
meillo@159 579 .\" XXX
meillo@159 580 however there is no alternative;
meillo@159 581 supporting MIME demands higher essential complexity.
meillo@58 582
meillo@134 583 .ig
meillo@134 584 XXX
meillo@134 585 Consider including text on scan listings here
meillo@58 586
meillo@134 587 Scan listings shall not contain body content. Hence, removed this feature.
meillo@134 588 Scan listings shall operator on message headers and non-message information
meillo@134 589 only. Displaying the beginning of the body complicates everything too much.
meillo@134 590 That's no surprise, because it's something completely different. If you
meillo@134 591 want to examine the body, then use show(1)/mhshow(1).
meillo@134 592 Changed the default scan formats accordingly.
meillo@134 593 .Ci 70b2643e0da8485174480c644ad9785c84f5bff4
meillo@134 594 ..
meillo@131 595
meillo@131 596
meillo@131 597
meillo@133 598
meillo@100 599 .H2 "Configure Options
meillo@58 600 .P
meillo@76 601 Customization is a double-edged sword.
meillo@76 602 It allows better suiting setups, but not for free.
meillo@76 603 There is the cost of code complexity to be able to customize.
meillo@76 604 There is the cost of less tested setups, because there are
meillo@72 605 more possible setups and especially corner-cases.
meillo@159 606 Additionally, there is the cost of choice itself.
meillo@76 607 The code complexity directly affects the developers.
meillo@72 608 Less tested code affects both, users and developers.
meillo@159 609 The problem of choice affects the users, for once by having to choose,
meillo@159 610 but also by more complex interfaces that require more documentation.
meillo@159 611 Whenever options add few advantages but increase the complexity of the
meillo@159 612 system, they should be considered for removal.
meillo@72 613 I have reduced the number of project-specific configure options from
meillo@72 614 fifteen to three.
meillo@74 615
meillo@76 616 .U3 "Mail Transfer Facilities
meillo@74 617 .P
meillo@85 618 With the removal of the mail transfer facilities five configure
meillo@85 619 options vanished:
meillo@85 620 .P
meillo@85 621 The switches
meillo@85 622 .Sw --with-tls
meillo@85 623 and
meillo@85 624 .Sw --with-cyrus-sasl
meillo@89 625 had activated the support for transfer encryption and authentication.
meillo@159 626 .\" XXX cf
meillo@159 627 .\" XXX gruende kurz wiederholen
meillo@85 628 This is not needed anymore.
meillo@85 629 .Ci fecd5d34f65597a4dfa16aeabea7d74b191532c3
meillo@85 630 .Ci 156d35f6425bea4c1ed3c4c79783dc613379c65b
meillo@85 631 .P
meillo@159 632 .\" XXX cf
meillo@159 633 .\" XXX ``For the same reason ...''
meillo@85 634 The configure switch
meillo@85 635 .Sw --enable-pop
meillo@85 636 activated the message retrieval facility.
meillo@85 637 The code area that would be conditionally compiled in for TLS and SASL
meillo@85 638 support had been small.
meillo@85 639 The conditionally compiled code area for POP support had been much larger.
meillo@85 640 Whereas the code base changes would only slightly change on toggling
meillo@85 641 TLS or SASL support, it changed much on toggling POP support.
meillo@85 642 The changes in the code base could hardly be overviewed.
meillo@159 643 By having POP support togglable, a second code base had been created,
meillo@85 644 one that needed to be tested.
meillo@85 645 This situation is basically similar for the conditional TLS and SASL
meillo@85 646 code, but there the changes are minor and can yet be overviewed.
meillo@85 647 Still, conditional compilation of a code base creates variations
meillo@85 648 of the original program.
meillo@85 649 More variations require more testing and maintenance work.
meillo@85 650 .P
meillo@85 651 Two other options only specified default configuration values:
meillo@100 652 .Sw --with-mts
meillo@164 653 defined the default transport service.
meillo@85 654 .Ci f6aa95b724fd8c791164abe7ee5468bf5c34f226
meillo@85 655 With
meillo@100 656 .Sw --with-smtpservers
meillo@164 657 default SMTP servers could be specified.
meillo@72 658 .Ci 128545e06224233b7e91fc4c83f8830252fe16c9
meillo@164 659 Both of them became irrelevant when the SMTP transport service was removed.
meillo@164 660 .\" XXX code ref
meillo@164 661 In mmh, all messages are handed over to
meillo@164 662 .Pn sendmail
meillo@164 663 for transportation.
meillo@164 664
meillo@72 665
meillo@74 666 .U3 "Backup Prefix
meillo@74 667 .P
meillo@76 668 The backup prefix is the string that was prepended to message
meillo@76 669 filenames to tag them as deleted.
meillo@76 670 By default it had been the comma character `\f(CW,\fP'.
meillo@159 671 .\" XXX Zeitlich ordnen
meillo@78 672 In July 2000, Kimmo Suominen introduced
meillo@78 673 the configure option
meillo@78 674 .Sw --with-hash-backup
meillo@78 675 to change the default to the hash symbol `\f(CW#\fP'.
meillo@78 676 The choice was probably personal preference, because first, the
meillo@78 677 option was named
meillo@78 678 .Sw --with-backup-prefix.
meillo@78 679 and had the prefix symbol as argument.
meillo@100 680 But giving the hash symbol as argument caused too many problems
meillo@100 681 for Autoconf,
meillo@100 682 thus the option was limited to use the hash symbol as the default prefix.
meillo@100 683 This supports the assumption, that the choice for the hash was
meillo@100 684 personal preference only.
meillo@100 685 Being related or not, words that start with the hash symbol
meillo@78 686 introduce a comment in the Unix shell.
meillo@72 687 Thus, the command line
meillo@72 688 .Cl "rm #13 #15
meillo@72 689 calls
meillo@72 690 .Pn rm
meillo@72 691 without arguments because the first hash symbol starts the comment
meillo@72 692 that reaches until the end of the line.
meillo@72 693 To delete the backup files,
meillo@72 694 .Cl "rm ./#13 ./#15"
meillo@72 695 needs to be used.
meillo@100 696 Using the hash as backup prefix can be seen as a precaution against
meillo@78 697 data loss.
meillo@78 698 .P
meillo@159 699 First, I removed the configure option but added the profile entry
meillo@72 700 .Pe backup-prefix ,
meillo@72 701 which allows to specify an arbitrary string as backup prefix.
meillo@72 702 .Ci 6c40d481d661d532dd527eaf34cebb6d3f8ed086
meillo@76 703 Profile entries are the common method to change mmh's behavior.
meillo@76 704 This change did not remove the choice but moved it to a location where
meillo@72 705 it suited better.
meillo@76 706 .P
meillo@78 707 Eventually, however, the new trash folder concept
meillo@154 708 (cf. Sec.
meillo@154 709 .Cf trash-folder )
meillo@164 710 removed the need for the backup prefix completely.
meillo@78 711 .Ci 8edc5aaf86f9f77124664f6801bc6c6cdf258173
meillo@133 712 .Ci ca0b3e830b86700d9e5e31b1784de2bdcaf58fc5
meillo@133 713
meillo@76 714
meillo@76 715 .U3 "Editor and Pager
meillo@74 716 .P
meillo@74 717 The two configure options
meillo@74 718 .CW --with-editor=EDITOR
meillo@74 719 .CW --with-pager=PAGER
meillo@74 720 were used to specify the default editor and pager at configure time.
meillo@109 721 Doing so at configure time made sense in the eighties,
meillo@76 722 when the set of available editors and pagers varied much across
meillo@76 723 different systems.
meillo@89 724 Today, the situation is more homogeneous.
meillo@74 725 The programs
meillo@74 726 .Pn vi
meillo@74 727 and
meillo@74 728 .Pn more
meillo@76 729 can be expected to be available on every Unix system,
meillo@74 730 as they are specified by POSIX since two decades.
meillo@74 731 (The specifications for
meillo@74 732 .Pn vi
meillo@74 733 and
meillo@74 734 .Pn more
meillo@74 735 appeared in
meillo@74 736 .[
meillo@74 737 posix 1987
meillo@74 738 .]
meillo@74 739 and,
meillo@74 740 .[
meillo@74 741 posix 1992
meillo@74 742 .]
meillo@74 743 respectively.)
meillo@74 744 As a first step, these two tools were hard-coded as defaults.
meillo@74 745 .Ci 5d43a99db70c12a673028c7758c20cbe3e13ef5f
meillo@74 746 Not changed were the
meillo@74 747 .Pe editor
meillo@74 748 and
meillo@74 749 .Pe moreproc
meillo@76 750 profile entries, which allowed the user to override the system defaults.
meillo@74 751 Later, the concept was reworked to respect the standard environment
meillo@74 752 variables
meillo@74 753 .Ev VISUAL
meillo@74 754 and
meillo@74 755 .Ev PAGER
meillo@76 756 if they are set.
meillo@74 757 Today, mmh determines the editor to use in the following order,
meillo@74 758 taking the first available and non-empty item:
meillo@74 759 .IP (1)
meillo@74 760 Environment variable
meillo@74 761 .Ev MMHEDITOR
meillo@74 762 .IP (2)
meillo@74 763 Profile entry
meillo@74 764 .Pe Editor
meillo@74 765 .IP (3)
meillo@74 766 Environment variable
meillo@74 767 .Ev VISUAL
meillo@74 768 .IP (4)
meillo@74 769 Environment variable
meillo@74 770 .Ev EDITOR
meillo@74 771 .IP (5)
meillo@74 772 Command
meillo@74 773 .Pn vi .
meillo@74 774 .P
meillo@76 775 .Ci f85f4b7ae62e3d05a945dcd46ead51f0a2a89a9b
meillo@76 776 .P
meillo@89 777 The pager to use is determined in a similar order,
meillo@74 778 also taking the first available and non-empty item:
meillo@74 779 .IP (1)
meillo@74 780 Environment variable
meillo@74 781 .Ev MMHPAGER
meillo@74 782 .IP (2)
meillo@74 783 Profile entry
meillo@74 784 .Pe Pager
meillo@74 785 (replaces
meillo@74 786 .Pe moreproc )
meillo@74 787 .IP (3)
meillo@74 788 Environment variable
meillo@74 789 .Ev PAGER
meillo@74 790 .IP (4)
meillo@74 791 Command
meillo@74 792 .Pn more .
meillo@74 793 .P
meillo@74 794 .Ci 0c4214ea2aec6497d0d67b436bbee9bc1d225f1e
meillo@74 795 .P
meillo@76 796 By respecting the
meillo@74 797 .Ev VISUAL /\c
meillo@74 798 .Ev EDITOR
meillo@74 799 and
meillo@74 800 .Ev PAGER
meillo@76 801 environment variables,
meillo@76 802 the new behavior confirms better to the common style on Unix systems.
meillo@76 803 Additionally, the new approach is more uniform and clearer to users.
meillo@72 804
meillo@72 805
meillo@76 806 .U3 "ndbm
meillo@72 807 .P
meillo@74 808 .Pn slocal
meillo@78 809 used to depend on
meillo@74 810 .I ndbm ,
meillo@74 811 a database library.
meillo@76 812 The database is used to store the `\fLMessage-ID\fP's of all
meillo@76 813 messages delivered.
meillo@74 814 This enables
meillo@74 815 .Pn slocal
meillo@74 816 to suppress delivering the same message to the same user twice.
meillo@74 817 (This features was enabled by the
meillo@74 818 .Sw -suppressdup
meillo@74 819 switch.)
meillo@74 820 .P
meillo@100 821 A variety of versions of the database library exist.
meillo@78 822 .[
meillo@78 823 wolter unix incompat notes dbm
meillo@78 824 .]
meillo@74 825 Complicated autoconf code was needed to detect them correctly.
meillo@74 826 Further more, the configure switches
meillo@74 827 .Sw --with-ndbm=ARG
meillo@74 828 and
meillo@74 829 .Sw --with-ndbmheader=ARG
meillo@74 830 were added to help with difficult setups that would
meillo@78 831 not be detected automatically or correctly.
meillo@74 832 .P
meillo@74 833 By removing the suppress duplicates feature of
meillo@74 834 .Pn slocal ,
meillo@74 835 the dependency on
meillo@74 836 .I ndbm
meillo@78 837 vanished and 120 lines of complex autoconf code could be saved.
meillo@74 838 .Ci ecd6d6a20cb7a1507e3a20d6c4cb3a1cf14c6bbf
meillo@89 839 The change removed functionality too, but that is minor to the
meillo@78 840 improvement by dropping the dependency and the complex autoconf code.
meillo@159 841 .\" XXX argument: slocal ist sowieso nicht teil vom mmh kern
meillo@72 842
meillo@74 843 .U3 "mh-e Support
meillo@72 844 .P
meillo@74 845 The configure option
meillo@74 846 .Sw --disable-mhe
meillo@74 847 was removed when the mh-e support was reworked.
meillo@74 848 Mh-e is the Emacs front-end to MH.
meillo@76 849 It requires MH to provide minor additional functions.
meillo@76 850 The
meillo@76 851 .Sw --disable-mhe
meillo@76 852 configure option could switch these extensions off.
meillo@76 853 After removing the support for old versions of mh-e,
meillo@74 854 only the
meillo@74 855 .Sw -build
meillo@76 856 switches of
meillo@74 857 .Pn forw
meillo@74 858 and
meillo@74 859 .Pn repl
meillo@76 860 are left to be mh-e extensions.
meillo@76 861 They are now always built in because they add little code and complexity.
meillo@76 862 In consequence, the
meillo@74 863 .Sw --disable-mhe
meillo@76 864 configure option was removed
meillo@72 865 .Ci a7ce7b4a580d77b6c2c4d980812beb589aa4c643
meillo@74 866 Removing the option removed a second code setup that would have
meillo@74 867 needed to be tested.
meillo@159 868 .\" XXX datum?
meillo@169 869 This change was first accomplished in nmh and thereafter merged into mmh.
meillo@76 870 .P
meillo@76 871 The interface changes in mmh require mh-e to be adjusted in order
meillo@76 872 to be able to use mmh as back-end.
meillo@76 873 This will require minor changes to mh-e, but removing the
meillo@76 874 .Sw -build
meillo@76 875 switches would require more rework.
meillo@72 876
meillo@74 877 .U3 "Masquerading
meillo@72 878 .P
meillo@74 879 The configure option
meillo@74 880 .Sw --enable-masquerade
meillo@76 881 could take up to three arguments:
meillo@76 882 `draft_from', `mmailid', and `username_extension'.
meillo@74 883 They activated different types of address masquerading.
meillo@74 884 All of them were implemented in the SMTP-speaking
meillo@74 885 .Pn post
meillo@76 886 command, which provided an MSA.
meillo@76 887 Address masquerading is an MTA's task and mmh does not cover
meillo@76 888 this field anymore.
meillo@76 889 Hence, true masquerading needs to be implemented in the external MTA.
meillo@74 890 .P
meillo@74 891 The
meillo@74 892 .I mmailid
meillo@74 893 masquerading type is the oldest one of the three and the only one
meillo@74 894 available in the original MH.
meillo@74 895 It provided a
meillo@74 896 .I username
meillo@74 897 to
meillo@74 898 .I fakeusername
meillo@76 899 mapping, based on the password file's GECOS field.
meillo@74 900 The man page
meillo@74 901 .Mp mh-tailor(5)
meillo@74 902 described the use case as being the following:
meillo@98 903 .QS
meillo@74 904 This is useful if you want the messages you send to always
meillo@74 905 appear to come from the name of an MTA alias rather than your
meillo@74 906 actual account name. For instance, many organizations set up
meillo@74 907 `First.Last' sendmail aliases for all users. If this is
meillo@74 908 the case, the GECOS field for each user should look like:
meillo@74 909 ``First [Middle] Last <First.Last>''
meillo@98 910 .QE
meillo@74 911 .P
meillo@74 912 As mmh sends outgoing mail via the local MTA only,
meillo@76 913 the best location to do such global rewrites is there.
meillo@74 914 Besides, the MTA is conceptionally the right location because it
meillo@74 915 does the reverse mapping for incoming mail (aliasing), too.
meillo@76 916 Further more, masquerading set up there is readily available for all
meillo@74 917 mail software on the system.
meillo@76 918 Hence, mmailid masquerading was removed.
meillo@74 919 .Ci 0836c8000ccb34b59410ef1c15b1b7feac70ce5f
meillo@74 920 .P
meillo@74 921 The
meillo@74 922 .I username_extension
meillo@76 923 masquerading type did not replace the username but would append a suffix,
meillo@76 924 specified by the
meillo@74 925 .Ev USERNAME_EXTENSION
meillo@76 926 environment variable, to it.
meillo@76 927 This provided support for the
meillo@74 928 .I user-extension
meillo@74 929 feature of qmail and the similar
meillo@74 930 .I "plussed user
meillo@74 931 processing of sendmail.
meillo@74 932 The decision to remove this username_extension masquerading was
meillo@74 933 motivated by the fact that
meillo@74 934 .Pn spost
meillo@76 935 hadn't supported it already.
meillo@76 936 .Ci 2abae0bfd0ad5bf898461e50aa4b466d641f23d9
meillo@76 937 Username extensions are possible in mmh, but less convenient to use.
meillo@159 938 .\" XXX covered by next paragraph
meillo@76 939 .\" XXX format file %(getenv USERNAME_EXTENSION)
meillo@74 940 .P
meillo@74 941 The
meillo@74 942 .I draft_from
meillo@74 943 masquerading type instructed
meillo@74 944 .Pn post
meillo@84 945 to use the value of the
meillo@84 946 .Hd From
meillo@84 947 header field as SMTP envelope sender.
meillo@76 948 Sender addresses could be replaced completely.
meillo@74 949 .Ci b14ea6073f77b4359aaf3fddd0e105989db9
meillo@76 950 Mmh offers a kind of masquerading similar in effect, but
meillo@74 951 with technical differences.
meillo@76 952 As mmh does not transfer messages itself, the local MTA has final control
meillo@76 953 over the sender's address. Any masquerading mmh introduces may be reverted
meillo@76 954 by the MTA.
meillo@76 955 In times of pedantic spam checking, an MTA will take care to use
meillo@76 956 sensible envelope sender addresses to keep its own reputation up.
meillo@84 957 Nonetheless, the MUA can set the
meillo@84 958 .Hd From
meillo@84 959 header field and thereby propose
meillo@76 960 a sender address to the MTA.
meillo@74 961 The MTA may then decide to take that one or generate the canonical sender
meillo@74 962 address for use as envelope sender address.
meillo@74 963 .P
meillo@74 964 In mmh, the MTA will always extract the recipient and sender from the
meillo@84 965 message header (\c
meillo@74 966 .Pn sendmail 's
meillo@74 967 .Sw -t
meillo@74 968 switch).
meillo@84 969 The
meillo@84 970 .Hd From
meillo@84 971 header field of the draft may be set arbitrary by the user.
meillo@74 972 If it is missing, the canonical sender address will be generated by the MTA.
meillo@74 973
meillo@74 974 .U3 "Remaining Options
meillo@74 975 .P
meillo@74 976 Two configure options remain in mmh.
meillo@74 977 One is the locking method to use:
meillo@74 978 .Sw --with-locking=[dot|fcntl|flock|lockf] .
meillo@76 979 The idea of removing all methods except the portable dot locking
meillo@76 980 and having that one as the default is appealing, but this change
meillo@76 981 requires deeper technical investigation into the topic.
meillo@76 982 The other option,
meillo@74 983 .Sw --enable-debug ,
meillo@74 984 compiles the programs with debugging symbols and does not strip them.
meillo@74 985 This option is likely to stay.
meillo@72 986
meillo@72 987
meillo@58 988
meillo@63 989
meillo@100 990 .H2 "Command Line Switches
meillo@58 991 .P
meillo@93 992 The command line switches of MH tools follow the X Window style.
meillo@93 993 They are words, introduced by a single dash.
meillo@93 994 For example:
meillo@93 995 .Cl "-truncate" .
meillo@93 996 Every program in mmh has two generic switches:
meillo@93 997 .Sw -help ,
meillo@93 998 to print a short message on how to use the program, and
meillo@159 999 .Sw -Version
meillo@164 1000 (with capital `V'), to tell what version of mmh the program belongs to.
meillo@93 1001 .P
meillo@93 1002 Switches change the behavior of programs.
meillo@93 1003 Programs that do one thing in one way require no switches.
meillo@93 1004 In most cases, doing something in exactly one way is too limiting.
meillo@97 1005 If there is basically one task to accomplish, but it should be done
meillo@93 1006 in various ways, switches are a good approach to alter the behavior
meillo@93 1007 of a program.
meillo@93 1008 Changing the behavior of programs provides flexibility and customization
meillo@97 1009 to users, but at the same time it complicates the code, documentation and
meillo@93 1010 usage of the program.
meillo@97 1011 .\" XXX: Ref
meillo@93 1012 Therefore, the number of switches should be kept small.
meillo@93 1013 A small set of well-chosen switches does no harm.
meillo@93 1014 But usually, the number of switches increases over time.
meillo@93 1015 Already in 1985, Rose and Romine have identified this as a major
meillo@93 1016 problem of MH:
meillo@93 1017 .[ [
meillo@93 1018 rose romine real work
meillo@93 1019 .], p. 12]
meillo@98 1020 .QS
meillo@93 1021 A complaint often heard about systems which undergo substantial development
meillo@93 1022 by many people over a number of years, is that more and more options are
meillo@93 1023 introduced which add little to the functionality but greatly increase the
meillo@93 1024 amount of information a user needs to know in order to get useful work done.
meillo@93 1025 This is usually referred to as creeping featurism.
meillo@93 1026 .QP
meillo@93 1027 Unfortunately MH, having undergone six years of off-and-on development by
meillo@93 1028 ten or so well-meaning programmers (the present authors included),
meillo@93 1029 suffers mightily from this.
meillo@98 1030 .QE
meillo@93 1031 .P
meillo@97 1032 Being reluctant to adding new switches \(en or `options',
meillo@97 1033 as Rose and Romine call them \(en is one part of a counter-action,
meillo@97 1034 the other part is removing hardly used switches.
meillo@97 1035 Nmh's tools had lots of switches already implemented,
meillo@97 1036 hence, cleaning up by removing some of them was the more important part
meillo@97 1037 of the counter-action.
meillo@93 1038 Removing existing functionality is always difficult because it
meillo@93 1039 breaks programs that use these functions.
meillo@93 1040 Also, for every obsolete feature, there'll always be someone who still
meillo@93 1041 uses it and thus opposes its removal.
meillo@93 1042 This puts the developer into the position,
meillo@93 1043 where sensible improvements to style are regarded as destructive acts.
meillo@97 1044 Yet, living with the featurism is far worse, in my eyes, because
meillo@97 1045 future needs will demand adding further features,
meillo@93 1046 worsening the situation more and more.
meillo@93 1047 Rose and Romine added in a footnote,
meillo@93 1048 ``[...]
meillo@93 1049 .Pn send
meillo@93 1050 will no doubt acquire an endless number of switches in the years to come.''
meillo@97 1051 Although clearly humorous, the comment points to the nature of the problem.
meillo@97 1052 Refusing to add any new switches would encounter the problem at its root,
meillo@97 1053 but this is not practical.
meillo@97 1054 New needs will require new switches and it would be unwise to block
meillo@97 1055 them strictly.
meillo@97 1056 Nevertheless, removing obsolete switches still is an effective approach
meillo@97 1057 to deal with the problem.
meillo@97 1058 Working on an experimental branch without an established user base,
meillo@97 1059 eased my work because I did not offend users when I removed existing
meillo@110 1060 functions.
meillo@93 1061 .P
meillo@93 1062 Rose and Romine counted 24 visible and 9 more hidden switches for
meillo@93 1063 .Pn send .
meillo@97 1064 In nmh, they increased up to 32 visible and 12 hidden ones.
meillo@97 1065 At the time of writing, no more than 7 visible switches and 1 hidden switch
meillo@97 1066 have remained in mmh's
meillo@97 1067 .Pn send .
meillo@97 1068 (These numbers include two generic switches, help and version.)
meillo@93 1069 .P
meillo@154 1070 The figure displays the number of switches for each of the tools
meillo@159 1071 that is available in both nmh and mmh.
meillo@100 1072 The tools are sorted by the number of switches they had in nmh.
meillo@100 1073 Visible and hidden switches were counted,
meillo@97 1074 but not the generic help and version switches.
meillo@93 1075 Whereas in the beginning of the project, the average tool had 11 switches,
meillo@93 1076 now it has no more than 5 \(en only half as many.
meillo@93 1077 If the `no' switches and similar inverse variant are folded onto
meillo@100 1078 their counter-parts, the average tool had 8 switches in pre-mmh times and
meillo@100 1079 has 4 now.
meillo@93 1080 The total number of functional switches in mmh dropped from 465
meillo@93 1081 to 234.
meillo@58 1082
meillo@93 1083 .KS
meillo@93 1084 .in 1c
meillo@93 1085 .so input/switches.grap
meillo@93 1086 .KE
meillo@58 1087
meillo@93 1088 .P
meillo@93 1089 A part of the switches vanished after functions were removed.
meillo@93 1090 This was the case for network mail transfer, for instance.
meillo@97 1091 Sometimes, however, the work flow was the other way:
meillo@97 1092 I looked through the
meillo@97 1093 .Mp mh-chart (7)
meillo@97 1094 man page to identify the tools with apparently too many switches.
meillo@97 1095 Then considering the value of each of the switches by examining
meillo@97 1096 the tool's man page and source code, aided by recherche and testing.
meillo@97 1097 This way, the removal of functions was suggested by the aim to reduce
meillo@97 1098 the number of switches per command.
meillo@97 1099
meillo@58 1100
meillo@93 1101 .U3 "Draft Folder Facility
meillo@93 1102 .P
meillo@100 1103 A change early in the project was the complete transition from
meillo@93 1104 the single draft message to the draft folder facility.
meillo@97 1105 .Ci 337338b404931f06f0db2119c9e145e8ca5a9860
meillo@164 1106 .\" XXX ref to section ...
meillo@109 1107 The draft folder facility was introduced in the mid-eighties, when
meillo@100 1108 Rose and Romine called it a ``relatively new feature''.
meillo@93 1109 .[
meillo@93 1110 rose romine real work
meillo@93 1111 .]
meillo@110 1112 Since then, the facility had existed but was inactive by default.
meillo@93 1113 The default activation and the related rework of the tools made it
meillo@93 1114 possible to remove the
meillo@93 1115 .Sw -[no]draftfolder ,
meillo@93 1116 and
meillo@93 1117 .Sw -draftmessage
meillo@93 1118 switches from
meillo@93 1119 .Pn comp ,
meillo@93 1120 .Pn repl ,
meillo@93 1121 .Pn forw ,
meillo@93 1122 .Pn dist ,
meillo@93 1123 .Pn whatnow ,
meillo@93 1124 and
meillo@93 1125 .Pn send .
meillo@97 1126 .Ci 337338b404931f06f0db2119c9e145e8ca5a9860
meillo@97 1127 The only flexibility removed with this change is having multiple
meillo@97 1128 draft folders within one profile.
meillo@97 1129 I consider this a theoretical problem only.
meillo@159 1130 At the same time, the
meillo@93 1131 .Sw -draft
meillo@93 1132 switch of
meillo@93 1133 .Pn anno ,
meillo@93 1134 .Pn refile ,
meillo@93 1135 and
meillo@93 1136 .Pn send
meillo@93 1137 was removed.
meillo@159 1138 The special treatment of \fIthe\fP draft message became irrelevant after
meillo@93 1139 the rework of the draft system.
meillo@159 1140 (cf. Sec.
meillo@154 1141 .Cf draft-folder )
meillo@164 1142 Furthermore,
meillo@95 1143 .Pn comp
meillo@164 1144 no longer needs a
meillo@95 1145 .Sw -file
meillo@164 1146 switch as the draft folder facility together with the
meillo@95 1147 .Sw -form
meillo@164 1148 switch are sufficient.
meillo@93 1149
meillo@95 1150
meillo@102 1151 .U3 "In Place Editing
meillo@93 1152 .P
meillo@93 1153 .Pn anno
meillo@93 1154 had the switches
meillo@93 1155 .Sw -[no]inplace
meillo@100 1156 to either annotate the message in place and thus preserve hard links,
meillo@93 1157 or annotate a copy to replace the original message, breaking hard links.
meillo@97 1158 Following the assumption that linked messages should truly be the
meillo@97 1159 same message, and annotating it should not break the link, the
meillo@93 1160 .Sw -[no]inplace
meillo@93 1161 switches were removed and the previous default
meillo@93 1162 .Sw -inplace
meillo@93 1163 was made the only behavior.
meillo@97 1164 .Ci c8195849d2e366c569271abb0f5f60f4ebf0b4d0
meillo@93 1165 The
meillo@93 1166 .Sw -[no]inplace
meillo@93 1167 switches of
meillo@93 1168 .Pn repl ,
meillo@93 1169 .Pn forw ,
meillo@93 1170 and
meillo@93 1171 .Pn dist
meillo@93 1172 could be removed, too, as they were simply passed through to
meillo@93 1173 .Pn anno .
meillo@93 1174 .P
meillo@93 1175 .Pn burst
meillo@93 1176 also had
meillo@93 1177 .Sw -[no]inplace
meillo@95 1178 switches, but with different meaning.
meillo@95 1179 With
meillo@95 1180 .Sw -inplace ,
meillo@95 1181 the digest had been replaced by the table of contents (i.e. the
meillo@110 1182 introduction text) and the burst messages were placed right
meillo@95 1183 after this message, renumbering all following messages.
meillo@95 1184 Also, any trailing text of the digest was lost, though,
meillo@95 1185 in practice, it usually consists of an end-of-digest marker only.
meillo@95 1186 Nontheless, this behavior appeared less elegant than the
meillo@95 1187 .Sw -noinplace
meillo@95 1188 behavior, which already had been the default.
meillo@95 1189 Nmh's
meillo@95 1190 .Mp burst (1)
meillo@95 1191 man page reads:
meillo@98 1192 .QS
meillo@164 1193 If
meillo@164 1194 .Sw -noinplace
meillo@164 1195 is given, each digest is preserved, no table
meillo@93 1196 of contents is produced, and the messages contained within
meillo@93 1197 the digest are placed at the end of the folder. Other messages
meillo@93 1198 are not tampered with in any way.
meillo@98 1199 .QE
meillo@95 1200 .LP
meillo@93 1201 The decision to drop the
meillo@93 1202 .Sw -inplace
meillo@95 1203 behavior was supported by the code complexity and the possible data loss
meillo@95 1204 it caused.
meillo@93 1205 .Sw -noinplace
meillo@95 1206 was chosen to be the definitive behavior.
meillo@97 1207 .Ci 68a686adeb39223a5e1ad35e4a24890ec053679d
meillo@93 1208
meillo@95 1209
meillo@95 1210 .U3 "Forms and Format Strings
meillo@93 1211 .P
meillo@95 1212 Historically, the tools that had
meillo@95 1213 .Sw -form
meillo@95 1214 switches to supply a form file had
meillo@95 1215 .Sw -format
meillo@95 1216 switches as well to supply the contents of a form file as a string
meillo@95 1217 on the command line directly.
meillo@95 1218 In consequence, the following two lines equaled:
meillo@95 1219 .VS
meillo@95 1220 scan -form scan.mailx
meillo@95 1221 scan -format "`cat .../scan.mailx`"
meillo@95 1222 VE
meillo@95 1223 The
meillo@95 1224 .Sw -format
meillo@95 1225 switches were dropped in favor for extending the
meillo@95 1226 .Sw -form
meillo@95 1227 switches.
meillo@97 1228 .Ci f51956be123db66b00138f80464d06f030dbb88d
meillo@95 1229 If their argument starts with an equal sign (`='),
meillo@95 1230 then the rest of the argument is taken as a format string,
meillo@95 1231 otherwise the arguments is treated as the name of a format file.
meillo@95 1232 Thus, now the following two lines equal:
meillo@95 1233 .VS
meillo@95 1234 scan -form scan.mailx
meillo@95 1235 scan -form "=`cat .../scan.mailx`"
meillo@95 1236 VE
meillo@95 1237 This rework removed the prefix collision between
meillo@95 1238 .Sw -form
meillo@95 1239 and
meillo@95 1240 .Sw -format .
meillo@95 1241 Now, typing
meillo@95 1242 .Sw -fo
meillo@95 1243 suffices to specify form or format string.
meillo@95 1244 .P
meillo@95 1245 The different meaning of
meillo@95 1246 .Sw -format
meillo@95 1247 for
meillo@95 1248 .Pn repl
meillo@95 1249 and
meillo@95 1250 .Pn forw
meillo@95 1251 was removed in mmh.
meillo@95 1252 .Pn forw
meillo@95 1253 was completely switched to MIME-type forwarding, thus removing the
meillo@95 1254 .Sw -[no]format .
meillo@97 1255 .Ci 6e271608b7b9c23771523f88d23a4d3593010cf1
meillo@95 1256 For
meillo@95 1257 .Pn repl ,
meillo@95 1258 the
meillo@95 1259 .Sw -[no]format
meillo@95 1260 switches were reworked to
meillo@95 1261 .Sw -[no]filter
meillo@95 1262 switches.
meillo@97 1263 .Ci 67411b1f95d6ec987b4c732459e1ba8a8ac192c6
meillo@95 1264 The
meillo@95 1265 .Sw -format
meillo@95 1266 switches of
meillo@95 1267 .Pn send
meillo@95 1268 and
meillo@95 1269 .Pn post ,
meillo@95 1270 which had a third meaning,
meillo@95 1271 were removed likewise.
meillo@97 1272 .Ci f3cb7cde0e6f10451b6848678d95860d512224b9
meillo@95 1273 Eventually, the ambiguity of the
meillo@95 1274 .Sw -format
meillo@95 1275 switches was resolved by not anymore having any such switch in mmh.
meillo@95 1276
meillo@95 1277
meillo@95 1278 .U3 "MIME Tools
meillo@95 1279 .P
meillo@95 1280 The MIME tools, which were once part of
meillo@100 1281 .Pn mhn
meillo@164 1282 .\" XXX
meillo@164 1283 (whatever that stood for),
meillo@95 1284 had several switches that added little practical value to the programs.
meillo@95 1285 The
meillo@95 1286 .Sw -[no]realsize
meillo@95 1287 switches of
meillo@95 1288 .Pn mhbuild
meillo@95 1289 and
meillo@95 1290 .Pn mhlist
meillo@97 1291 were removed, doing real size calculations always now
meillo@97 1292 .Ci 8d8f1c3abc586c005c904e52c4adbfe694d2201c ,
meillo@159 1293 as nmh's
meillo@159 1294 .Mp mhbuild (1)
meillo@159 1295 man page states
meillo@95 1296 ``This provides an accurate count at the expense of a small delay.''
meillo@95 1297 This small delay is not noticable on modern systems.
meillo@95 1298 .P
meillo@95 1299 The
meillo@95 1300 .Sw -[no]check
meillo@95 1301 switches were removed together with the support for
meillo@95 1302 .Hd Content-MD5
meillo@95 1303 header fields.
meillo@95 1304 .[
meillo@95 1305 rfc 1864
meillo@95 1306 .]
meillo@97 1307 .Ci 31dc797eb5178970d68962ca8939da3fd9a8efda
meillo@154 1308 (cf. Sec.
meillo@154 1309 .Cf content-md5 )
meillo@95 1310 .P
meillo@95 1311 The
meillo@95 1312 .Sw -[no]ebcdicsafe
meillo@95 1313 and
meillo@95 1314 .Sw -[no]rfc934mode
meillo@95 1315 switches of
meillo@95 1316 .Pn mhbuild
meillo@95 1317 were removed because they are considered obsolete.
meillo@97 1318 .Ci 01a3480928da485b4d6109d36d751dfa71799d58
meillo@97 1319 .Ci 3363e2624dce0eb8164cf8b3f1ab385c8ff72e88
meillo@95 1320 .P
meillo@95 1321 Content caching of external MIME parts, activated with the
meillo@95 1322 .Sw -rcache
meillo@95 1323 and
meillo@95 1324 .Sw -wcache
meillo@95 1325 switches was completely removed.
meillo@97 1326 .Ci d1fefd9f614e4dc3cda16da6c69133c1b2005269
meillo@97 1327 External MIME parts are rare today, having a caching facility
meillo@159 1328 for them appears to be unnecessary.
meillo@95 1329 .P
meillo@95 1330 In pre-MIME times,
meillo@95 1331 .Pn mhl
meillo@95 1332 had covered many tasks that are part of MIME handling today.
meillo@95 1333 Therefore,
meillo@95 1334 .Pn mhl
meillo@95 1335 could be simplified to a large extend, reducing the number of its
meillo@95 1336 switches from 21 to 6.
meillo@97 1337 .Ci 350ad6d3542a07639213cf2a4fe524e829c1e7b6
meillo@97 1338 .Ci 0e46503be3c855bddaeae3843e1b659279c35d70
meillo@95 1339
meillo@95 1340
meillo@95 1341
meillo@95 1342
meillo@95 1343 .U3 "Header Printing
meillo@95 1344 .P
meillo@95 1345 .Pn folder 's
meillo@95 1346 data output is self-explaining enough that
meillo@159 1347 displaying the header line makes little sense.
meillo@95 1348 Hence, the
meillo@95 1349 .Sw -[no]header
meillo@95 1350 switch was removed and headers are never printed.
meillo@97 1351 .Ci 601cc73d1fa05ce96faa728f036d6c51b91701c7
meillo@95 1352 .P
meillo@95 1353 In
meillo@95 1354 .Pn mhlist ,
meillo@95 1355 the
meillo@95 1356 .Sw -[no]header
meillo@95 1357 switches were removed, too.
meillo@97 1358 .Ci b24f96523aaf60e44e04a3ffb1d22e69a13a602f
meillo@95 1359 But in this case headers are always printed,
meillo@95 1360 because the output is not self-explaining.
meillo@95 1361 .P
meillo@95 1362 .Pn scan
meillo@95 1363 also had
meillo@95 1364 .Sw -[no]header
meillo@95 1365 switches.
meillo@95 1366 Printing the header had been sensible until the introduction of
meillo@95 1367 format strings made it impossible to display the column headings.
meillo@95 1368 Only the folder name and the current date remained to be printed.
meillo@95 1369 As this information can be perfectly retrieved by
meillo@95 1370 .Pn folder
meillo@95 1371 and
meillo@95 1372 .Pn date ,
meillo@95 1373 consequently, the switches were removed.
meillo@97 1374 .Ci c477dc5d1d03fa6d9a8ab3dd3508c63cbddc044e
meillo@95 1375 .P
meillo@95 1376 By removing all
meillo@95 1377 .Sw -header
meillo@95 1378 switches, the collision with
meillo@95 1379 .Sw -help
meillo@95 1380 on the first two letters was resolved.
meillo@95 1381 Currently,
meillo@95 1382 .Sw -h
meillo@95 1383 evaluates to
meillo@95 1384 .Sw -help
meillo@95 1385 for all tools of mmh.
meillo@95 1386
meillo@95 1387
meillo@139 1388 .U3 "Suppressing Edits or the Invocation of the WhatNow Shell
meillo@95 1389 .P
meillo@95 1390 The
meillo@95 1391 .Sw -noedit
meillo@100 1392 switch of
meillo@95 1393 .Pn comp ,
meillo@95 1394 .Pn repl ,
meillo@95 1395 .Pn forw ,
meillo@95 1396 .Pn dist ,
meillo@95 1397 and
meillo@95 1398 .Pn whatnow
meillo@95 1399 was removed, but it can now be replaced by specifying
meillo@95 1400 .Sw -editor
meillo@95 1401 with an empty argument.
meillo@97 1402 .Ci 75fca31a5b9d5c1a99c74ab14c94438d8852fba9
meillo@95 1403 (Specifying
meillo@159 1404 .Cl "-editor /bin/true
meillo@95 1405 is nearly the same, only differing by the previous editor being set.)
meillo@95 1406 .P
meillo@95 1407 The more important change is the removal of the
meillo@95 1408 .Sw -nowhatnowproc
meillo@95 1409 switch.
meillo@97 1410 .Ci ee4f43cf2ef0084ec698e4e87159a94c01940622
meillo@95 1411 This switch had introduced an awkward behavior, as explained in nmh's
meillo@95 1412 man page for
meillo@95 1413 .Mp comp (1):
meillo@98 1414 .QS
meillo@164 1415 The
meillo@164 1416 .Sw -editor
meillo@164 1417 .Ar editor
meillo@164 1418 switch indicates the editor to use for
meillo@164 1419 the initial edit. Upon exiting from the editor,
meillo@164 1420 .Pn comp
meillo@164 1421 will invoke the
meillo@164 1422 .Pn whatnow
meillo@164 1423 program. See
meillo@164 1424 .Mp whatnow (1)
meillo@164 1425 for a discussion of available options.
meillo@164 1426 The invocation of this program can be
meillo@164 1427 inhibited by using the
meillo@164 1428 .Sw -nowhatnowproc
meillo@164 1429 switch. (In truth of fact, it is the
meillo@164 1430 .Pn whatnow
meillo@164 1431 program which starts the initial edit.
meillo@164 1432 Hence,
meillo@164 1433 .Sw -nowhatnowproc
meillo@164 1434 will prevent any edit from occurring.)
meillo@98 1435 .QE
meillo@95 1436 .P
meillo@95 1437 Effectively, the
meillo@95 1438 .Sw -nowhatnowproc
meillo@100 1439 switch creates only a draft message.
meillo@95 1440 As
meillo@159 1441 .Cl "-whatnowproc /bin/true
meillo@95 1442 causes the same behavior, the
meillo@95 1443 .Sw -nowhatnowproc
meillo@95 1444 switch was removed for being redundant.
meillo@100 1445 Likely, the
meillo@95 1446 .Sw -nowhatnowproc
meillo@100 1447 switch was intended to be used by front-ends.
meillo@95 1448
meillo@95 1449
meillo@95 1450
meillo@95 1451 .U3 "Various
meillo@95 1452 .BU
meillo@139 1453 With the removal of MMDF maildrop format support,
meillo@139 1454 .Pn packf
meillo@139 1455 and
meillo@139 1456 .Pn rcvpack
meillo@139 1457 no longer needed their
meillo@139 1458 .Sw -mbox
meillo@139 1459 and
meillo@139 1460 .Sw -mmdf
meillo@139 1461 switches.
meillo@139 1462 .Sw -mbox
meillo@154 1463 is the sole behavior now.
meillo@139 1464 .Ci 3916ab66ad5d183705ac12357621ea8661afd3c0
meillo@139 1465 In the same go,
meillo@139 1466 .Pn packf
meillo@139 1467 and
meillo@139 1468 .Pn rcvpack
meillo@154 1469 were reworked and their
meillo@139 1470 .Sw -file
meillo@139 1471 switch became unnecessary.
meillo@139 1472 .Ci ca1023716d4c2ab890696f3e41fa0d94267a940e
meillo@139 1473
meillo@139 1474 .BU
meillo@139 1475 Mmh's tools will no longer clear the screen (\c
meillo@139 1476 .Pn scan 's
meillo@139 1477 and
meillo@139 1478 .Pn mhl 's
meillo@139 1479 .Sw -[no]clear
meillo@139 1480 switches
meillo@139 1481 .Ci e57b17343dcb3ff373ef4dd089fbe778f0c7c270
meillo@139 1482 .Ci 943765e7ac5693ae177fd8d2b5a2440e53ce816e ).
meillo@139 1483 Neither will
meillo@139 1484 .Pn mhl
meillo@139 1485 ring the bell (\c
meillo@139 1486 .Sw -[no]bell
meillo@139 1487 .Ci e11983f44e59d8de236affa5b0d0d3067c192e24 )
meillo@139 1488 nor page the output itself (\c
meillo@139 1489 .Sw -length
meillo@139 1490 .Ci 5b9d883db0318ed2b84bb82dee880d7381f99188 ).
meillo@159 1491 .\" XXX Ref
meillo@139 1492 Generally, the pager to use is no longer specified with the
meillo@139 1493 .Sw -[no]moreproc
meillo@139 1494 command line switches for
meillo@139 1495 .Pn mhl
meillo@139 1496 and
meillo@139 1497 .Pn show /\c
meillo@139 1498 .Pn mhshow .
meillo@139 1499 .Ci 39e87a75b5c2d3572ec72e717720b44af291e88a
meillo@139 1500
meillo@139 1501 .BU
meillo@96 1502 In order to avoid prefix collisions among switch names, the
meillo@95 1503 .Sw -version
meillo@95 1504 switch was renamed to
meillo@95 1505 .Sw -Version
meillo@95 1506 (with capital `V').
meillo@97 1507 .Ci 32b2354dbaf4bf934936eb5b102a4a3d2fdd209a
meillo@95 1508 Every program has the
meillo@95 1509 .Sw -version
meillo@95 1510 switch but its first three letters collided with the
meillo@95 1511 .Sw -verbose
meillo@95 1512 switch, present in many programs.
meillo@95 1513 The rename solved this problem once for all.
meillo@95 1514 Although this rename breaks a basic interface, having the
meillo@95 1515 .Sw -V
meillo@95 1516 abbreviation to display the version information, isn't all too bad.
meillo@139 1517
meillo@95 1518 .BU
meillo@95 1519 .Sw -[no]preserve
meillo@95 1520 of
meillo@95 1521 .Pn refile
meillo@168 1522 was removed
meillo@168 1523 .Ci 8edc5aaf86f9f77124664f6801bc6c6cdf258173
meillo@168 1524 because what use was it anyway?
meillo@168 1525 Quoting nmh's man page
meillo@165 1526 .Mp refile (1):
meillo@98 1527 .QS
meillo@95 1528 Normally when a message is refiled, for each destination
meillo@95 1529 folder it is assigned the number which is one above the current
meillo@95 1530 highest message number in that folder. Use of the
meillo@164 1531 .Sw -preserv
meillo@164 1532 [sic!] switch will override this message renaming, and try
meillo@95 1533 to preserve the number of the message. If a conflict for a
meillo@164 1534 particular folder occurs when using the
meillo@164 1535 .Sw -preserve
meillo@164 1536 switch, then
meillo@164 1537 .Pn refile
meillo@164 1538 will use the next available message number which
meillo@95 1539 is above the message number you wish to preserve.
meillo@98 1540 .QE
meillo@139 1541
meillo@95 1542 .BU
meillo@95 1543 The removal of the
meillo@95 1544 .Sw -[no]reverse
meillo@95 1545 switches of
meillo@95 1546 .Pn scan
meillo@97 1547 .Ci 8edc5aaf86f9f77124664f6801bc6c6cdf258173
meillo@95 1548 is a bug fix, supported by the comments
meillo@95 1549 ``\-[no]reverse under #ifdef BERK (I really HATE this)''
meillo@95 1550 by Rose and
meillo@95 1551 ``Lists messages in reverse order with the `\-reverse' switch.
meillo@95 1552 This should be considered a bug.'' by Romine in the documentation.
meillo@159 1553 .\" XXX Ref: welche datei genau.
meillo@97 1554 The question remains why neither Rose and Romine had fixed this
meillo@109 1555 bug in the eighties when they wrote these comments nor has anyone
meillo@95 1556 thereafter.
meillo@93 1557
meillo@93 1558
meillo@93 1559 .ig
meillo@93 1560
meillo@95 1561 forw: [no]dashstuffing(mhl)
meillo@93 1562
meillo@95 1563 mhshow: [no]pause [no]serialonly
meillo@93 1564
meillo@93 1565 mhmail: resent queued
meillo@93 1566 inc: snoop, (pop)
meillo@93 1567
meillo@95 1568 mhl: [no]faceproc folder sleep
meillo@95 1569 [no]dashstuffing(forw) digest list volume number issue number
meillo@93 1570
meillo@95 1571 prompter: [no]doteof
meillo@93 1572
meillo@93 1573 refile: [no]preserve [no]unlink [no]rmmproc
meillo@93 1574
meillo@95 1575 send: [no]forward [no]mime [no]msgid
meillo@93 1576 [no]push split [no]unique (sasl) width snoop [no]dashstuffing
meillo@93 1577 attach attachformat
meillo@93 1578 whatnow: (noedit) attach
meillo@93 1579
meillo@93 1580 slocal: [no]suppressdups
meillo@93 1581
meillo@95 1582 spost: [no]filter [no]backup width [no]push idanno
meillo@93 1583 [no]check(whom) whom(whom)
meillo@93 1584
meillo@93 1585 whom: ???
meillo@93 1586
meillo@95 1587 ..
meillo@93 1588
meillo@93 1589
meillo@93 1590 .ig
meillo@93 1591
meillo@93 1592 .P
meillo@93 1593 In the best case, all switches are unambiguous on the first character,
meillo@93 1594 or on the three-letter prefix for the `no' variants.
meillo@96 1595 Reducing switch prefix collisions, shortens the necessary prefix length
meillo@93 1596 the user must type.
meillo@93 1597 Having less switches helps best.
meillo@93 1598
meillo@93 1599 ..
meillo@58 1600
meillo@95 1601
meillo@102 1602 .\" XXX: whatnow prompt commands
meillo@102 1603
meillo@102 1604
meillo@95 1605
meillo@95 1606
meillo@133 1607 .\" --------------------------------------------------------------
meillo@74 1608 .H1 "Modernizing
meillo@102 1609 .P
meillo@164 1610 In the more than thirty years of MH's existence, its code base was
meillo@159 1611 increasingly extended.
meillo@118 1612 New features entered the project and became alternatives to the
meillo@118 1613 existing behavior.
meillo@118 1614 Relicts from several decades have gathered in the code base,
meillo@118 1615 but seldom obsolete features were dropped.
meillo@118 1616 This section describes the removing of old code
meillo@118 1617 and the modernizing of the default setup.
meillo@118 1618 It focuses on the functional aspect only;
meillo@154 1619 the non-functional aspects of code style are discussed in Sec.
meillo@154 1620 .Cf code-style .
meillo@58 1621
meillo@58 1622
meillo@100 1623 .H2 "Code Relicts
meillo@0 1624 .P
meillo@159 1625 My position regarding the removal of obsolete functions of mmh,
meillo@159 1626 .\" XXX ``in order to remove old code,''
meillo@159 1627 is much more revolutional than the nmh community appreciates.
meillo@159 1628 Working on an experimental version, I was quickly able to drop
meillo@104 1629 functionality I considered ancient.
meillo@104 1630 The need for consensus with peers would have slowed this process down.
meillo@104 1631 Without the need to justify my decisions, I was able to rush forward.
meillo@110 1632 In December 2011, Paul Vixie motivated the nmh developers to just
meillo@159 1633 .\" XXX ugs
meillo@104 1634 do the work:
meillo@104 1635 .[
meillo@104 1636 paul vixie edginess nmh-workers
meillo@104 1637 .]
meillo@104 1638 .QS
meillo@104 1639 let's stop walking on egg shells with this code base. there's no need to
meillo@104 1640 discuss whether to keep using vfork, just note in [sic!] passing, [...]
meillo@104 1641 we don't need a separate branch for removing vmh
meillo@104 1642 or ridding ourselves of #ifdef's or removing posix replacement functions
meillo@164 1643 or depending on pure ansi/posix ``libc''.
meillo@104 1644 .QP
meillo@164 1645 these things should each be a day or two of work and the ``main branch''
meillo@104 1646 should just be modern. [...]
meillo@104 1647 let's push forward, aggressively.
meillo@104 1648 .QE
meillo@104 1649 .LP
meillo@104 1650 I did so already in the months before.
meillo@104 1651 I pushed forward.
meillo@159 1652 .\" XXX semicolon ?
meillo@104 1653 I simply dropped the cruft.
meillo@104 1654 .P
meillo@104 1655 The decision to drop a feature was based on literature research and
meillo@159 1656 careful thinking, but whether having had contact with this particular
meillo@104 1657 feature within my own computer life served as a rule of thumb.
meillo@159 1658 I explained my reasons in the commit messages
meillo@109 1659 in the version control system.
meillo@104 1660 Hence, others can comprehend my view and argue for undoing the change
meillo@104 1661 if I have missed an important aspect.
meillo@109 1662 I was quick in dropping parts.
meillo@159 1663 I rather re-included falsely dropped parts than going at a slower pace.
meillo@109 1664 Mmh is experimental work; it required tough decisions.
meillo@159 1665 .\" XXX ``exp. work'' schon oft gesagt
meillo@12 1666
meillo@102 1667
meillo@104 1668 .U3 "Forking
meillo@12 1669 .P
meillo@109 1670 Being a tool chest, MH creates many processes.
meillo@104 1671 In earlier times
meillo@104 1672 .Fu fork()
meillo@104 1673 had been an expensive system call, because the process's image needed
meillo@159 1674 to be completely duplicated at once.
meillo@159 1675 This expensive work was especially unnecessary in the commonly occuring
meillo@159 1676 case wherein the image is replaced by a call to
meillo@104 1677 .Fu exec()
meillo@104 1678 right after having forked the child process.
meillo@104 1679 The
meillo@104 1680 .Fu vfork()
meillo@104 1681 system call was invented to speed up this particular case.
meillo@104 1682 It completely omits the duplication of the image.
meillo@104 1683 On old systems this resulted in significant speed ups.
meillo@104 1684 Therefore MH used
meillo@104 1685 .Fu vfork()
meillo@104 1686 whenever possible.
meillo@12 1687 .P
meillo@104 1688 Modern memory management units support copy-on-write semantics, which make
meillo@104 1689 .Fu fork()
meillo@104 1690 almost as fast as
meillo@104 1691 .Fu vfork() .
meillo@104 1692 The man page of
meillo@104 1693 .Mp vfork (2)
meillo@104 1694 in FreeBSD 8.0 states:
meillo@104 1695 .QS
meillo@104 1696 This system call will be eliminated when proper system sharing mechanisms
meillo@104 1697 are implemented. Users should not depend on the memory sharing semantics
meillo@104 1698 of vfork() as it will, in that case, be made synonymous to fork(2).
meillo@104 1699 .QE
meillo@104 1700 .LP
meillo@104 1701 Vixie supports the removal with the note that ``the last
meillo@104 1702 system on which fork was so slow that an mh user would notice it, was
meillo@104 1703 Eunice. that was 1987''.
meillo@104 1704 .[
meillo@104 1705 nmh-workers vixie edginess
meillo@104 1706 .]
meillo@104 1707 I replaced all calls to
meillo@104 1708 .Fu vfork()
meillo@104 1709 with calls to
meillo@104 1710 .Fu fork() .
meillo@109 1711 .Ci 40821f5c1316e9205a08375e7075909cc9968e7d
meillo@104 1712 .P
meillo@104 1713 Related to the costs of
meillo@104 1714 .Fu fork()
meillo@104 1715 is the probability of its success.
meillo@109 1716 In the eighties, on heavy loaded systems, calls to
meillo@104 1717 .Fu fork()
meillo@104 1718 were prone to failure.
meillo@104 1719 Hence, many of the
meillo@104 1720 .Fu fork()
meillo@104 1721 calls in the code were wrapped into loops to retry the
meillo@104 1722 .Fu fork()
meillo@159 1723 several times, to increase the chances to succeed, eventually.
meillo@109 1724 On modern systems, a failing
meillo@104 1725 .Fu fork()
meillo@109 1726 call is unusual.
meillo@104 1727 Hence, in the rare case when
meillo@104 1728 .Fu fork()
meillo@104 1729 fails, mmh programs simply abort.
meillo@109 1730 .Ci 5fbf37ee68e018998ada61eeab73e035b26834b6
meillo@12 1731
meillo@12 1732
meillo@109 1733 .U3 "Header Fields
meillo@104 1734 .BU
meillo@84 1735 The
meillo@84 1736 .Hd Encrypted
meillo@104 1737 header field was introduced by RFC\|822,
meillo@109 1738 but already marked as legacy in RFC\|2822.
meillo@109 1739 Today, OpenPGP provides the basis for standardized exchange of encrypted
meillo@104 1740 messages [RFC\|4880, RFC\|3156].
meillo@109 1741 Hence, the support for
meillo@104 1742 .Hd Encrypted
meillo@104 1743 header fields is removed in mmh.
meillo@109 1744 .Ci 064527f7b57ab050e5af13e15ad99aeeab125857
meillo@104 1745 .BU
meillo@159 1746 The native support for
meillo@84 1747 .Hd Face
meillo@104 1748 header fields has been removed, as well.
meillo@109 1749 .Ci 8e5be81f784682822f5e868c1bf3c8624682bd23
meillo@104 1750 This feature is similar to the
meillo@84 1751 .Hd X-Face
meillo@84 1752 header field in its intent,
meillo@21 1753 but takes a different approach to store the image.
meillo@84 1754 Instead of encoding the image data directly into the header field,
meillo@109 1755 it contains the hostname and UDP port where the image
meillo@109 1756 date can be retrieved.
meillo@159 1757 There is even a third Face system,
meillo@109 1758 which is the successor of
meillo@109 1759 .Hd X-Face ,
meillo@109 1760 although it re-uses the
meillo@104 1761 .Hd Face
meillo@109 1762 header field.
meillo@109 1763 It was invented in 2005 and supports colored PNG images.
meillo@104 1764 None of the Face systems described here is popular today.
meillo@104 1765 Hence, mmh has no direct support for them.
meillo@104 1766 .BU
meillo@154 1767 .Id content-md5
meillo@104 1768 The
meillo@104 1769 .Hd Content-MD5
meillo@104 1770 header field was introduced by RFC\|1864.
meillo@104 1771 It provides detection of data corruption during the transfer.
meillo@104 1772 But it can not ensure verbatim end-to-end delivery of the contents
meillo@104 1773 [RFC\|1864].
meillo@104 1774 The proper approach to verify content integrity in an
meillo@166 1775 end-to-end relationship is the use of digital signatures.
meillo@104 1776 .\" XXX (RFCs FIXME).
meillo@104 1777 On the other hand, transfer protocols should detect corruption during
meillo@109 1778 the transmission.
meillo@109 1779 The TCP includes a checksum field therefore.
meillo@104 1780 These two approaches in combinations render the
meillo@104 1781 .Hd Content-MD5
meillo@104 1782 header field superfluous.
meillo@109 1783 Not a single one out of 4\|200 messages from two decades
meillo@109 1784 in an nmh-workers mailing list archive contains a
meillo@104 1785 .Hd Content-MD5
meillo@104 1786 header field.
meillo@104 1787 Neither did any of the 60\|000 messages in my personal mail storage.
meillo@104 1788 Removing the support for this header field,
meillo@104 1789 removed the last place where MD5 computation was needed.
meillo@109 1790 .Ci 31dc797eb5178970d68962ca8939da3fd9a8efda
meillo@104 1791 Hence, the MD5 code could be removed as well.
meillo@104 1792 Over 500 lines of code vanished by this one change.
meillo@104 1793
meillo@104 1794
meillo@104 1795 .U3 "MMDF maildrop support
meillo@21 1796 .P
meillo@104 1797 This type of format is conceptionally similar to the mbox format,
meillo@139 1798 but uses a different message delimiter (`\fL\\1\\1\\1\\1\fP',
meillo@139 1799 commonly written as `\fL^A^A^A^A\fP', instead of `\fLFrom\0\fP').
meillo@104 1800 Mbox is the de-facto standard maildrop format on Unix,
meillo@159 1801 whereas the MMDF maildrop format is now forgotten.
meillo@159 1802 By dropping the MMDF maildrop format support,
meillo@159 1803 mbox became the only packed mailbox format supported in mmh.
meillo@104 1804 .P
meillo@109 1805 The simplifications within the code were moderate.
meillo@109 1806 Mainly, the reading and writing of MMDF mailbox files was removed.
meillo@109 1807 But also, switches of
meillo@109 1808 .Pn packf
meillo@104 1809 and
meillo@109 1810 .Pn rcvpack
meillo@109 1811 could be removed.
meillo@109 1812 .Ci 3916ab66ad5d183705ac12357621ea8661afd3c0
meillo@109 1813 In the message parsing function
meillo@109 1814 .Fn sbr/m_getfld.c ,
meillo@109 1815 knowledge of MMDF packed mail boxes was removed.
meillo@109 1816 .Ci 684ec30d81e1223a282764452f4902ed4ad1c754
meillo@109 1817 Further code structure simplifications may be possible there,
meillo@109 1818 because only one single packed mailbox format is left to be supported.
meillo@104 1819 I have not worked on them yet because
meillo@104 1820 .Fu m_getfld()
meillo@104 1821 is heavily optimized and thus dangerous to touch.
meillo@104 1822 The risk of damaging the intricate workings of the optimized code is
meillo@104 1823 too high.
meillo@104 1824
meillo@12 1825
meillo@101 1826 .U3 "Prompter's Control Keys
meillo@20 1827 .P
meillo@20 1828 The program
meillo@20 1829 .Pn prompter
meillo@104 1830 queries the user to fill in a message form.
meillo@104 1831 When used by
meillo@20 1832 .Pn comp
meillo@104 1833 as
meillo@104 1834 .Cl "comp -editor prompter" ,
meillo@20 1835 the resulting behavior is similar to
meillo@20 1836 .Pn mailx .
meillo@51 1837 Apparently,
meillo@20 1838 .Pn prompter
meillo@104 1839 hadn't been touched lately.
meillo@104 1840 Otherwise it's hardly explainable why it
meillo@20 1841 still offered the switches
meillo@84 1842 .Sw -erase
meillo@84 1843 .Ar chr
meillo@20 1844 and
meillo@84 1845 .Sw -kill
meillo@84 1846 .Ar chr
meillo@20 1847 to name the characters for command line editing.
meillo@21 1848 The times when this had been necessary are long time gone.
meillo@20 1849 Today these things work out-of-the-box, and if not, are configured
meillo@20 1850 with the standard tool
meillo@20 1851 .Pn stty .
meillo@104 1852 The switches are removed now
meillo@104 1853 .Ci 0bd9750710cdbab80cfb4036dd87af20afe1552f .
meillo@20 1854
meillo@104 1855
meillo@109 1856 .U3 "Hardcopy Terminal Support
meillo@21 1857 .P
meillo@109 1858 More of a funny anecdote is a check for being connected to a
meillo@109 1859 hardcopy terminal.
meillo@159 1860 It remained in the code until spring 2012, when I finally removed it
meillo@104 1861 .Ci b7764c4a6b71d37918a97594d866258f154017ca .
meillo@21 1862 .P
meillo@109 1863 The check only prevented a pager to be placed between the printing
meillo@104 1864 program (\c
meillo@104 1865 .Pn mhl )
meillo@104 1866 and the terminal.
meillo@109 1867 In nmh, this could have been ensured statically with the
meillo@104 1868 .Sw -nomoreproc
meillo@109 1869 at the command line, too.
meillo@121 1870 In mmh, setting the profile entry
meillo@104 1871 .Pe Pager
meillo@104 1872 or the environment variable
meillo@104 1873 .Ev PAGER
meillo@104 1874 to
meillo@109 1875 .Pn cat
meillo@159 1876 is sufficient.
meillo@104 1877
meillo@104 1878
meillo@21 1879
meillo@12 1880
meillo@58 1881 .H2 "Attachments
meillo@22 1882 .P
meillo@101 1883 The mind model of email attachments is unrelated to MIME.
meillo@101 1884 Although the MIME RFCs (2045 through 2049) define the technical
meillo@109 1885 requirements for having attachments, they do not mention the word
meillo@101 1886 ``attachment''.
meillo@101 1887 Instead of attachments, MIME talks about ``multi-part message bodies''
meillo@101 1888 [RFC\|2045], a more general concept.
meillo@101 1889 Multi-part messages are messages
meillo@101 1890 ``in which one or more different
meillo@101 1891 sets of data are combined in a single body''
meillo@101 1892 [RFC\|2046].
meillo@101 1893 MIME keeps its descriptions generic;
meillo@101 1894 it does not imply specific usage models.
meillo@109 1895 One usage model became prevalent: attachments.
meillo@101 1896 The idea is having a main text document with files of arbitrary kind
meillo@101 1897 attached to it.
meillo@101 1898 In MIME terms, this is a multi-part message having a text part first
meillo@110 1899 and parts of arbitrary type following.
meillo@101 1900 .P
meillo@101 1901 MH's MIME support is a direct implementation of the RFCs.
meillo@101 1902 The perception of the topic described in the RFCs is clearly visible
meillo@101 1903 in MH's implementation.
meillo@159 1904 .\" XXX rewrite ``no idea''.
meillo@159 1905 As a result,
meillo@159 1906 MH had all the MIME features but no idea of attachments.
meillo@109 1907 But users don't need all the MIME features,
meillo@109 1908 they want convenient attachment handling.
meillo@109 1909
meillo@102 1910
meillo@102 1911 .U3 "Composing MIME Messages
meillo@102 1912 .P
meillo@102 1913 In order to improve the situation on the message composing side,
meillo@102 1914 Jon Steinhart had added an attachment system to nmh in 2002.
meillo@101 1915 .Ci 7480dbc14bc90f2d872d434205c0784704213252
meillo@102 1916 In the file
meillo@102 1917 .Fn docs/README-ATTACHMENTS ,
meillo@102 1918 he described his motivation to do so as such:
meillo@101 1919 .QS
meillo@159 1920 Although nmh contains the necessary functionality for MIME message
meillo@159 1921 handing [sic!], the interface to this functionality is pretty obtuse.
meillo@101 1922 There's no way that I'm ever going to convince my partner to write
meillo@101 1923 .Pn mhbuild
meillo@101 1924 composition files!
meillo@101 1925 .QE
meillo@102 1926 .LP
meillo@102 1927 With this change, the mind model of attachments entered nmh.
meillo@102 1928 In the same document:
meillo@101 1929 .QS
meillo@101 1930 These changes simplify the task of managing attachments on draft files.
meillo@101 1931 They allow attachments to be added, listed, and deleted.
meillo@101 1932 MIME messages are automatically created when drafts with attachments
meillo@101 1933 are sent.
meillo@101 1934 .QE
meillo@102 1935 .LP
meillo@102 1936 Unfortunately, the attachment system,
meillo@102 1937 like any new facilities in nmh,
meillo@110 1938 was inactive by default.
meillo@101 1939 .P
meillo@101 1940 During my work in Argentina, I tried to improve the attachment system.
meillo@102 1941 But, because of great opposition in the nmh community,
meillo@102 1942 my patch died as a proposal on the mailing list, after long discussions.
meillo@101 1943 .[
meillo@101 1944 nmh-workers attachment proposal
meillo@101 1945 .]
meillo@110 1946 In January 2012, I extended the patch and applied it to mmh.
meillo@101 1947 .Ci 8ff284ff9167eff8f5349481529332d59ed913b1
meillo@102 1948 In mmh, the attachment system is active by default.
meillo@102 1949 Instead of command line switches, the
meillo@102 1950 .Pe Attachment-Header
meillo@102 1951 profile entry is used to specify
meillo@102 1952 the name of the attachment header field.
meillo@102 1953 It is pre-defined to
meillo@102 1954 .Hd Attach .
meillo@101 1955 .P
meillo@159 1956 To add an attachment to a draft, a header line needs to be added:
meillo@101 1957 .VS
meillo@101 1958 To: bob
meillo@101 1959 Subject: The file you wanted
meillo@101 1960 Attach: /path/to/the/file-bob-wanted
meillo@101 1961 --------
meillo@101 1962 Here it is.
meillo@101 1963 VE
meillo@101 1964 The header field can be added to the draft manually in the editor,
meillo@102 1965 or by using the `attach' command at the WhatNow prompt, or
meillo@102 1966 non-interactively with
meillo@101 1967 .Pn anno :
meillo@101 1968 .VS
meillo@102 1969 anno -append -nodate -component Attach -text /path/to/attachment
meillo@101 1970 VE
meillo@102 1971 Drafts with attachment headers are converted to MIME automatically by
meillo@102 1972 .Pn send .
meillo@102 1973 The conversion to MIME is invisible to the user.
meillo@159 1974 The draft stored in the draft folder is always in source form with
meillo@101 1975 attachment headers.
meillo@101 1976 If the MIMEification fails, for instance because the file to attach
meillo@101 1977 is not accessible, the original draft is not changed.
meillo@101 1978 .P
meillo@102 1979 The attachment system handles the forwarding of messages, too.
meillo@101 1980 If the attachment header value starts with a plus character (`+'),
meillo@101 1981 like in
meillo@101 1982 .Cl "Attach: +bob 30 42" ,
meillo@159 1983 the given messages in the specified folder will be attached.
meillo@101 1984 This allowed to simplify
meillo@101 1985 .Pn forw .
meillo@101 1986 .Ci f41f04cf4ceca7355232cf7413e59afafccc9550
meillo@101 1987 .P
meillo@101 1988 Closely related to attachments is non-ASCII text content,
meillo@101 1989 because it requires MIME too.
meillo@102 1990 In nmh, the user needed to call `mime' at the WhatNow prompt
meillo@101 1991 to have the draft converted to MIME.
meillo@102 1992 This was necessary whenever the draft contained non-ASCII characters.
meillo@101 1993 If the user did not call `mime', a broken message would be sent.
meillo@101 1994 Therefore, the
meillo@101 1995 .Pe automimeproc
meillo@101 1996 profile entry could be specified to have the `mime' command invoked
meillo@102 1997 automatically each time.
meillo@159 1998 Unfortunately, this approach conflicted with attachment system
meillo@101 1999 because the draft would already be in MIME format at the time
meillo@101 2000 when the attachment system wanted to MIMEify it.
meillo@102 2001 To use nmh's attachment system, `mime' must not be called at the
meillo@102 2002 WhatNow prompt and
meillo@101 2003 .Pe automimeproc
meillo@102 2004 must not be set in the profile.
meillo@101 2005 But then the case of non-ASCII text without attachment headers was
meillo@101 2006 not caught.
meillo@102 2007 All in all, the solution was complex and irritating.
meillo@168 2008 My patch from December 2010
meillo@168 2009 .[
meillo@168 2010 nmh-workers attachment proposal
meillo@168 2011 .]
meillo@168 2012 would have simplified the situation.
meillo@102 2013 .P
meillo@101 2014 Mmh's current solution is even more elaborate.
meillo@101 2015 Any necessary MIMEification is done automatically.
meillo@101 2016 There is no `mime' command at the WhatNow prompt anymore.
meillo@102 2017 The draft will be converted automatically to MIME when either an
meillo@102 2018 attachment header or non-ASCII text is present.
meillo@159 2019 Furthermore, the hash character (`#') is not special any more
meillo@159 2020 at line beginnings in the draft message.
meillo@159 2021 .\" XXX REF ?
meillo@159 2022 Users need not concern themselves with the whole topic at all.
meillo@101 2023 .P
meillo@102 2024 Although the new approach does not anymore support arbitrary MIME
meillo@102 2025 compositions directly, the full power of
meillo@101 2026 .Pn mhbuild
meillo@101 2027 can still be accessed.
meillo@102 2028 Given no attachment headers are included, the user can create
meillo@101 2029 .Pn mhbuild
meillo@102 2030 composition drafts like in nmh.
meillo@101 2031 Then, at the WhatNow prompt, he needs to invoke
meillo@101 2032 .Cl "edit mhbuild
meillo@101 2033 to convert it to MIME.
meillo@110 2034 Because the resulting draft does neither contain non-ASCII characters
meillo@102 2035 nor has it attachment headers, the attachment system will not touch it.
meillo@101 2036 .P
meillo@159 2037 The approach taken in mmh is tailored towards today's most common case:
meillo@159 2038 a text part, possibly with attachments.
meillo@159 2039 This case was simplified.
meillo@102 2040
meillo@112 2041
meillo@102 2042 .U3 "MIME Type Guessing
meillo@102 2043 .P
meillo@159 2044 From the programmer's point of view, the use of
meillo@101 2045 .Pn mhbuild
meillo@159 2046 composition drafts had one notable advantage over attachment headers:
meillo@159 2047 The user provides the appropriate MIME types for files to include.
meillo@102 2048 The attachment system needs to find out the correct MIME type itself.
meillo@102 2049 This is a difficult task, yet it spares the user irritating work.
meillo@102 2050 Determining the correct MIME type of content is partly mechanical,
meillo@102 2051 partly intelligent work.
meillo@102 2052 Forcing the user to find out the correct MIME type,
meillo@102 2053 forces him to do partly mechanical work.
meillo@102 2054 Letting the computer do the work, can lead to bad choices for difficult
meillo@102 2055 content.
meillo@102 2056 For mmh, the latter option was chosen.
meillo@102 2057 .P
meillo@102 2058 Determining the MIME type by the suffix of the file name is a dumb
meillo@102 2059 approach, yet it is simple to implement and provides good results
meillo@102 2060 for the common cases.
meillo@102 2061 Mmh implements this approach in the
meillo@102 2062 .Pn print-mimetype
meillo@102 2063 script.
meillo@112 2064 .Ci 4b5944268ea0da7bb30598a27857304758ea9b44
meillo@102 2065 Using it is the default choice.
meillo@102 2066 .P
meillo@112 2067 A far better, though less portable, approach is the use of
meillo@102 2068 .Pn file .
meillo@102 2069 This standard tool tries to determine the type of files.
meillo@102 2070 Unfortunately, its capabilities and accuracy varies from system to system.
meillo@102 2071 Additionally, its output was only intended for human beings,
meillo@102 2072 but not to be used by programs.
meillo@102 2073 It varies much.
meillo@102 2074 Nevertheless, modern versions of GNU
meillo@102 2075 .Pn file ,
meillo@102 2076 which is prevalent on the popular GNU/Linux systems,
meillo@159 2077 provide MIME type output in machine-readable form.
meillo@102 2078 Although this solution is highly system-dependent,
meillo@102 2079 it solves the difficult problem well.
meillo@102 2080 On systems where GNU
meillo@102 2081 .Pn file ,
meillo@102 2082 version 5.04 or higher, is available it should be used.
meillo@102 2083 One needs to specify the following profile entry to do so:
meillo@112 2084 .Ci 3baec236a39c5c89a9bda8dbd988d643a21decc6
meillo@102 2085 .VS
meillo@102 2086 Mime-Type-Query: file -b --mime
meillo@102 2087 VE
meillo@102 2088 .LP
meillo@102 2089 Other versions of
meillo@102 2090 .Pn file
meillo@102 2091 might possibly be usable with wrapper scripts to reformat the output.
meillo@102 2092 The diversity among
meillo@102 2093 .Pn file
meillo@102 2094 implementations is great; one needs to check the local variant.
meillo@102 2095 .P
meillo@102 2096 If no MIME type can be determined, text content gets sent as
meillo@102 2097 `text/plain' and anything else under the generic fall-back type
meillo@102 2098 `application/octet-stream'.
meillo@102 2099 It is not possible in mmh to override the automatic MIME type guessing
meillo@102 2100 for a specific file.
meillo@159 2101 To do so, either the user would need to know in advance for which file
meillo@159 2102 the automatic guessing fails, or the system would require interaction.
meillo@102 2103 I consider both cases impractical.
meillo@102 2104 The existing solution should be sufficient.
meillo@102 2105 If not, the user may always fall back to
meillo@102 2106 .Pn mhbuild
meillo@102 2107 composition drafts and ignore the attachment system.
meillo@101 2108
meillo@102 2109
meillo@102 2110 .U3 "Storing Attachments
meillo@102 2111 .P
meillo@169 2112 Extracting MIME parts of a message and storing them to disk is performed by
meillo@108 2113 .Pn mhstore .
meillo@108 2114 The program has two operation modes,
meillo@108 2115 .Sw -auto
meillo@108 2116 and
meillo@108 2117 .Sw -noauto .
meillo@108 2118 With the former one, each part is stored under the filename given in the
meillo@108 2119 MIME part's meta information, if available.
meillo@108 2120 This naming information is usually available for modern attachments.
meillo@108 2121 If no filename is available, this MIME part is stored as if
meillo@108 2122 .Sw -noauto
meillo@108 2123 would have been specified.
meillo@108 2124 In the
meillo@108 2125 .Sw -noauto
meillo@108 2126 mode, the parts are processed according to rules, defined by
meillo@108 2127 .Pe mhstore-store-*
meillo@108 2128 profile entries.
meillo@108 2129 These rules define generic filename templates for storing
meillo@108 2130 or commands to post-process the contents in arbitrary ways.
meillo@108 2131 If no matching rule is available the part is stored under a generic
meillo@108 2132 filename, built from message number, MIME part number, and MIME type.
meillo@108 2133 .P
meillo@108 2134 The
meillo@108 2135 .Sw -noauto
meillo@108 2136 mode had been the default in nmh because it was considered safe,
meillo@108 2137 in contrast to the
meillo@108 2138 .Sw -auto
meillo@108 2139 mode.
meillo@108 2140 In mmh,
meillo@108 2141 .Sw -auto
meillo@108 2142 is not dangerous anymore.
meillo@108 2143 Two changes were necessary:
meillo@108 2144 .BU
meillo@108 2145 Any directory path is removed from the proposed filename.
meillo@108 2146 Thus, the files are always stored in the expected directory.
meillo@108 2147 .Ci 41b6eadbcecf63c9a66aa5e582011987494abefb
meillo@108 2148 .BU
meillo@108 2149 Tar files are not extracted automatically any more.
meillo@108 2150 Thus, the rest of the file system will not be touched.
meillo@108 2151 .Ci 94c80042eae3383c812d9552089953f9846b1bb6
meillo@108 2152 .LP
meillo@108 2153 Now, the outcome of mmh's
meillo@108 2154 .Cl "mhstore -auto
meillo@110 2155 can be foreseen from the output of
meillo@108 2156 .Cl "mhlist -verbose" .
meillo@108 2157 .P
meillo@108 2158 The
meillo@108 2159 .Sw -noauto
meillo@108 2160 mode is seen to be more powerful but less convenient.
meillo@108 2161 On the other hand,
meillo@108 2162 .Sw -auto
meillo@108 2163 is safe now and
meillo@108 2164 storing attachments under their original name is intuitive.
meillo@108 2165 Hence,
meillo@108 2166 .Sw -auto
meillo@108 2167 serves better as the default option.
meillo@108 2168 .Ci 3410b680416c49a7617491af38bc1929855a331d
meillo@108 2169 .P
meillo@108 2170 Files are stored into the directory given by the
meillo@108 2171 .Pe Nmh-Storage
meillo@108 2172 profile entry, if set, or
meillo@108 2173 into the current working directory, otherwise.
meillo@108 2174 Storing to different directories is only possible with
meillo@108 2175 .Pe mhstore-store-*
meillo@108 2176 profile entries.
meillo@108 2177 .P
meillo@108 2178 Still, in both modes, existing files get overwritten silently.
meillo@108 2179 This can be considered a bug.
meillo@108 2180 Yet, each other behavior has its draw-backs, too.
meillo@108 2181 Refusing to replace files requires adding a
meillo@108 2182 .Sw -force
meillo@108 2183 option.
meillo@108 2184 Users will likely need to invoke
meillo@108 2185 .Pn mhstore
meillo@108 2186 a second time with
meillo@159 2187 .Sw -force .
meillo@159 2188 Eventually, only the user can decide in the specific case.
meillo@108 2189 This requires interaction, which I like to avoid if possible.
meillo@108 2190 Appending a unique suffix to the filename is another bad option.
meillo@108 2191 For now, the behavior remains as it is.
meillo@108 2192 .P
meillo@108 2193 In mmh, only MIME parts of type message are special in
meillo@108 2194 .Pn mhstore 's
meillo@108 2195 .Sw -auto
meillo@108 2196 mode.
meillo@108 2197 Instead of storing message/rfc822 parts as files to disk,
meillo@108 2198 they are stored as messages into the current mail folder.
meillo@159 2199 The same applies to message/partial, although the parts are
meillo@159 2200 automatically reassembled beforehand.
meillo@159 2201 MIME parts of type message/external-body are not automatically retrieved
meillo@159 2202 anymore.
meillo@159 2203 Instead, information on how to retrieve them is output.
meillo@108 2204 Not supporting this rare case saved nearly one thousand lines of code.
meillo@108 2205 .Ci 55e1d8c654ee0f7c45b9361ce34617983b454c32
meillo@108 2206 .\" XXX mention somewhere else too: (The profile entry `nmh-access-ftp'
meillo@108 2207 .\" and sbr/ruserpass.c for reading ~/.netrc are gone now.)
meillo@159 2208 `application/octet-stream; type=tar' is not special anymore.
meillo@108 2209 Automatically extracting such MIME parts had been the dangerous part
meillo@108 2210 of the
meillo@108 2211 .Sw -auto
meillo@108 2212 mode.
meillo@108 2213 .Ci 94c80042eae3383c812d9552089953f9846b1bb6
meillo@108 2214
meillo@102 2215
meillo@102 2216
meillo@102 2217 .U3 "Showing MIME Messages
meillo@102 2218 .P
meillo@114 2219 The program
meillo@114 2220 .Pn mhshow
meillo@114 2221 had been written to display MIME messages.
meillo@114 2222 It implemented the conceptional view of the MIME RFCs.
meillo@114 2223 Nmh's
meillo@114 2224 .Pn mhshow
meillo@114 2225 handled each MIME part independently, presenting them separately
meillo@114 2226 to the user.
meillo@114 2227 This does not match today's understanding of email attachments,
meillo@114 2228 where displaying a message is seen to be a single, integrated operation.
meillo@114 2229 Today, email messages are expected to consist of a main text part
meillo@114 2230 plus possibly attachments.
meillo@114 2231 They are not any more seen to be arbitrary MIME hierarchies with
meillo@114 2232 information on how to display the individual parts.
meillo@114 2233 I adjusted
meillo@114 2234 .Pn mhshow 's
meillo@114 2235 behavior to the modern view on the topic.
meillo@114 2236 .P
meillo@159 2237 One should note that this section completely ignores the original
meillo@114 2238 .Pn show
meillo@114 2239 program, because it was not capable to display MIME messages
meillo@114 2240 and is no longer part of mmh.
meillo@114 2241 Although
meillo@114 2242 .Pn mhshow
meillo@114 2243 was renamed to
meillo@114 2244 .Pn show
meillo@114 2245 in mmh, this section uses the name
meillo@114 2246 .Pn mhshow ,
meillo@114 2247 in order to avoid confusion.
meillo@159 2248 .\" XXX ref to other section
meillo@114 2249 .P
meillo@114 2250 In mmh, the basic idea is that
meillo@114 2251 .Pn mhshow
meillo@114 2252 should display a message in one single pager session.
meillo@114 2253 Therefore,
meillo@114 2254 .Pn mhshow
meillo@114 2255 invokes a pager session for all its output,
meillo@114 2256 whenever it prints to a terminal.
meillo@114 2257 .Ci a4197ea6ffc5c1550e8b52d5a654bcaaaee04a4e
meillo@114 2258 In consequence,
meillo@114 2259 .Pn mhl
meillo@114 2260 does no more invoke a pager.
meillo@114 2261 .Ci 0e46503be3c855bddaeae3843e1b659279c35d70
meillo@114 2262 With
meillo@114 2263 .Pn mhshow
meillo@114 2264 replacing the original
meillo@114 2265 .Pn show ,
meillo@114 2266 output from
meillo@114 2267 .Pn mhl
meillo@114 2268 does not go to the terminal directly, but through
meillo@114 2269 .Pn mhshow .
meillo@114 2270 Hence,
meillo@114 2271 .Pn mhl
meillo@114 2272 does not need to invoke a pager.
meillo@114 2273 The one and only job of
meillo@114 2274 .Pn mhl
meillo@114 2275 is to format messages or parts of them.
meillo@114 2276 The only place in mmh, where a pager is invoked is
meillo@114 2277 .Pn mhshow .
meillo@114 2278 .P
meillo@114 2279 .Pe mhshow-show-*
meillo@114 2280 profile entries can be used to display MIME parts in a specific way.
meillo@114 2281 For instance, PDF and Postscript files could be converted to plain text
meillo@114 2282 to display them in the terminal.
meillo@169 2283 In mmh, MIME parts will always be displayed serially.
meillo@114 2284 The request to display the MIME type `multipart/parallel' in parallel
meillo@114 2285 is ignored.
meillo@114 2286 It is simply treated as `multipart/mixed'.
meillo@114 2287 .Ci d0581ba306a7299113a346f9b4c46ce97bc4cef6
meillo@114 2288 This could already be requested with the, now removed,
meillo@114 2289 .Sw -serialonly
meillo@114 2290 switch of
meillo@114 2291 .Pn mhshow .
meillo@114 2292 As MIME parts are always processed exclusively , i.e. serially,
meillo@114 2293 the `%e' escape in
meillo@114 2294 .Pe mhshow-show-*
meillo@114 2295 profile entries became useless and was thus removed.
meillo@114 2296 .Ci a20d405db09b7ccca74d3e8c57550883da49e1ae
meillo@114 2297 .P
meillo@114 2298 In the intended setup, only text content would be displayed.
meillo@114 2299 Non-text content would be converted to text by appropriate
meillo@114 2300 .Pe mhshow-show-*
meillo@114 2301 profile entries before, if possible and wanted.
meillo@114 2302 All output would be displayed in a single pager session.
meillo@114 2303 Other kinds of attachments are ignored.
meillo@114 2304 With
meillo@114 2305 .Pe mhshow-show-*
meillo@114 2306 profile entries for them, they can be displayed serially along
meillo@114 2307 the message.
meillo@114 2308 For parallel display, the attachments need to be stored to disk first.
meillo@114 2309 .P
meillo@114 2310 To display text content in foreign charsets, they need to be converted
meillo@114 2311 to the native charset.
meillo@114 2312 Therefore,
meillo@114 2313 .Pe mhshow-charset-*
meillo@114 2314 profile entries used to be needed.
meillo@169 2315 In mmh, the conversion is performed automatically by piping the
meillo@169 2316 text through the
meillo@114 2317 .Pn iconv
meillo@114 2318 command, if necessary.
meillo@114 2319 .Ci 2433122c20baccb10b70b49c04c6b0497b5b3b60
meillo@114 2320 Custom
meillo@114 2321 .Pe mhshow-show-*
meillo@114 2322 rules for textual content might need a
meillo@114 2323 .Cl "iconv -f %c %f |
meillo@114 2324 prefix to have the text converted to the native charset.
meillo@114 2325 .P
meillo@121 2326 Although the conversion of foreign charsets to the native one
meillo@114 2327 has improved, it is not consistent enough.
meillo@114 2328 Further work needs to be done and
meillo@114 2329 the basic concepts in this field need to be re-thought.
meillo@114 2330 Though, the default setup of mmh displays message in foreign charsets
meillo@114 2331 correctly without the need to configure anything.
meillo@114 2332
meillo@114 2333
meillo@114 2334 .ig
meillo@114 2335
meillo@114 2336 .P
meillo@114 2337 mhshow/mhstore: Removed support for retrieving message/external-body parts.
meillo@114 2338 These tools won't download the contents automatically anymore. Instead,
meillo@114 2339 they print the information needed to get the contents. If someone should
meillo@114 2340 really receive one of those rare message/external-body messages, he can
meillo@114 2341 do the job manually. We save nearly a thousand lines of code. That's worth
meillo@114 2342 it!
meillo@114 2343 (The profile entry `nmh-access-ftp' and sbr/ruserpass.c for reading
meillo@114 2344 ~/.netrc are gone now.)
meillo@114 2345 .Ci 55e1d8c654ee0f7c45b9361ce34617983b454c32
meillo@114 2346
meillo@114 2347 ..
meillo@102 2348
meillo@58 2349
meillo@58 2350
meillo@166 2351 .H2 "Signing and Encrypting
meillo@22 2352 .P
meillo@166 2353 Nmh offers no direct support for digital signatures and message encryption.
meillo@157 2354 This functionality needed to be added through third-party software.
meillo@166 2355 In mmh, the functionality should be included because it
meillo@166 2356 is a part of modern email and likely wanted by users of mmh.
meillo@157 2357 A fresh mmh installation should support signing and encrypting
meillo@157 2358 out-of-the-box.
meillo@157 2359 Therefore, Neil Rickert's
meillo@157 2360 .Pn mhsign
meillo@157 2361 and
meillo@157 2362 .Pn mhpgp
meillo@157 2363 scripts
meillo@157 2364 .[
meillo@157 2365 neil rickert mhsign mhpgp
meillo@157 2366 .]
meillo@157 2367 were included into mmh.
meillo@157 2368 The scripts fit well into the mmh, because they are lightweight and
meillo@157 2369 of style similar to the existing tools.
meillo@157 2370 Additionally, no licensing difficulties appeared,
meillo@157 2371 as they are part of the public domain.
meillo@112 2372 .P
meillo@157 2373 The scripts were written for nmh, hence I needed to adjust them according
meillo@157 2374 to the differences of mmh.
meillo@157 2375 For instance, I removed the use of the backup prefix and dropped support
meillo@157 2376 for old PGP features.
meillo@157 2377 .P
meillo@157 2378 .Pn mhsign
meillo@157 2379 handles the signing and encrypting part.
meillo@157 2380 It comprises about 250 lines of shell code and interfaces between
meillo@157 2381 .Pn gnupg
meillo@157 2382 and
meillo@157 2383 the MH system.
meillo@157 2384 It was meant to be invoked at the WhatNow prompt, but in mmh,
meillo@157 2385 .Pn send
meillo@157 2386 does the job automatically.
meillo@157 2387 Special header fields were introduced to request the action.
meillo@157 2388 If a draft contains the
meillo@157 2389 .Hd Sign
meillo@157 2390 header field,
meillo@157 2391 .Pn send
meillo@157 2392 will sign it.
meillo@157 2393 The key to be used is either chosen automatically or specified by the
meillo@157 2394 .Pe Pgpkey
meillo@157 2395 profile entry.
meillo@157 2396 .Pn send
meillo@157 2397 always signes messages using the PGP/MIME standard, \" REF XXX
meillo@157 2398 but by manually invoking
meillo@157 2399 .Pn mhsign ,
meillo@157 2400 old-style non-MIME signatures can be created as well.
meillo@157 2401 To sign an outgoing message, the draft needs to contain a
meillo@157 2402 .Hd Enc
meillo@157 2403 header field.
meillo@157 2404 Public keys of all recipients are taken from the gnupg keyring or
meillo@157 2405 from an overrides files, called
meillo@157 2406 .Fn pgpkeys .
meillo@157 2407 Unless public keys are found for all recipients,
meillo@157 2408 .Pn send
meillo@157 2409 will refuse to encrypt and send it.
meillo@157 2410 Currently, messages with hidden (BCC) recipients can not be encrypted.
meillo@157 2411 This corner-case requires a more complex solution.
meillo@157 2412 Covering it is left to do.
meillo@157 2413 .P
meillo@157 2414 The integrated message signing and encrypting support is one of the
meillo@157 2415 most recent features in mmh.
meillo@157 2416 Feedback from users and the experience I will gather myself
meillo@157 2417 will direct the further development of the facility.
meillo@157 2418 It is worthwhile to consider adding
meillo@157 2419 .Sw -[no]sign
meillo@157 2420 and
meillo@157 2421 .Sw -[no]enc
meillo@157 2422 switches to
meillo@157 2423 .Pn send ,
meillo@157 2424 to override the corresponding header fields.
meillo@157 2425 The profile entry:
meillo@157 2426 .VS
meillo@157 2427 send: -sign
meillo@157 2428 VE
meillo@157 2429 .LP
meillo@157 2430 would then activate signing of all outgoing messages.
meillo@157 2431 With the present approach, the line
meillo@157 2432 .VS
meillo@157 2433 Send:
meillo@157 2434 VE
meillo@157 2435 .LP
meillo@157 2436 needs to be added to all message forms to achieve the same result.
meillo@157 2437 Yet, the integration of
meillo@157 2438 .Pn mhsign
meillo@157 2439 into mmh is too recent to have enough experience to decide this
meillo@157 2440 question now.
meillo@157 2441 .P
meillo@157 2442 .Pn mhpgp
meillo@157 2443 is the contrary part to
meillo@157 2444 .Pn mhsign .
meillo@157 2445 It verifies signatures and decrypts messages.
meillo@157 2446 .P
meillo@157 2447 FIXME: Add it to mmh first, then write about it here.
meillo@157 2448 .P
meillo@157 2449 The integration of
meillo@157 2450 .Pn mhpgp
meillo@157 2451 into
meillo@157 2452 .Pn show ,
meillo@157 2453 to automatically verify signatures and decrypt messages as needed,
meillo@157 2454 is a task left open.
meillo@157 2455 .Pn show 's
meillo@157 2456 current structure does not allow such an integration on basis of
meillo@157 2457 the existing code.
meillo@157 2458 Extensive programming work is required. ... FIXME
meillo@157 2459
meillo@157 2460
meillo@58 2461
meillo@58 2462
meillo@102 2463
meillo@133 2464 .H2 "Draft and Trash Folder
meillo@131 2465 .P
meillo@58 2466
meillo@131 2467 .U3 "Draft Folder
meillo@154 2468 .Id draft-folder
meillo@131 2469 .P
meillo@131 2470 In the beginning, MH had the concept of a draft message.
meillo@131 2471 This is the file
meillo@131 2472 .Fn draft
meillo@131 2473 in the MH directory, which is treated special.
meillo@131 2474 On composing a message, this draft file was used.
meillo@131 2475 When starting to compose another message before the former one was sent,
meillo@131 2476 the user had to decide among:
meillo@131 2477 .BU
meillo@168 2478 Using the old draft to finish and send it before starting with a new one.
meillo@131 2479 .BU
meillo@168 2480 Discarding the old draft and replacing it with a new one.
meillo@131 2481 .BU
meillo@168 2482 Preserving the old draft by refiling it to a folder.
meillo@131 2483 .P
meillo@168 2484 It was only possible to work in alternation on multiple drafts.
meillo@131 2485 Therefore, the current draft needed to be refiled to a folder and
meillo@168 2486 another one re-used for editing.
meillo@131 2487 Working on multiple drafts at the same time was impossible.
meillo@131 2488 The usual approach of switching to a different MH context did not
meillo@168 2489 help anything.
meillo@131 2490 .P
meillo@131 2491 The draft folder facility exists to
meillo@131 2492 allow true parallel editing of drafts, in a straight forward way.
meillo@131 2493 It was introduced by Marshall T. Rose, already in 1984.
meillo@131 2494 Similar to other new features, the draft folder was inactive by default.
meillo@131 2495 Even in nmh, the highly useful draft folder was not available
meillo@131 2496 out-of-the-box.
meillo@131 2497 At least, Richard Coleman added the man page
meillo@131 2498 .Mp mh-draft (5)
meillo@131 2499 to better document the feature.
meillo@131 2500 .P
meillo@131 2501 Not using the draft folder facility has the single advantage of having
meillo@131 2502 the draft file at a static location.
meillo@131 2503 This is simple in simple cases but the concept does not scale for more
meillo@131 2504 complex cases.
meillo@131 2505 The concept of the draft message is too limited for the problem.
meillo@131 2506 Therefore the draft folder was introduced.
meillo@131 2507 It is the more powerful and more natural concept.
meillo@131 2508 The draft folder is a folder like any other folder in MH.
meillo@131 2509 Its messages can be listed like any other messages.
meillo@131 2510 A draft message is no longer a special case.
meillo@131 2511 Tools do not need special switches to work on the draft message.
meillo@131 2512 Hence corner-cases were removed.
meillo@131 2513 .P
meillo@131 2514 The trivial part of the work was activating the draft folder with a
meillo@131 2515 default name.
meillo@131 2516 I chose the name
meillo@131 2517 .Fn +drafts
meillo@131 2518 for obvious reasons.
meillo@131 2519 In consequence, the command line switches
meillo@131 2520 .Sw -draftfolder
meillo@131 2521 and
meillo@131 2522 .Sw -draftmessage
meillo@131 2523 could be removed.
meillo@131 2524 More difficult but also more improving was updating the tools to the
meillo@131 2525 new concept.
meillo@131 2526 For nearly three decades, the tools needed to support two draft handling
meillo@131 2527 approaches.
meillo@131 2528 By fully switching to the draft folder, the tools could be simplified
meillo@131 2529 by dropping the awkward draft message handling code.
meillo@131 2530 .Sw -draft
meillo@131 2531 switches were removed because operating on a draft message is no longer
meillo@131 2532 special.
meillo@131 2533 It became indistinguishable to operating on any other message.
meillo@168 2534 .Ci 337338b404931f06f0db2119c9e145e8ca5a9860
meillo@168 2535 .P
meillo@168 2536 There is no more need to query the user for draft handling
meillo@168 2537 .Ci 2d48b455c303a807041c35e4248955f8bec59eeb .
meillo@131 2538 It is always possible to add another new draft.
meillo@131 2539 Refiling drafts is without difference to refiling other messages.
meillo@168 2540 All of these special cases are gone.
meillo@131 2541 Yet, one draft-related switch remained.
meillo@131 2542 .Pn comp
meillo@131 2543 still has
meillo@131 2544 .Sw -[no]use
meillo@131 2545 for switching between two modes:
meillo@131 2546 .BU
meillo@168 2547 .Sw -use
meillo@168 2548 to modify an existing draft.
meillo@131 2549 .BU
meillo@168 2550 .Sw -nouse
meillo@168 2551 to compose a new draft, possibly taking some existing message as template.
meillo@131 2552 .P
meillo@131 2553 In either case, the behavior of
meillo@131 2554 .Pn comp
meillo@131 2555 is deterministic.
meillo@131 2556 .P
meillo@131 2557 .Pn send
meillo@131 2558 now operates on the current message in the draft folder by default.
meillo@131 2559 As message and folder can both be overridden by specifying them on
meillo@131 2560 the command line, it is possible to send any message in the mail storage
meillo@131 2561 by simply specifying its number and folder.
meillo@131 2562 In contrast to the other tools,
meillo@131 2563 .Pn send
meillo@131 2564 takes the draft folder as its default folder.
meillo@131 2565 .P
meillo@131 2566 Dropping the draft message concept in favor for the draft folder concept,
meillo@131 2567 removed special cases with regular cases.
meillo@131 2568 This simplified the source code of the tools, as well as the concepts.
meillo@131 2569 In mmh, draft management does not break with the MH concepts
meillo@131 2570 but applies them.
meillo@133 2571 .Cl "scan +drafts" ,
meillo@133 2572 for instance, is a truly natural request.
meillo@169 2573 Most of the work was already performed by Rose in the eighties.
meillo@133 2574 The original improvement of mmh is dropping the old draft message approach
meillo@133 2575 and thus simplifying the tools, the documentation and the system as a whole.
meillo@131 2576 Although my part in the draft handling improvement was small,
meillo@133 2577 it was an important one.
meillo@131 2578
meillo@131 2579
meillo@131 2580 .U3 "Trash Folder
meillo@154 2581 .Id trash-folder
meillo@131 2582 .P
meillo@131 2583 Similar to the situation for drafts is the situation for removed messages.
meillo@131 2584 Historically, a message was ``deleted'' by prepending a specific
meillo@164 2585 \fIbackup prefix\fP, usually the comma character, to the file name.
meillo@164 2586 The specific file would then be ignored by MH because only files with
meillo@164 2587 names consisting of digits only are treated as messages.
meillo@131 2588 Although files remained in the file system,
meillo@168 2589 the messages were no longer visible in MH.
meillo@168 2590 To truly delete them, a maintenance job was needed.
meillo@168 2591 Usually a cron job was installed to delete them after a grace time.
meillo@131 2592 For instance:
meillo@131 2593 .VS
meillo@131 2594 find $HOME/Mail -type f -name ',*' -ctime +7 -delete
meillo@131 2595 VE
meillo@168 2596 In such a setup, the original message could be restored
meillo@131 2597 within the grace time interval by stripping the
meillo@164 2598 backup prefix from the file name.
meillo@168 2599 But the user could not rely on this statement.
meillo@168 2600 If the last message of a folder with six messages (\fL1-6\fP) was removed,
meillo@131 2601 message
meillo@131 2602 .Fn 6 ,
meillo@168 2603 became file
meillo@131 2604 .Fn ,6 .
meillo@168 2605 If then a new message entered the same folder, it would be named with
meillo@168 2606 the number one above the highest existing message number.
meillo@168 2607 In this case the message would be named
meillo@131 2608 .Fn 6
meillo@131 2609 then.
meillo@168 2610 If this new message would be removed as well,
meillo@168 2611 then the backup of the former message is overwritten.
meillo@168 2612 Hence, the ability to restore removed messages did not only depend on
meillo@131 2613 the ``sweeping cron job'' but also on the removing of further messages.
meillo@131 2614 It is undesirable to have such obscure and complex mechanisms.
meillo@168 2615 The user should be given a small set of clear assertions, such as
meillo@131 2616 ``Removed files are restorable within a seven-day grace time.''
meillo@131 2617 With the addition ``... unless a message with the same name in the
meillo@131 2618 same folder is removed before.'' the statement becomes complex.
meillo@131 2619 A user will hardly be able to keep track of any removal to know
meillo@131 2620 if the assertion still holds true for a specific file.
meillo@164 2621 In practice, the real mechanism is unclear to the user.
meillo@131 2622 The consequences of further removals are not obvious.
meillo@131 2623 .P
meillo@131 2624 Further more, the backup files are scattered within the whole mail storage.
meillo@131 2625 This complicates managing them.
meillo@164 2626 It is possible with the help of
meillo@131 2627 .Pn find ,
meillo@131 2628 but everything would be more convenient
meillo@131 2629 if the deleted messages would be collected in one place.
meillo@131 2630 .P
meillo@131 2631 The profile entry
meillo@131 2632 .Pe rmmproc
meillo@131 2633 (previously named
meillo@131 2634 .Pe Delete-Prog )
meillo@131 2635 was introduced very early to improve the situation.
meillo@164 2636 It could be set to any command, which would be executed to remove
meillo@131 2637 the specified messages.
meillo@168 2638 This would override the default action described above.
meillo@168 2639 Refiling the to-be-removed files to a trash folder is the usual example.
meillo@131 2640 Nmh's man page
meillo@131 2641 .Mp rmm (1)
meillo@131 2642 proposes to set the
meillo@131 2643 .Pe rmmproc
meillo@131 2644 to
meillo@131 2645 .Cl "refile +d
meillo@168 2646 to move messages to the trash folder,
meillo@131 2647 .Fn +d ,
meillo@131 2648 instead of renaming them with the backup prefix.
meillo@131 2649 The man page proposes additionally the expunge command
meillo@131 2650 .Cl "rm `mhpath +d all`
meillo@168 2651 to empty the trash folder.
meillo@131 2652 .P
meillo@131 2653 Removing messages in such a way has advantages.
meillo@131 2654 The mail storage is prevented from being cluttered with removed messages
meillo@131 2655 because they are all collected in one place.
meillo@131 2656 Existing and removed messages are thus separated more strictly.
meillo@131 2657 No backup files are silently overwritten.
meillo@164 2658 But most important is the ability to keep removed messages in the MH domain.
meillo@131 2659 Messages in the trash folder can be listed like those in any other folder.
meillo@131 2660 Deleted messages can be displayed like any other messages.
meillo@169 2661 .Pn refile
meillo@169 2662 can restore deleted messages.
meillo@131 2663 All operations on deleted files are still covered by the MH tools.
meillo@131 2664 The trash folder is just like any other folder in the mail storage.
meillo@131 2665 .P
meillo@131 2666 Similar to the draft folder case, I dropped the old backup prefix approach
meillo@131 2667 in favor for replacing it by the better suiting trash folder system.
meillo@131 2668 Hence,
meillo@131 2669 .Pn rmm
meillo@131 2670 calls
meillo@131 2671 .Pn refile
meillo@131 2672 to move the to-be-removed message to the trash folder,
meillo@131 2673 .Fn +trash
meillo@131 2674 by default.
meillo@164 2675 To sweep it clean, the user can use
meillo@131 2676 .Cl "rmm -unlink +trash a" ,
meillo@131 2677 where the
meillo@131 2678 .Sw -unlink
meillo@131 2679 switch causes the files to be unlinked.
meillo@168 2680 .Ci 8edc5aaf86f9f77124664f6801bc6c6cdf258173
meillo@168 2681 .Ci ca0b3e830b86700d9e5e31b1784de2bdcaf58fc5
meillo@131 2682 .P
meillo@168 2683 Dropping the legacy approach and converting to the new approach completely
meillo@131 2684 simplified the code base.
meillo@131 2685 The relationship between
meillo@131 2686 .Pn rmm
meillo@131 2687 and
meillo@131 2688 .Pn refile
meillo@131 2689 was inverted.
meillo@131 2690 In mmh,
meillo@131 2691 .Pn rmm
meillo@131 2692 invokes
meillo@131 2693 .Pn refile ,
meillo@131 2694 which used to be the other way round.
meillo@131 2695 Yet, the relationship is simpler now.
meillo@164 2696 Loops, like described in nmh's man page for
meillo@131 2697 .Mp refile (1),
meillo@164 2698 can no longer occur:
meillo@131 2699 .QS
meillo@131 2700 Since
meillo@131 2701 .Pn refile
meillo@131 2702 uses your
meillo@131 2703 .Pe rmmproc
meillo@131 2704 to delete the message, the
meillo@131 2705 .Pe rmmproc
meillo@131 2706 must NOT call
meillo@131 2707 .Pn refile
meillo@131 2708 without specifying
meillo@131 2709 .Sw -normmproc
meillo@131 2710 or you will create an infinite loop.
meillo@131 2711 .QE
meillo@131 2712 .LP
meillo@131 2713 .Pn rmm
meillo@131 2714 either unlinks a message with
meillo@131 2715 .Fu unlink()
meillo@131 2716 or invokes
meillo@131 2717 .Pn refile
meillo@131 2718 to move it to the trash folder.
meillo@131 2719 .Pn refile
meillo@131 2720 does not invoke any tools.
meillo@131 2721 .P
meillo@136 2722 By generalizing the message removal in the way that it became covered
meillo@136 2723 by the MH concepts made the whole system more powerful.
meillo@131 2724
meillo@131 2725
meillo@131 2726
meillo@131 2727
meillo@131 2728
meillo@133 2729 .H2 "Modern Defaults
meillo@133 2730 .P
meillo@133 2731 Nmh has a bunch of convenience-improving features inactive by default,
meillo@133 2732 although one can expect every new user wanting to have them active.
meillo@133 2733 The reason they are inactive by default is the wish to stay compatible
meillo@133 2734 with old versions.
meillo@136 2735 But what is the definition for old versions?
meillo@136 2736 Still, the highly useful draft folder facility has not been activated
meillo@136 2737 by default although it was introduced over twenty-five years ago.
meillo@133 2738 .[
meillo@133 2739 rose romine real work
meillo@133 2740 .]
meillo@136 2741 The community seems not to care.
meillo@136 2742 This is one of several examples that require new users to first build up
meillo@136 2743 a profile before they can access the modern features of nmh.
meillo@136 2744 Without an extensive profile, the setup is hardly usable
meillo@133 2745 for modern emailing.
meillo@133 2746 The point is not the customization of the setup,
meillo@136 2747 but the need to activate generally useful facilities.
meillo@133 2748 .P
meillo@133 2749 Yet, the real problem lies less in enabling the features, as this is
meillo@133 2750 straight forward as soon as one knows what he wants.
meillo@168 2751 The real problem is that new users need deep insight into the project
meillo@168 2752 to find out about inactive features nmh already provides.
meillo@133 2753 To give an example, I needed one year of using nmh
meillo@133 2754 before I became aware of the existence of the attachment system.
meillo@133 2755 One could argue that this fact disqualifies my reading of the
meillo@133 2756 documentation.
meillo@133 2757 If I would have installed nmh from source back then, I could agree.
meillo@133 2758 Yet, I had used a prepackaged version and had expected that it would
meillo@133 2759 just work.
meillo@133 2760 Nevertheless, I had been convinced by the concepts of MH already
meillo@133 2761 and I am a software developer,
meillo@133 2762 still I required a lot of time to discover the cool features.
meillo@133 2763 How can we expect users to be even more advanced than me,
meillo@133 2764 just to allow them use MH in a convenient and modern way?
meillo@133 2765 Unless they are strongly convinced of the concepts, they will fail.
meillo@133 2766 I have seen friends of me giving up disappointed
meillo@133 2767 before they truly used the system,
meillo@133 2768 although they had been motivated in the beginning.
meillo@133 2769 They suffer hard enough to get used to the toolchest approach,
meillo@133 2770 we should spare them further inconveniences.
meillo@133 2771 .P
meillo@136 2772 Maintaining compatibility for its own sake is bad,
meillo@136 2773 because the code base collects more and more compatibility code.
meillo@136 2774 Sticking to the compatiblity code means remaining limited;
meillo@168 2775 whereas adjusting to the changes renders the compatibility unnecessary.
meillo@168 2776 Keeping unused alternatives in the code is a bad choice as they likely
meillo@136 2777 gather bugs, by not being well tested.
meillo@136 2778 Also, the increased code size and the greater number of conditions
meillo@136 2779 increase the maintenance costs.
meillo@133 2780 If any MH implementation would be the back-end of widespread
meillo@133 2781 email clients with large user bases, compatibility would be more
meillo@133 2782 important.
meillo@133 2783 Yet, it appears as if this is not the case.
meillo@133 2784 Hence, compatibility is hardly important for technical reasons.
meillo@133 2785 Its importance originates rather from personal reasons.
meillo@133 2786 Nmh's user base is small and old.
meillo@133 2787 Changing the interfaces would cause inconvenience to long-term users of MH.
meillo@133 2788 It would force them to change their many years old MH configurations.
meillo@168 2789 I do understand this aspect, but by sticking to the old users,
meillo@168 2790 new users are kept away.
meillo@133 2791 Yet, the future lies in new users.
meillo@168 2792 In consequence, mmh invites new users by providing a convenient
meillo@168 2793 and modern setup, readily usable out-of-the-box.
meillo@133 2794 .P
meillo@136 2795 In mmh, all modern features are active by default and many previous
meillo@136 2796 approaches are removed or only accessible in manual ways.
meillo@136 2797 New default features include:
meillo@133 2798 .BU
meillo@133 2799 The attachment system (\c
meillo@133 2800 .Hd Attach ).
meillo@133 2801 .Ci 8ff284ff9167eff8f5349481529332d59ed913b1
meillo@133 2802 .BU
meillo@133 2803 The draft folder facility (\c
meillo@133 2804 .Fn +drafts ).
meillo@133 2805 .Ci 337338b404931f06f0db2119c9e145e8ca5a9860
meillo@133 2806 .BU
meillo@133 2807 The unseen sequence (`u')
meillo@133 2808 .Ci c2360569e1d8d3678e294eb7c1354cb8bf7501c1
meillo@133 2809 and the sequence negation prefix (`!').
meillo@133 2810 .Ci db74c2bd004b2dc9bf8086a6d8bf773ac051f3cc
meillo@133 2811 .BU
meillo@133 2812 Quoting the original message in the reply.
meillo@133 2813 .Ci 67411b1f95d6ec987b4c732459e1ba8a8ac192c6
meillo@133 2814 .BU
meillo@133 2815 Forwarding messages using MIME.
meillo@133 2816 .Ci 6e271608b7b9c23771523f88d23a4d3593010cf1
meillo@136 2817 .P
meillo@136 2818 In consequence, a setup with a profile that defines only the path to the
meillo@136 2819 mail storage, is already convenient to use.
meillo@168 2820 Again, Paul Vixie's ``edginess'' call supports the direction I took:
meillo@136 2821 ``the `main branch' should just be modern''.
meillo@136 2822 .[
meillo@136 2823 paul vixie edginess nmh-workers
meillo@136 2824 .]
meillo@131 2825
meillo@133 2826
meillo@133 2827
meillo@133 2828
meillo@133 2829
meillo@133 2830 .\" --------------------------------------------------------------
meillo@131 2831 .H1 "Styling
meillo@22 2832 .P
meillo@118 2833 Kernighan and Pike have emphasized the importance of style in the
meillo@118 2834 preface of their book:
meillo@118 2835 .[ [
meillo@118 2836 kernighan pike practice of programming
meillo@118 2837 .], p. x]
meillo@118 2838 .QS
meillo@118 2839 Chapter 1 discusses programming style.
meillo@118 2840 Good style is so important to good programming that we have chose
meillo@118 2841 to cover it first.
meillo@118 2842 .QE
meillo@168 2843 This section covers changes in mmh that were guided by the desire
meillo@118 2844 to improve on style.
meillo@118 2845 Many of them follow the rules given in the quoted book.
meillo@118 2846 .[
meillo@118 2847 kernighan pike practice of programming
meillo@118 2848 .]
meillo@118 2849
meillo@118 2850
meillo@127 2851
meillo@127 2852
meillo@127 2853 .H2 "Code Style
meillo@154 2854 .Id code-style
meillo@118 2855 .P
meillo@118 2856 .U3 "Indentation Style
meillo@118 2857 .P
meillo@118 2858 Indentation styles are the holy cow of programmers.
meillo@168 2859 Kernighan and Pike
meillo@118 2860 .[ [
meillo@118 2861 kernighan pike practice of programming
meillo@118 2862 .], p. 10]
meillo@168 2863 wrote:
meillo@118 2864 .QS
meillo@118 2865 Programmers have always argued about the layout of programs,
meillo@118 2866 but the specific style is much less important than its consistent
meillo@118 2867 application.
meillo@121 2868 Pick one style, preferably ours, use it consistently, and don't waste
meillo@118 2869 time arguing.
meillo@118 2870 .QE
meillo@118 2871 .P
meillo@118 2872 I agree that the constant application is most important,
meillo@118 2873 but I believe that some styles have advantages over others.
meillo@118 2874 For instance the indentation with tab characters only.
meillo@118 2875 Tab characters directly map to the nesting level \(en
meillo@118 2876 one tab, one level.
meillo@118 2877 Tab characters are flexible because developers can adjust them to
meillo@118 2878 whatever width they like to have.
meillo@118 2879 There is no more need to run
meillo@118 2880 .Pn unexpand
meillo@118 2881 or
meillo@118 2882 .Pn entab
meillo@118 2883 programs to ensure the correct mixture of leading tabs and spaces.
meillo@118 2884 The simple rules are: (1) Leading whitespace must consist of tabs only.
meillo@118 2885 (2) Any other whitespace should consist of spaces.
meillo@121 2886 These two rules ensure the integrity of the visual appearance.
meillo@121 2887 Although reformatting existing code should be avoided, I did it.
meillo@136 2888 I did not waste time arguing; I just reformated the code.
meillo@118 2889 .Ci a485ed478abbd599d8c9aab48934e7a26733ecb1
meillo@118 2890
meillo@118 2891 .U3 "Comments
meillo@118 2892 .P
meillo@118 2893 Section 1.6 of
meillo@118 2894 .[ [
meillo@118 2895 kernighan pike practice of programming
meillo@118 2896 .], p. 23]
meillo@118 2897 demands: ``Don't belabor the obvious.''
meillo@122 2898 Hence, I simply removed all the comments in the following code excerpt:
meillo@118 2899 .VS
meillo@120 2900 context_replace(curfolder, folder); /* update current folder */
meillo@120 2901 seq_setcur(mp, mp->lowsel); /* update current message */
meillo@120 2902 seq_save(mp); /* synchronize message sequences */
meillo@120 2903 folder_free(mp); /* free folder/message structure */
meillo@120 2904 context_save(); /* save the context file */
meillo@120 2905
meillo@120 2906 [...]
meillo@120 2907
meillo@120 2908 int c; /* current character */
meillo@120 2909 char *cp; /* miscellaneous character pointer */
meillo@120 2910
meillo@120 2911 [...]
meillo@120 2912
meillo@120 2913 /* NUL-terminate the field */
meillo@120 2914 *cp = '\0';
meillo@118 2915 VE
meillo@120 2916 .Ci 426543622b377fc5d091455cba685e114b6df674
meillo@118 2917 .P
meillo@136 2918 The program code explains enough itself, already.
meillo@136 2919
meillo@118 2920
meillo@118 2921 .U3 "Names
meillo@118 2922 .P
meillo@118 2923 Kernighan and Pike suggest:
meillo@118 2924 ``Use active names for functions''.
meillo@118 2925 .[ [
meillo@118 2926 kernighan pike practice of programming
meillo@118 2927 .], p. 4]
meillo@118 2928 One application of this rule was the rename of
meillo@118 2929 .Fu check_charset()
meillo@118 2930 to
meillo@118 2931 .Fu is_native_charset() .
meillo@118 2932 .Ci 8d77b48284c58c135a6b2787e721597346ab056d
meillo@118 2933 The same change fixed a violation of ``Be accurate'' as well.
meillo@118 2934 The code did not match the expectation the function suggested,
meillo@118 2935 as it, for whatever reason, only compared the first ten characters
meillo@118 2936 of the charset name.
meillo@118 2937 .P
meillo@118 2938 More important than using active names is using descriptive names.
meillo@145 2939 .VS
meillo@145 2940 m_unknown(in); /* the MAGIC invocation... */
meillo@145 2941 VE
meillo@145 2942 Renaming the obscure
meillo@118 2943 .Fu m_unknown()
meillo@145 2944 function was a delightful event, although it made the code less funny.
meillo@118 2945 .Ci 611d68d19204d7cbf5bd585391249cb5bafca846
meillo@118 2946 .P
meillo@118 2947 Magic numbers are generally considered bad style.
meillo@118 2948 Obviously, Kernighan and Pike agree:
meillo@118 2949 ``Give names to magic numbers''.
meillo@118 2950 .[ [
meillo@118 2951 kernighan pike practice of programming
meillo@118 2952 .], p. 19]
meillo@118 2953 One such change was naming the type of input \(en mbox or mail folder \(en
meillo@118 2954 to be scanned:
meillo@118 2955 .VS
meillo@118 2956 #define SCN_MBOX (-1)
meillo@118 2957 #define SCN_FOLD 0
meillo@118 2958 VE
meillo@118 2959 .Ci 7ffb36d28e517a6f3a10272056fc127592ab1c19
meillo@118 2960 .P
meillo@118 2961 The argument
meillo@118 2962 .Ar outnum
meillo@118 2963 of the function
meillo@118 2964 .Fu scan()
meillo@118 2965 in
meillo@118 2966 .Fn uip/scansbr.c
meillo@118 2967 defines the number of the message to be created.
meillo@118 2968 If no message is to be created, the argument is misused to transport
meillo@118 2969 program logic.
meillo@118 2970 This lead to obscure code.
meillo@118 2971 I improved the clarity of the code by introducing two variables:
meillo@118 2972 .VS
meillo@118 2973 int incing = (outnum > 0);
meillo@118 2974 int ismbox = (outnum != 0);
meillo@118 2975 VE
meillo@118 2976 They cover the magic values and are used for conditions.
meillo@118 2977 The variable
meillo@118 2978 .Ar outnum
meillo@118 2979 is only used when it holds an ordinary message number.
meillo@118 2980 .Ci b8b075c77be7794f3ae9ff0e8cedb12b48fd139f
meillo@118 2981 The clarity improvement of the change showed detours in the program logic
meillo@118 2982 of related code parts.
meillo@118 2983 Having the new variables with descriptive names, a more
meillo@121 2984 straight forward implementation became apparent.
meillo@169 2985 Before the code was clarified, the possibility to improve had not be seen.
meillo@118 2986 .Ci aa60b0ab5e804f8befa890c0a6df0e3143ce0723
meillo@118 2987
meillo@133 2988
meillo@133 2989
meillo@133 2990 .H2 "Structural Rework
meillo@133 2991 .P
meillo@136 2992 Although the stylistic changes described up to here improve the
meillo@136 2993 readability of the source code, all of them are changes ``in the small''.
meillo@136 2994 Structural changes affect a much larger area.
meillo@136 2995 They are more difficult to do but lead to larger improvements,
meillo@136 2996 especially as they influence the outer shape of the tools as well.
meillo@118 2997 .P
meillo@118 2998 At the end of their chapter on style,
meillo@118 2999 Kernighan and Pike ask: ``But why worry about style?''
meillo@136 3000 Following are two examples of structural rework that show
meillo@136 3001 why style is important in the first place.
meillo@136 3002
meillo@136 3003
meillo@136 3004 .U3 "Rework of \f(CWanno\fP
meillo@118 3005 .P
meillo@120 3006 Until 2002,
meillo@120 3007 .Pn anno
meillo@120 3008 had six functional command line switches,
meillo@120 3009 .Sw -component
meillo@120 3010 and
meillo@120 3011 .Sw -text ,
meillo@168 3012 which have an argument each,
meillo@120 3013 and the two pairs of flags,
meillo@120 3014 .Sw -[no]date
meillo@120 3015 and
meillo@120 3016 .Sw -[no]inplace .
meillo@120 3017 Then Jon Steinhart introduced his attachment system.
meillo@120 3018 In need for more advanced annotation handling, he extended
meillo@120 3019 .Pn anno .
meillo@120 3020 He added five more switches:
meillo@120 3021 .Sw -draft ,
meillo@120 3022 .Sw -list ,
meillo@120 3023 .Sw -delete ,
meillo@120 3024 .Sw -append ,
meillo@120 3025 and
meillo@120 3026 .Sw -number ,
meillo@120 3027 the last one taking an argument.
meillo@121 3028 .Ci 7480dbc14bc90f2d872d434205c0784704213252
meillo@120 3029 Later,
meillo@120 3030 .Sw -[no]preserve
meillo@120 3031 was added.
meillo@121 3032 .Ci d9b1d57351d104d7ec1a5621f090657dcce8cb7f
meillo@120 3033 Then, the Synopsis section of the man page
meillo@120 3034 .Mp anno (1)
meillo@120 3035 read:
meillo@120 3036 .VS
meillo@120 3037 anno [+folder] [msgs] [-component field] [-inplace | -noinplace]
meillo@120 3038 [-date | -nodate] [-draft] [-append] [-list] [-delete]
meillo@120 3039 [-number [num|all]] [-preserve | -nopreserve] [-version]
meillo@120 3040 [-help] [-text body]
meillo@120 3041 VE
meillo@120 3042 .LP
meillo@120 3043 The implementation followed the same structure.
meillo@120 3044 Problems became visible when
meillo@120 3045 .Cl "anno -list -number 42
meillo@120 3046 worked on the current message instead on message number 42,
meillo@120 3047 and
meillo@120 3048 .Cl "anno -list -number l:5
meillo@124 3049 did not work on the last five messages but failed with the mysterious
meillo@120 3050 error message: ``anno: missing argument to -list''.
meillo@121 3051 Yet, the invocation matched the specification in the man page.
meillo@120 3052 There, the correct use of
meillo@120 3053 .Sw -number
meillo@120 3054 was defined as being
meillo@120 3055 .Cl "[-number [num|all]]
meillo@120 3056 and the textual description for the combination with
meillo@120 3057 .Sw -list
meillo@120 3058 read:
meillo@120 3059 .QS
meillo@164 3060 The
meillo@164 3061 .Sw -list
meillo@164 3062 option produces a listing of the field bodies for
meillo@120 3063 header fields with names matching the specified component,
meillo@164 3064 one per line. The listing is numbered, starting at 1, if the
meillo@164 3065 .Sw -number
meillo@164 3066 option is also used.
meillo@120 3067 .QE
meillo@120 3068 .LP
meillo@120 3069 The problem was manifold.
meillo@120 3070 The code required a numeric argument to the
meillo@120 3071 .Sw -number
meillo@120 3072 switch.
meillo@120 3073 If it was missing or non-numeric,
meillo@120 3074 .Pn anno
meillo@120 3075 aborted with an error message that had an off-by-one error,
meillo@120 3076 printing the switch one before the failing one.
meillo@120 3077 Semantically, the argument to the
meillo@120 3078 .Sw -number
meillo@120 3079 switch is only necessary in combination with
meillo@120 3080 .Sw -delete ,
meillo@120 3081 but not with
meillo@120 3082 .Sw -list .
meillo@120 3083 In the former case it is even necessary.
meillo@120 3084 .P
meillo@120 3085 Trying to fix these problems on the surface would not have solved it truly.
meillo@120 3086 The problems discovered originate from a discrepance between the semantic
meillo@120 3087 structure of the problem and the structure implemented in the program.
meillo@120 3088 Such structural differences can not be cured on the surface.
meillo@120 3089 They need to be solved by adjusting the structure of the implementation
meillo@120 3090 to the structure of the problem.
meillo@120 3091 .P
meillo@120 3092 In 2002, the new switches
meillo@120 3093 .Sw -list
meillo@120 3094 and
meillo@120 3095 .Sw -delete
meillo@120 3096 were added in the same way, the
meillo@120 3097 .Sw -number
meillo@120 3098 switch for instance had been added.
meillo@120 3099 Yet, they are of structural different type.
meillo@120 3100 Semantically,
meillo@120 3101 .Sw -list
meillo@120 3102 and
meillo@120 3103 .Sw -delete
meillo@120 3104 introduce modes of operation.
meillo@120 3105 Historically,
meillo@120 3106 .Pn anno
meillo@120 3107 had only one operation mode: adding header fields.
meillo@120 3108 With the extension, it got two moder modes:
meillo@120 3109 listing and deleting header fields.
meillo@120 3110 The structure of the code changes did not pay respect to this
meillo@120 3111 fundamental change to
meillo@120 3112 .Pn anno 's
meillo@120 3113 behavior.
meillo@120 3114 Neither the implementation nor the documentation did clearly
meillo@120 3115 define them as being exclusive modes of operation.
meillo@120 3116 Having identified the problem, I solved it by putting structure into
meillo@120 3117 .Pn anno
meillo@120 3118 and its documentation.
meillo@120 3119 .Ci d54c8db8bdf01e8381890f7729bc0ef4a055ea11
meillo@120 3120 .P
meillo@120 3121 The difference is visible in both, the code and the documentation.
meillo@121 3122 The following code excerpt:
meillo@120 3123 .VS
meillo@120 3124 int delete = -2; /* delete header element if set */
meillo@120 3125 int list = 0; /* list header elements if set */
meillo@120 3126 [...]
meillo@121 3127 case DELETESW: /* delete annotations */
meillo@121 3128 delete = 0;
meillo@121 3129 continue;
meillo@121 3130 case LISTSW: /* produce a listing */
meillo@121 3131 list = 1;
meillo@121 3132 continue;
meillo@120 3133 VE
meillo@121 3134 .LP
meillo@121 3135 was replaced by:
meillo@120 3136 .VS
meillo@120 3137 static enum { MODE_ADD, MODE_DEL, MODE_LIST } mode = MODE_ADD;
meillo@120 3138 [...]
meillo@121 3139 case DELETESW: /* delete annotations */
meillo@121 3140 mode = MODE_DEL;
meillo@121 3141 continue;
meillo@121 3142 case LISTSW: /* produce a listing */
meillo@121 3143 mode = MODE_LIST;
meillo@121 3144 continue;
meillo@120 3145 VE
meillo@120 3146 .LP
meillo@121 3147 The replacement code does not only reflect the problem's structure better,
meillo@121 3148 it is easier to understand as well.
meillo@121 3149 The same applies to the documentation.
meillo@120 3150 The man page was completely reorganized to propagate the same structure.
meillo@121 3151 This is visible in the Synopsis section:
meillo@120 3152 .VS
meillo@120 3153 anno [+folder] [msgs] [-component field] [-text body]
meillo@120 3154 [-append] [-date | -nodate] [-preserve | -nopreserve]
meillo@120 3155 [-Version] [-help]
meillo@120 3156
meillo@120 3157 anno -delete [+folder] [msgs] [-component field] [-text
meillo@120 3158 body] [-number num | all ] [-preserve | -nopreserve]
meillo@120 3159 [-Version] [-help]
meillo@120 3160
meillo@120 3161 anno -list [+folder] [msgs] [-component field] [-number]
meillo@120 3162 [-Version] [-help]
meillo@120 3163 VE
meillo@121 3164 .\" XXX think about explaining the -preserve rework?
meillo@118 3165
meillo@58 3166
meillo@58 3167
meillo@133 3168 .U3 "Path Conversion
meillo@133 3169 .P
meillo@134 3170 Four kinds of path names can appear in MH:
meillo@134 3171 .IP (1)
meillo@134 3172 Absolute Unix directory paths, like
meillo@134 3173 .Fn /etc/passwd .
meillo@134 3174 .IP (2)
meillo@134 3175 Relative Unix directory paths, like
meillo@134 3176 .Fn ./foo/bar .
meillo@134 3177 .IP (3)
meillo@134 3178 Absolute MH folder paths, like
meillo@134 3179 .Fn +friends/phil .
meillo@134 3180 .IP (4)
meillo@134 3181 Relative MH folder paths, like
meillo@134 3182 .Fn @subfolder .
meillo@134 3183 .P
meillo@134 3184 The last type, relative MH folder paths, are hardly documented.
meillo@134 3185 Nonetheless, they are useful for large mail storages.
meillo@134 3186 The current mail folder is specified as `\c
meillo@134 3187 .Fn @ ',
meillo@134 3188 just like the current directory is specified as `\c
meillo@134 3189 .Fn . '.
meillo@134 3190 .P
meillo@134 3191 To allow MH tools to understand all four notations,
meillo@134 3192 they need to convert between them.
meillo@134 3193 In nmh, these path name conversion functions were located in the files
meillo@134 3194 .Fn sbr/path.c
meillo@134 3195 (``return a pathname'') and
meillo@134 3196 .Fn sbr/m_maildir.c
meillo@134 3197 (``get the path for the mail directory'').
meillo@134 3198 The seven functions in the two files were documented with no more
meillo@134 3199 than two comments, which described obvious information.
meillo@134 3200 The function signatures were neither explaining:
meillo@134 3201 .VS
meillo@134 3202 char *path(char *, int);
meillo@134 3203 char *pluspath(char *);
meillo@134 3204 char *m_mailpath(char *);
meillo@134 3205 char *m_maildir(char *);
meillo@134 3206 VE
meillo@134 3207 .P
meillo@134 3208 My investigation provides the following description:
meillo@134 3209 .BU
meillo@134 3210 The second parameter of
meillo@134 3211 .Fu path()
meillo@134 3212 defines the type of path given as first parameter.
meillo@134 3213 Directory paths are converted to absolute directory paths.
meillo@134 3214 Folder paths are converted to absolute folder paths.
meillo@134 3215 Folder paths must not include a leading `@' character.
meillo@134 3216 Leading plus characters are preserved.
meillo@134 3217 The result is a pointer to newly allocated memory.
meillo@134 3218 .BU
meillo@134 3219 .Fu pluspath()
meillo@134 3220 is a convenience-wrapper to
meillo@134 3221 .Fu path() ,
meillo@134 3222 to convert folder paths only.
meillo@134 3223 This function can not be used for directory paths.
meillo@134 3224 An empty string parameter causes a buffer overflow.
meillo@134 3225 .BU
meillo@134 3226 .Fu m_mailpath()
meillo@134 3227 converts directory paths to absolute directory paths.
meillo@134 3228 The characters `+' or `@' at the beginning of the path name are
meillo@134 3229 treated literal, i.e. as the first character of a relative directory path.
meillo@134 3230 Hence, this function can not be used for folder paths.
meillo@134 3231 In any case, the result is an absolute directory path.
meillo@134 3232 The result is a pointer to newly allocated memory.
meillo@134 3233 .BU
meillo@134 3234 .Fu m_maildir()
meillo@134 3235 returns the parameter unchanged if it is an absolute directory path
meillo@134 3236 or begins with the entry `.' or `..'.
meillo@134 3237 All other strings are prepended with the current working directory.
meillo@134 3238 Hence, this functions can not be used for folder paths.
meillo@134 3239 The result is either an absolute directory path or a relative
meillo@134 3240 directory path, starting with a dot.
meillo@134 3241 In contrast to the other functions, the result is a pointer to
meillo@134 3242 static memory.
meillo@134 3243 .P
meillo@134 3244 The situation was obscure, irritating, error-prone, and non-orthogonal.
meillo@134 3245 No clear terminology was used to name the different kinds of path names.
meillo@134 3246 The first argument of
meillo@134 3247 .Fu m_mailpath() ,
meillo@134 3248 for instance, was named
meillo@134 3249 .Ar folder ,
meillo@134 3250 though
meillo@134 3251 .Fu m_mailpath()
meillo@134 3252 can not be used for MH folders.
meillo@134 3253 .P
meillo@134 3254 I reworked the path name conversion completely, introducing clarity.
meillo@134 3255 First of all, the terminology needed to be defined.
meillo@134 3256 A path name is either in the Unix domain, then it is called
meillo@134 3257 \fIdirectory path\fP, `dirpath' for short, or it is in the MH domain,
meillo@134 3258 then it is called \fIfolder path\fP, `folpath' for short.
meillo@134 3259 The two terms need to be used with strict distinction.
meillo@134 3260 Having a clear terminology is often an indicator of having understood
meillo@134 3261 the problem itself.
meillo@134 3262 Second, I exploited the concept of path type indicators.
meillo@134 3263 By requesting every path name to start with a clear type identifier,
meillo@134 3264 conversion between the types can be fully automated.
meillo@134 3265 Thus the tools can accept paths of any type from the user.
meillo@134 3266 Therefore, it was necessary to require relative directory paths to be
meillo@134 3267 prefixed with a dot character.
meillo@134 3268 In consequence, the dot character could no longer be an alias for the
meillo@134 3269 current message.
meillo@134 3270 .Ci cff0e16925e7edbd25b8b9d6d4fbdf03e0e60c01
meillo@134 3271 Third, I created three new functions to replace the previous mess:
meillo@134 3272 .BU
meillo@134 3273 .Fu expandfol()
meillo@134 3274 converts folder paths to absolute folder paths,
meillo@134 3275 without the leading plus character.
meillo@134 3276 Directory paths are simply passed through.
meillo@134 3277 This function is to be used for folder paths only, thus the name.
meillo@134 3278 The result is a pointer to static memory.
meillo@134 3279 .BU
meillo@134 3280 .Fu expanddir()
meillo@134 3281 converts directory paths to absolute directory paths.
meillo@134 3282 Folder paths are treated as relative directory paths.
meillo@134 3283 This function is to be used for directory paths only, thus the name.
meillo@134 3284 The result is a pointer to static memory.
meillo@134 3285 .BU
meillo@134 3286 .Fu toabsdir()
meillo@134 3287 converts any type of path to an absolute directory path.
meillo@134 3288 This is the function of choice for path conversion.
meillo@134 3289 Absolute directory paths are the most general representation of a
meillo@134 3290 path name.
meillo@134 3291 The result is a pointer to static memory.
meillo@134 3292 .P
meillo@134 3293 The new functions have names that indicate their use.
meillo@134 3294 Two of the functions convert relative to absolute path names of the
meillo@134 3295 same type.
meillo@134 3296 The third function converts any path name type to the most general one,
meillo@134 3297 the absolute directory path.
meillo@134 3298 All of the functions return pointers to static memory.
meillo@134 3299 All three functions are implemented in
meillo@134 3300 .Fn sbr/path.c .
meillo@134 3301 .Fn sbr/m_maildir.c
meillo@134 3302 is removed.
meillo@168 3303 .Ci d39e2c447b0d163a5a63f480b23d06edb7a73aa0
meillo@134 3304 .P
meillo@134 3305 Along with the path conversion rework, I also replaced
meillo@134 3306 .Fu getfolder(FDEF)
meillo@134 3307 with
meillo@134 3308 .Fu getdeffol()
meillo@134 3309 and
meillo@134 3310 .Fu getfolder(FCUR)
meillo@134 3311 with
meillo@134 3312 .Fu getcurfol() ,
meillo@134 3313 which is only a convenience wrapper for
meillo@134 3314 .Fu expandfol("@") .
meillo@134 3315 This code was moved from
meillo@134 3316 .Fn sbr/getfolder.c
meillo@134 3317 to
meillo@134 3318 .Fn sbr/path.c .
meillo@168 3319 .Ci d39e2c447b0d163a5a63f480b23d06edb7a73aa0
meillo@134 3320 .P
meillo@134 3321 The related function
meillo@134 3322 .Fu etcpath()
meillo@134 3323 was moved to
meillo@134 3324 .Fn sbr/path.c ,
meillo@168 3325 too
meillo@168 3326 .Ci b4c29794c12099556151d93a860ee51badae2e35 .
meillo@134 3327 Previously, it had been located in
meillo@134 3328 .Fn config/config.c ,
meillo@134 3329 for whatever reasons.
meillo@134 3330 .P
meillo@134 3331 .Fn sbr/path.c
meillo@134 3332 now contains all path handling code.
meillo@134 3333 Only 173 lines of code were needed to replace the previous 252 lines.
meillo@134 3334 The readability of the code is highly improved.
meillo@134 3335 Additionally, each of the six exported and one static functions
meillo@134 3336 is introduced by an explaining comment.
meillo@133 3337
meillo@133 3338
meillo@133 3339
meillo@133 3340
meillo@133 3341 .H2 "Profile Reading
meillo@133 3342 .P
meillo@138 3343 The MH profile contains the configuration for the user-specific MH setup.
meillo@138 3344 MH tools read the profile right after starting up,
meillo@138 3345 as it contains the location of the user's mail storage
meillo@138 3346 and similar settings that influence the whole setup.
meillo@138 3347 Further more, the profile contains the default switches for the tools,
meillo@138 3348 hence, it must be read before the command line switches are processed.
meillo@138 3349 .P
meillo@138 3350 For historic reasons, some MH tools did not read the profile and context.
meillo@138 3351 Among them were
meillo@138 3352 .Pn post /\c
meillo@138 3353 .Pn spost ,
meillo@138 3354 .Pn mhmail ,
meillo@138 3355 and
meillo@138 3356 .Pn slocal .
meillo@138 3357 The reason why these tools ignored the profile were not clearly stated.
meillo@138 3358 During the discussion on the nmh-workers mailing list,
meillo@138 3359 .[
meillo@138 3360 nmh-workers levine post profile
meillo@138 3361 .]
meillo@138 3362 David Levine posted an explanation, quoting John Romine:
meillo@138 3363 .QS
meillo@138 3364 I asked John Romine and here's what he had to say, which
meillo@138 3365 agrees and provides an example that convinces me:
meillo@138 3366 .QS
meillo@164 3367 My take on this is that
meillo@164 3368 .Pn post
meillo@164 3369 should not be called by users directly, and it doesn't read the
meillo@164 3370 .Fn .mh_profile
meillo@138 3371 (only front-end UI programs read the profile).
meillo@138 3372 .QP
meillo@164 3373 For example, there can be contexts where
meillo@164 3374 .Pn post
meillo@164 3375 is called by a helper program (like `\c
meillo@164 3376 .Pn mhmail ')
meillo@164 3377 which may be run by a non-MH user.
meillo@164 3378 We don't want this to prompt the user to create an MH profile, etc.
meillo@138 3379 .QP
meillo@164 3380 My suggestion would be to have
meillo@164 3381 .Pn send
meillo@164 3382 pass a (hidden) `\c
meillo@164 3383 .Sw -fileproc
meillo@164 3384 .Ar proc '
meillo@164 3385 option to
meillo@164 3386 .Pn post
meillo@164 3387 if needed.
meillo@164 3388 You could also
meillo@164 3389 use an environment variable (I think
meillo@164 3390 .Pn send /\c
meillo@164 3391 .Pn whatnow
meillo@164 3392 do this).
meillo@138 3393 .QE
meillo@164 3394 I think that's the way to go.
meillo@164 3395 My personal preference is to use a command line option,
meillo@164 3396 not an environment variable.
meillo@138 3397 .QE
meillo@138 3398 .P
meillo@138 3399 To solve the problem of
meillo@138 3400 .Pn post
meillo@138 3401 not honoring the
meillo@138 3402 .Pe fileproc
meillo@138 3403 profile entry,
meillo@138 3404 the community roughly agreed that a switch
meillo@138 3405 .Sw -fileproc
meillo@138 3406 should be added to
meillo@138 3407 .Pn post
meillo@138 3408 to be able to pass a different fileproc.
meillo@138 3409 I strongly disagree with this approach because it does not solve
meillo@138 3410 the problem; it only removes a single symptom.
meillo@138 3411 The problem is that
meillo@138 3412 .Pn post
meillo@138 3413 does not behave as expected.
meillo@138 3414 But all programs should behave as expected.
meillo@138 3415 Clear and simple concepts are a precondition for this.
meillo@138 3416 Hence, the real solution is having all MH tools read the profile.
meillo@138 3417 .P
meillo@138 3418 Yet, the problem has a further aspect.
meillo@138 3419 It mainly originates in
meillo@138 3420 .Pn mhmail .
meillo@138 3421 .Pn mhmail
meillo@138 3422 was intended to be a replacement for
meillo@138 3423 .Pn mailx
meillo@138 3424 on systems with MH installations.
meillo@138 3425 .Pn mhmail
meillo@138 3426 should have been able to use just like
meillo@138 3427 .Pn mailx ,
meillo@138 3428 but sending the message via MH's
meillo@138 3429 .Pn post
meillo@138 3430 instead of
meillo@138 3431 .Pn sendmail .
meillo@138 3432 Using
meillo@138 3433 .Pn mhmail
meillo@138 3434 should not be influenced by the question whether the user had
meillo@138 3435 MH set up for himself or not.
meillo@138 3436 .Pn mhmail
meillo@138 3437 did not read the profile as this requests the user to set up MH
meillo@138 3438 if not done yet.
meillo@138 3439 As
meillo@138 3440 .Pn mhmail
meillo@138 3441 used
meillo@138 3442 .Pn post ,
meillo@138 3443 .Pn post
meillo@138 3444 could not read the profile neither.
meillo@138 3445 This is the reason why
meillo@138 3446 .Pn post
meillo@138 3447 does not read the profile.
meillo@138 3448 This is the reason for the actual problem.
meillo@138 3449 It was not much of a problem because
meillo@138 3450 .Pn post
meillo@138 3451 was not intended to be used by users directly.
meillo@138 3452 .Pn send
meillo@138 3453 is the interactive front-end to
meillo@138 3454 .Pn post .
meillo@138 3455 .Pn send
meillo@138 3456 read the profile and passed all relevant values on the command line to
meillo@138 3457 .Pn post
meillo@138 3458 \(en an awkward solution.
meillo@138 3459 .P
meillo@138 3460 The important insight is that
meillo@138 3461 .Pn mhmail
meillo@138 3462 is no true MH tool.
meillo@138 3463 The concepts broke because this outlandish tool was treated as any other
meillo@138 3464 MH tool.
meillo@138 3465 Instead it should have been treated accordingly to its foreign style.
meillo@138 3466 The solution is not to prevent the tools reading the profile but
meillo@138 3467 to instruct them reading a different profile.
meillo@138 3468 .Pn mhmail
meillo@138 3469 could have set up a well-defined profile and caused all MH tools
meillo@138 3470 in the session use it by exporting an environment variable.
meillo@138 3471 With this approach, no special cases would have been introduced,
meillo@138 3472 no surprises would have been caused.
meillo@138 3473 By writing a clean-profile-wrapper, the concept could have been
meillo@138 3474 generalized orthogonally to the whole MH toolchest.
meillo@138 3475 Then Rose's motivation behind the decision that
meillo@138 3476 .Pn post
meillo@138 3477 ignores the profile, as quoted by Jeffrey Honig,
meillo@138 3478 .[
meillo@138 3479 nmh-workers post profile
meillo@138 3480 .]
meillo@138 3481 would have become possible:
meillo@138 3482 .QS
meillo@138 3483 when you run mh commands in a script, you want all the defaults to be
meillo@138 3484 what the man page says.
meillo@138 3485 when you run a command by hand, then you want your own defaults...
meillo@138 3486 .QE
meillo@138 3487 .LP
meillo@138 3488 Yet, I consider this explanation short-sighted.
meillo@138 3489 We should rather regard theses two cases as just two different MH setups,
meillo@138 3490 based on two different profiles.
meillo@138 3491 Mapping such problems on the concepts of switching between different
meillo@138 3492 profiles, solves them once for all.
meillo@138 3493 .P
meillo@138 3494 In mmh, the wish to have
meillo@138 3495 .Pn mhmail
meillo@138 3496 as as replacement for
meillo@138 3497 .Pn mailx
meillo@138 3498 is considered obsolete.
meillo@138 3499 Mmh's
meillo@138 3500 .Pn mhmail
meillo@138 3501 does no longer cover this use-case.
meillo@138 3502 Currently,
meillo@138 3503 .Pn mhmail
meillo@138 3504 is in a transition state.
meillo@138 3505 .Ci 32d4f9daaa70519be3072479232ff7be0500d009
meillo@138 3506 It may become a front-end to
meillo@138 3507 .Pn comp ,
meillo@138 3508 which provides an interface more convenient in some cases.
meillo@138 3509 In this case,
meillo@138 3510 .Pn mhmail
meillo@138 3511 will become an ordinary MH tool, reading the profile.
meillo@138 3512 If, however, this idea will not convince, then
meillo@138 3513 .Pn mhmail
meillo@138 3514 will be removed.
meillo@138 3515 .P
meillo@138 3516 Every program in the mmh toolchest reads the profile.
meillo@138 3517 The only exception is
meillo@138 3518 .Pn slocal ,
meillo@138 3519 which is not considered part of the mmh toolchest.
meillo@138 3520 This MDA is only distributed with mmh, currently.
meillo@138 3521 Mmh has no
meillo@138 3522 .Pn post
meillo@138 3523 program, but
meillo@138 3524 .Pn spost ,
meillo@138 3525 which now reads the profile.
meillo@138 3526 .Ci 3e017a7abbdf69bf0dff7a4073275961eda1ded8
meillo@138 3527 With this change,
meillo@138 3528 .Pn send
meillo@138 3529 and
meillo@138 3530 .Pn spost
meillo@138 3531 can be considered to be merged.
meillo@138 3532 .Pn spost
meillo@169 3533 is only invoked directly by the to-be-changed
meillo@138 3534 .Pn mhmail
meillo@138 3535 implementation and by
meillo@138 3536 .Pn rcvdist ,
meillo@138 3537 which will require rework.
meillo@138 3538 .P
meillo@138 3539 The
meillo@138 3540 .Fu context_foil()
meillo@138 3541 function to pretend to have read an empty profile was removed.
meillo@138 3542 .Ci 68af8da96bea87a5541988870130b6209ce396f6
meillo@138 3543 All mmh tools read the profile.
meillo@133 3544
meillo@133 3545
meillo@127 3546
meillo@121 3547 .H2 "Standard Libraries
meillo@22 3548 .P
meillo@121 3549 MH is one decade older than the POSIX and ANSI C standards.
meillo@121 3550 Hence, MH included own implementations of functions
meillo@121 3551 that are standardized and thus widely available today,
meillo@121 3552 but were not back then.
meillo@121 3553 Today, twenty years after the POSIX and ANSI C were published,
meillo@121 3554 developers can expect system to comply with these standards.
meillo@121 3555 In consequence, MH-specific replacements for standard functions
meillo@121 3556 can and should be dropped.
meillo@121 3557 Kernighan and Pike advise: ``Use standard libraries.''
meillo@121 3558 .[ [
meillo@121 3559 kernighan pike practice of programming
meillo@121 3560 .], p. 196]
meillo@121 3561 Actually, MH had followed this advice in history,
meillo@121 3562 but it had not adjusted to the changes in this field.
meillo@121 3563 The
meillo@121 3564 .Fu snprintf()
meillo@121 3565 function, for instance, was standardized with C99 and is available
meillo@121 3566 almost everywhere because of its high usefulness.
meillo@123 3567 In project's own implementation of
meillo@121 3568 .Fu snprintf()
meillo@123 3569 was dropped in March 2012 in favor for using the one of the
meillo@123 3570 standard library.
meillo@123 3571 .Ci 0052f1024deb0a0a2fc2e5bacf93d45a5a9c9b32
meillo@123 3572 Such decisions limit the portability of mmh
meillo@121 3573 if systems don't support these standardized and widespread functions.
meillo@123 3574 This compromise is made because mmh focuses on the future.
meillo@121 3575 .P
meillo@123 3576 I am not yet thirty years old and my C and Unix experience comprises
meillo@123 3577 only half a dozen years.
meillo@121 3578 Hence, I need to learn about the history in retrospective.
meillo@121 3579 I have not used those ancient constructs myself.
meillo@121 3580 I have not suffered from their incompatibilities.
meillo@121 3581 I have not longed for standardization.
meillo@121 3582 All my programming experience is from a time when ANSI C and POSIX
meillo@121 3583 were well established already.
meillo@121 3584 I have only read a lot of books about the (good) old times.
meillo@121 3585 This puts me in a difficult positions when working with old code.
meillo@123 3586 I need to freshly acquire knowledge about old code constructs and ancient
meillo@123 3587 programming styles, whereas older programmers know these things by
meillo@123 3588 heart from their own experience.
meillo@121 3589 .P
meillo@123 3590 Being aware of the situation, I rather let people with more historic
meillo@123 3591 experience replace ancient code constructs with standardized ones.
meillo@121 3592 Lyndon Nerenberg covered large parts of this task for the nmh project.
meillo@121 3593 He converted project-specific functions to POSIX replacements,
meillo@121 3594 also removing the conditionals compilation of now standardized features.
meillo@123 3595 Ken Hornstein and David Levine had their part in the work, too.
meillo@121 3596 Often, I only needed to pull over changes from nmh into mmh.
meillo@121 3597 These changes include many commits; these are among them:
meillo@121 3598 .Ci 768b5edd9623b7238e12ec8dfc409b82a1ed9e2d
meillo@121 3599 .Ci 0052f1024deb0a0a2fc2e5bacf93d45a5a9c9b32 .
meillo@102 3600 .P
meillo@123 3601 During my own work, I tidied up the \fIMH standard library\fP,
meillo@123 3602 .Fn libmh.a ,
meillo@123 3603 which is located in the
meillo@123 3604 .Fn sbr
meillo@123 3605 (``subroutines'') directory in the source tree.
meillo@123 3606 The MH library includes functions that mmh tools usually need.
meillo@123 3607 Among them are MH-specific functions for profile, context, sequence,
meillo@123 3608 and folder handling, but as well
meillo@123 3609 MH-independent functions, such as auxiliary string functions,
meillo@123 3610 portability interfaces and error-checking wrappers for critical
meillo@123 3611 functions of the standard library.
meillo@123 3612 .P
meillo@123 3613 I have replaced the
meillo@121 3614 .Fu atooi()
meillo@121 3615 function with calls to
meillo@123 3616 .Fu strtoul()
meillo@139 3617 with the third parameter, the base, set to eight.
meillo@121 3618 .Fu strtoul()
meillo@123 3619 is part of C89 and thus considered safe to use.
meillo@121 3620 .Ci c490c51b3c0f8871b6953bd0c74551404f840a74
meillo@102 3621 .P
meillo@121 3622 I did remove project-included fallback implementations of
meillo@121 3623 .Fu memmove()
meillo@121 3624 and
meillo@121 3625 .Fu strerror() ,
meillo@121 3626 although Peter Maydell had re-included them into nmh in 2008
meillo@121 3627 to support SunOS 4.
meillo@121 3628 Nevertheless, these functions are part of ANSI C.
meillo@121 3629 Systems that do not even provide full ANSI C support should not
meillo@121 3630 put a load on mmh.
meillo@121 3631 .Ci b067ff5c465a5d243ce5a19e562085a9a1a97215
meillo@121 3632 .P
meillo@121 3633 The
meillo@121 3634 .Fu copy()
meillo@121 3635 function copies the string in argument one to the location in two.
meillo@121 3636 In contrast to
meillo@121 3637 .Fu strcpy() ,
meillo@121 3638 it returns a pointer to the terminating null-byte in the destination area.
meillo@123 3639 The code was adjusted to replace
meillo@121 3640 .Fu copy()
meillo@123 3641 with
meillo@121 3642 .Fu strcpy() ,
meillo@121 3643 except within
meillo@121 3644 .Fu concat() ,
meillo@121 3645 where
meillo@121 3646 .Fu copy()
meillo@123 3647 was more convenient.
meillo@123 3648 Therefore, the definition of
meillo@121 3649 .Fu copy()
meillo@123 3650 was moved into the source file of
meillo@121 3651 .Fu concat()
meillo@123 3652 and its visibility is now limited to it.
meillo@121 3653 .Ci 552fd7253e5ee9e554c5c7a8248a6322aa4363bb
meillo@121 3654 .P
meillo@121 3655 The function
meillo@121 3656 .Fu r1bindex()
meillo@121 3657 had been a generalized version of
meillo@121 3658 .Fu basename()
meillo@121 3659 with minor differences.
meillo@121 3660 As all calls to
meillo@121 3661 .Fu r1bindex()
meillo@121 3662 had the slash (`/') as delimiter anyway,
meillo@121 3663 replacing
meillo@121 3664 .Fu r1bindex()
meillo@121 3665 with the more specific and better-named function
meillo@121 3666 .Fu basename()
meillo@121 3667 became desirable.
meillo@121 3668 Unfortunately, many of the 54 calls to
meillo@121 3669 .Fu r1bindex()
meillo@123 3670 depended on a special behavior,
meillo@121 3671 which differed from the POSIX specification for
meillo@121 3672 .Fu basename() .
meillo@121 3673 Hence,
meillo@121 3674 .Fu r1bindex()
meillo@121 3675 was kept but renamed to
meillo@123 3676 .Fu mhbasename() ,
meillo@123 3677 fixing the delimiter to the slash.
meillo@121 3678 .Ci 240013872c392fe644bd4f79382d9f5314b4ea60
meillo@121 3679 For possible uses of
meillo@121 3680 .Fu r1bindex()
meillo@121 3681 with a different delimiter,
meillo@121 3682 the ANSI C function
meillo@121 3683 .Fu strrchr()
meillo@121 3684 provides the core functionality.
meillo@121 3685 .P
meillo@121 3686 The
meillo@121 3687 .Fu ssequal()
meillo@121 3688 function \(en apparently for ``substring equal'' \(en
meillo@121 3689 was renamed to
meillo@121 3690 .Fu isprefix() ,
meillo@121 3691 because this is what it actually checks.
meillo@121 3692 .Ci c20b4fa14515c7ab388ce35411d89a7a92300711
meillo@121 3693 Its source file had included the following comments, no joke.
meillo@121 3694 .VS
meillo@121 3695 /*
meillo@121 3696 * THIS CODE DOES NOT WORK AS ADVERTISED.
meillo@121 3697 * It is actually checking if s1 is a PREFIX of s2.
meillo@121 3698 * All calls to this function need to be checked to see
meillo@121 3699 * if that needs to be changed. Prefix checking is cheaper, so
meillo@121 3700 * should be kept if it's sufficient.
meillo@121 3701 */
meillo@121 3702
meillo@121 3703 /*
meillo@121 3704 * Check if s1 is a substring of s2.
meillo@121 3705 * If yes, then return 1, else return 0.
meillo@121 3706 */
meillo@121 3707 VE
meillo@123 3708 Two months later, it was completely removed by replacing it with
meillo@123 3709 .Fu strncmp() .
meillo@123 3710 .Ci b0b1dd37ff515578cf7cba51625189eb34a196cb
meillo@121 3711
meillo@102 3712
meillo@102 3713
meillo@102 3714
meillo@133 3715
meillo@133 3716 .H2 "User Data Locations
meillo@133 3717 .P
meillo@133 3718 In nmh, a personal setup consists of the MH profile and the MH directory.
meillo@133 3719 The profile is a file named
meillo@133 3720 .Fn \&.mh_profile
meillo@133 3721 in the user's home directory.
meillo@133 3722 It contains the static configuration.
meillo@133 3723 It also contains the location of the MH directory in the profile entry
meillo@133 3724 .Pe Path .
meillo@133 3725 The MH directory contains the mail storage and is the first
meillo@133 3726 place to search for personal forms, scan formats, and similar
meillo@133 3727 configuration files.
meillo@133 3728 The location of the MH directory can be chosen freely by the user.
meillo@133 3729 The default and usual name is a directory named
meillo@133 3730 .Fn Mail
meillo@133 3731 in the home directory.
meillo@133 3732 .P
meillo@133 3733 The way MH data is splitted between profile and MH directory is a legacy.
meillo@133 3734 It is only sensible in a situation where the profile is the only
meillo@133 3735 configuration file.
meillo@133 3736 Why else should the mail storage and the configuration files be intermixed?
meillo@133 3737 They are different kinds of data:
meillo@133 3738 The data to be operated on and the configuration to change how
meillo@133 3739 tools operate.
meillo@133 3740 Splitting the configuration between the profile and the MH directory
meillo@133 3741 is bad.
meillo@133 3742 Merging the mail storage and the configuration in one directory is bad
meillo@133 3743 as well.
meillo@133 3744 As the mail storage and the configuration were not separated sensibly
meillo@133 3745 in the first place, I did it now.
meillo@133 3746 .P
meillo@133 3747 Personal mmh data is grouped by type, resulting in two distinct parts:
meillo@133 3748 The mail storage and the configuration.
meillo@133 3749 In mmh, the mail storage directory still contains all the messages,
meillo@133 3750 but, in exception of public sequences files, nothing else.
meillo@133 3751 In difference to nmh, the auxiliary configuration files are no longer
meillo@133 3752 located there.
meillo@133 3753 Therefore, the directory is no longer called the user's \fIMH directory\fP
meillo@133 3754 but his \fImail storage\fP.
meillo@133 3755 Its location is still user-chosen, with the default name
meillo@133 3756 .Fn Mail ,
meillo@133 3757 in the user's home directory.
meillo@133 3758 In mmh, the configuration is grouped together in
meillo@133 3759 the hidden directory
meillo@133 3760 .Fn \&.mmh
meillo@133 3761 in the user's home directory.
meillo@133 3762 This \fImmh directory\fP contains the context file, personal forms,
meillo@133 3763 scan formats, and the like, but also the user's profile, now named
meillo@133 3764 .Fn profile .
meillo@133 3765 The location of the profile is no longer fixed to
meillo@133 3766 .Fn $HOME/.mh_profile
meillo@133 3767 but to
meillo@133 3768 .Fn $HOME/.mmh/profile .
meillo@133 3769 Having both, the file
meillo@133 3770 .Fn $HOME/.mh_profile
meillo@133 3771 and the configuration directory
meillo@133 3772 .Fn $HOME/.mmh
meillo@133 3773 appeared to be inconsistent.
meillo@133 3774 The approach chosen for mmh is consistent, simple, and familiar to
meillo@133 3775 Unix users.
meillo@168 3776 .Ci 7030d7edb099bff36ded7548bb5380f7acab4f9b
meillo@133 3777 .P
meillo@168 3778 MH allows users to have multiple MH setups.
meillo@133 3779 Therefore, it is necessary to select a different profile.
meillo@133 3780 The profile is the single entry point to access the rest of a
meillo@133 3781 personal MH setup.
meillo@133 3782 In nmh, the environment variable
meillo@133 3783 .Ev MH
meillo@133 3784 could be used to specifiy a different profile.
meillo@133 3785 To operate in the same MH setup with a separate context,
meillo@133 3786 the
meillo@133 3787 .Ev MHCONTEXT
meillo@133 3788 environment variable could be used.
meillo@133 3789 This allows having own current folders and current messages in
meillo@133 3790 each terminal, for instance.
meillo@133 3791 In mmh, three environment variables are used.
meillo@133 3792 .Ev MMH
meillo@133 3793 overrides the default location of the mmh directory (\c
meillo@133 3794 .Fn .mmh ).
meillo@133 3795 .Ev MMHP
meillo@133 3796 and
meillo@133 3797 .Ev MMHC
meillo@133 3798 override the paths to the profile and context files, respectively.
meillo@133 3799 This approach allows the set of personal configuration files to be chosen
meillo@133 3800 independently from the profile, context, and mail storage.
meillo@168 3801 .Ci 7030d7edb099bff36ded7548bb5380f7acab4f9b
meillo@133 3802 .P
meillo@133 3803 The separation of the files by type is sensible and convenient.
meillo@133 3804 The new approach has no functional disadvantages,
meillo@133 3805 as every setup I can imagine can be implemented with both approaches,
meillo@133 3806 possibly even easier with the new approach.
meillo@133 3807 The main achievement of the change is the clear and sensible split
meillo@133 3808 between mail storage and configuration.
meillo@133 3809
meillo@133 3810
meillo@133 3811
meillo@133 3812
meillo@133 3813
meillo@118 3814 .H2 "Modularization
meillo@118 3815 .P
meillo@123 3816 The source code of the mmh tools is located in the
meillo@122 3817 .Fn uip
meillo@123 3818 (``user interface programs'') directory.
meillo@123 3819 Each tools has a source file with the same name.
meillo@122 3820 For example,
meillo@122 3821 .Pn rmm
meillo@122 3822 is built from
meillo@122 3823 .Fn uip/rmm.c .
meillo@123 3824 Some source files are used for multiple programs.
meillo@122 3825 For example
meillo@122 3826 .Fn uip/scansbr.c
meillo@123 3827 is used for both,
meillo@122 3828 .Pn scan
meillo@122 3829 and
meillo@122 3830 .Pn inc .
meillo@122 3831 In nmh, 49 tools were built from 76 source files.
meillo@123 3832 This is a ratio of 1.6 source files per program.
meillo@123 3833 32 programs depended on multiple source files;
meillo@123 3834 17 programs depended on one source file only.
meillo@122 3835 In mmh, 39 tools are built from 51 source files.
meillo@123 3836 This is a ratio of 1.3 source files per program.
meillo@123 3837 18 programs depend on multiple source files;
meillo@123 3838 21 programs depend on one source file only.
meillo@123 3839 (These numbers and the ones in the following text ignore the MH library
meillo@123 3840 as well as shell scripts and multiple names for the same program.)
meillo@122 3841 .P
meillo@123 3842 Splitting the source code of a large program into multiple files can
meillo@122 3843 increase the readability of its source code.
meillo@124 3844 Most of the mmh tools, however, are simple and straight-forward programs.
meillo@122 3845 With the exception of the MIME handling tools,
meillo@122 3846 .Pn pick
meillo@122 3847 is the largest tools.
meillo@122 3848 It contains 1\|037 lines of source code (measured with
meillo@122 3849 .Pn sloccount ), excluding the MH library.
meillo@122 3850 Only the MIME handling tools (\c
meillo@122 3851 .Pn mhbuild ,
meillo@122 3852 .Pn mhstore ,
meillo@122 3853 .Pn show ,
meillo@122 3854 etc.)
meillo@122 3855 are larger.
meillo@122 3856 Splitting programs with less than 1\|000 lines of code into multiple
meillo@123 3857 source files seldom leads to better readability.
meillo@123 3858 For such tools, splitting makes sense
meillo@122 3859 when parts of the code are reused in other programs,
meillo@122 3860 and the reused code fragment is not general enough
meillo@122 3861 for including it in the MH library,
meillo@124 3862 or, if the code has dependencies on a library that only few programs need.
meillo@122 3863 .Fn uip/packsbr.c ,
meillo@122 3864 for instance, provides the core program logic for the
meillo@122 3865 .Pn packf
meillo@122 3866 and
meillo@122 3867 .Pn rcvpack
meillo@122 3868 programs.
meillo@122 3869 .Fn uip/packf.c
meillo@122 3870 and
meillo@122 3871 .Fn uip/rcvpack.c
meillo@122 3872 mainly wrap the core function appropriately.
meillo@122 3873 No other tools use the folder packing functions.
meillo@123 3874 As another example,
meillo@123 3875 .Fn uip/termsbr.c
meillo@123 3876 provides termcap support, which requires linking with a termcap or
meillo@123 3877 curses library.
meillo@123 3878 Including
meillo@123 3879 .Fn uip/termsbr.c
meillo@123 3880 into the MH library would require every program to be linked with
meillo@123 3881 termcap or curses, although only few of the programs require it.
meillo@122 3882 .P
meillo@122 3883 The task of MIME handling is complex enough that splitting its code
meillo@122 3884 into multiple source files improves the readability.
meillo@122 3885 The program
meillo@122 3886 .Pn mhstore ,
meillo@122 3887 for instance, is compiled out of seven source files with 2\|500
meillo@122 3888 lines of code in summary.
meillo@122 3889 The main code file
meillo@122 3890 .Fn uip/mhstore.c
meillo@123 3891 consists of 800 lines; the other 1\|700 lines of code are reused in
meillo@123 3892 other MIME handling tools.
meillo@123 3893 It seems to be worthwhile to bundle the generic MIME handling code into
meillo@123 3894 a MH-MIME library, as a companion to the MH standard library.
meillo@122 3895 This is left open for the future.
meillo@122 3896 .P
meillo@169 3897 The work already accomplished focussed on the non-MIME tools.
meillo@122 3898 The amount of code compiled into each program was reduced.
meillo@123 3899 This eases the understanding of the code base.
meillo@122 3900 In nmh,
meillo@122 3901 .Pn comp
meillo@122 3902 was built from six source files:
meillo@122 3903 .Fn comp.c ,
meillo@122 3904 .Fn whatnowproc.c ,
meillo@122 3905 .Fn whatnowsbr.c ,
meillo@122 3906 .Fn sendsbr.c ,
meillo@122 3907 .Fn annosbr.c ,
meillo@122 3908 and
meillo@122 3909 .Fn distsbr.c .
meillo@122 3910 In mmh, it builds from only two:
meillo@122 3911 .Fn comp.c
meillo@122 3912 and
meillo@122 3913 .Fn whatnowproc.c .
meillo@123 3914 In nmh's
meillo@123 3915 .Pn comp ,
meillo@123 3916 the core function of
meillo@122 3917 .Pn whatnow ,
meillo@122 3918 .Pn send ,
meillo@122 3919 and
meillo@122 3920 .Pn anno
meillo@123 3921 were compiled into
meillo@122 3922 .Pn comp .
meillo@123 3923 This saved the need to execute these programs with
meillo@122 3924 .Fu fork()
meillo@122 3925 and
meillo@122 3926 .Fu exec() ,
meillo@122 3927 two expensive system calls.
meillo@122 3928 Whereis this approach improved the time performance,
meillo@122 3929 it interweaved the source code.
meillo@122 3930 Core functionalities were not encapsulated into programs but into
meillo@122 3931 function, which were then wrapped by programs.
meillo@122 3932 For example,
meillo@122 3933 .Fn uip/annosbr.c
meillo@122 3934 included the function
meillo@122 3935 .Fu annotate() .
meillo@122 3936 Each program that wanted to annotate messages, included the source file
meillo@123 3937 .Fn uip/annosbr.c
meillo@123 3938 and called
meillo@123 3939 .Fu annotate() .
meillo@123 3940 Because the function
meillo@123 3941 .Fu annotate()
meillo@123 3942 was used like the tool
meillo@123 3943 .Pn anno ,
meillo@123 3944 it had seven parameters, reflecting the command line switches of the tool.
meillo@122 3945 When another pair of command line switches was added to
meillo@122 3946 .Pn anno ,
meillo@122 3947 a rather ugly hack was implemented to avoid adding another parameter
meillo@122 3948 to the function.
meillo@122 3949 .Ci d9b1d57351d104d7ec1a5621f090657dcce8cb7f
meillo@122 3950 .P
meillo@122 3951 Separation simplifies the understanding of program code
meillo@122 3952 because the area influenced by any particular statement is smaller.
meillo@122 3953 The separating on the program-level is more strict than the separation
meillo@122 3954 on the function level.
meillo@122 3955 In mmh, the relevant code of
meillo@122 3956 .Pn comp
meillo@122 3957 comprises the two files
meillo@122 3958 .Fn uip/comp.c
meillo@122 3959 and
meillo@122 3960 .Fn uip/whatnowproc.c ,
meillo@123 3961 together 210 lines of code.
meillo@122 3962 In nmh,
meillo@122 3963 .Pn comp
meillo@122 3964 comprises six files with 2\|450 lines.
meillo@123 3965 Not all of the code in these six files was actually used by
meillo@122 3966 .Pn comp ,
meillo@123 3967 but the code reader needed to read all of the code first to know which
meillo@123 3968 parts were used.
meillo@122 3969 .P
meillo@123 3970 As I have read a lot in the code base during the last two years,
meillo@123 3971 I learned about the easy and the difficult parts.
meillo@123 3972 Code is easy to understand if:
meillo@123 3973 .BU
meillo@139 3974 The influenced code area is small.
meillo@123 3975 .BU
meillo@139 3976 The boundaries are strictly defined.
meillo@123 3977 .BU
meillo@139 3978 The code is written straight-forward.
meillo@123 3979 .P
meillo@123 3980 .\" XXX move this paragraph somewhere else?
meillo@123 3981 Reading
meillo@122 3982 .Pn rmm 's
meillo@122 3983 source code in
meillo@122 3984 .Fn uip/rmm.c
meillo@122 3985 is my recommendation for a beginner's entry point into the code base of nmh.
meillo@122 3986 The reasons are that the task of
meillo@122 3987 .Pn rmm
meillo@122 3988 is straight forward and it consists of one small source code file only,
meillo@122 3989 yet its source includes code constructs typical for MH tools.
meillo@122 3990 With the introduction of the trash folder in mmh,
meillo@122 3991 .Pn rmm
meillo@122 3992 became a bit more complex, because it invokes
meillo@122 3993 .Pn refile .
meillo@122 3994 Still, it is a good example for a simple tool with clear sources.
meillo@122 3995 .P
meillo@122 3996 Understanding
meillo@122 3997 .Pn comp
meillo@122 3998 requires to read 210 lines of code in mmh, but ten times as much in nmh.
meillo@123 3999 Due to the aforementioned hack in
meillo@122 4000 .Pn anno
meillo@122 4001 to save the additional parameter, information passed through the program's
meillo@122 4002 source base in obscure ways.
meillo@123 4003 Thus, understanding
meillo@122 4004 .Pn comp ,
meillo@123 4005 required understanding the inner workings of
meillo@122 4006 .Fn uip/annosbr.c
meillo@122 4007 first.
meillo@123 4008 To be sure to fully understand a program, its whole source code needs
meillo@122 4009 to be examined.
meillo@123 4010 Not doing so is a leap of faith, assuming that the developers
meillo@122 4011 have avoided obscure programming techniques.
meillo@122 4012 By separating the tools on the program-level, the boundaries are
meillo@122 4013 clearly visible and technically enforced.
meillo@122 4014 The interfaces are calls to
meillo@122 4015 .Fu exec()
meillo@122 4016 rather than arbitrary function calls.
meillo@123 4017 .P
meillo@123 4018 But the real problem is another:
meillo@123 4019 Nmh violates the golden ``one tool, one job'' rule of the Unix philosophy.
meillo@123 4020 Understanding
meillo@122 4021 .Pn comp
meillo@123 4022 requires understanding
meillo@123 4023 .Fn uip/annosbr.c
meillo@123 4024 and
meillo@123 4025 .Fn uip/sendsbr.c
meillo@123 4026 because
meillo@123 4027 .Pn comp
meillo@123 4028 does annotate and send messages.
meillo@123 4029 In nmh, there surely exists the tool
meillo@122 4030 .Pn send ,
meillo@123 4031 which does (almost) only send messages.
meillo@123 4032 But
meillo@122 4033 .Pn comp
meillo@123 4034 and
meillo@122 4035 .Pn repl
meillo@122 4036 and
meillo@122 4037 .Pn forw
meillo@122 4038 and
meillo@122 4039 .Pn dist
meillo@122 4040 and
meillo@122 4041 .Pn whatnow
meillo@122 4042 and
meillo@123 4043 .Pn viamail ,
meillo@123 4044 they all (!) have the same message sending function included, too.
meillo@123 4045 In result,
meillo@123 4046 .Pn comp
meillo@123 4047 sends messages without using
meillo@123 4048 .Pn send .
meillo@123 4049 The situation is the same as if
meillo@123 4050 .Pn grep
meillo@123 4051 would page without
meillo@123 4052 .Pn more
meillo@123 4053 just because both programs are part of the same code base.
meillo@123 4054 .P
meillo@122 4055 The clear separation on the surface \(en the toolchest approach \(en
meillo@123 4056 is violated on the level below.
meillo@122 4057 This violation is for the sake of time performance.
meillo@122 4058 On systems where
meillo@122 4059 .Fu fork()
meillo@122 4060 and
meillo@122 4061 .Fu exec()
meillo@122 4062 are expensive, the quicker response might be noticable.
meillo@124 4063 In the old times, sacrificing readability and conceptional beauty for
meillo@124 4064 speed might even have been a must to prevent MH from being unusably slow.
meillo@122 4065 Whatever the reasons had been, today they are gone.
meillo@123 4066 No longer should we sacrifice readability or conceptional beauty.
meillo@122 4067 No longer should we violate the Unix philosophy's ``one tool, one job''
meillo@122 4068 guideline.
meillo@123 4069 No longer should we keep speed improvements that became unnecessary.
meillo@122 4070 .P
meillo@123 4071 Therefore, mmh's
meillo@123 4072 .Pn comp
meillo@123 4073 does no longer send messages.
meillo@123 4074 In mmh, different jobs are divided among separate programs that
meillo@122 4075 invoke each other as needed.
meillo@123 4076 In consequence,
meillo@123 4077 .Pn comp
meillo@123 4078 invokes
meillo@123 4079 .Pn whatnow
meillo@123 4080 which thereafter invokes
meillo@123 4081 .Pn send .
meillo@168 4082 .Ci 3df5ab3c116e6d4a2fb4bb5cc9dfc5f781825815
meillo@168 4083 .Ci c73c00bfccd22ec77e9593f47462aeca4a8cd9c0
meillo@123 4084 The clear separation on the surface is maintained on the level below.
meillo@123 4085 Human users and the tools use the same interface \(en
meillo@123 4086 annotations, for example, are made by invoking
meillo@123 4087 .Pn anno ,
meillo@123 4088 no matter if requested by programs or by human beings.
meillo@168 4089 .Ci 469a4163c2a1a43731d412eaa5d9cae7d670c48b
meillo@168 4090 .Ci aed384169af5204b8002d06e7a22f89197963d2d
meillo@168 4091 .Ci 3caf9e298a8861729ca8b8a84f57022b6f3ea742
meillo@123 4092 The decrease of tools built from multiple source files and thus
meillo@123 4093 the decrease of
meillo@123 4094 .Fn uip/*sbr.c
meillo@123 4095 files confirm the improvement.
meillo@168 4096 .Ci 9e6d91313f01c96b4058d6bf419a8ca9a207bc33
meillo@168 4097 .ci 81744a46ac9f845d6c2b9908074d269275178d2e
meillo@168 4098 .Ci f0f858069d21111f0dbea510044593f89c9b0829
meillo@168 4099 .Ci 0503a6e9be34f24858b55b555a5c948182b9f24b
meillo@168 4100 .Ci 27826f9353e0f0b04590b7d0f8f83e60462b90f0
meillo@168 4101 .Ci d1da1f94ce62160aebb30df4063ccbc53768656b
meillo@168 4102 .Ci c42222869e318fff5dec395eca3e776db3075455
meillo@123 4103 .P
meillo@145 4104 .\" XXX move this paragraph up somewhere
meillo@123 4105 One disadvantage needs to be taken with this change:
meillo@123 4106 The compiler can no longer check the integrity of the interfaces.
meillo@123 4107 By changing the command line interfaces of tools, it is
meillo@123 4108 the developer's job to adjust the invocations of these tools as well.
meillo@123 4109 As this is a manual task and regression tests, which could detect such
meillo@124 4110 problems, are not available yet, it is prone to errors.
meillo@123 4111 These errors will not be detected at compile time but at run time.
meillo@123 4112 Installing regression tests is a task left to do.
meillo@123 4113 In the best case, a uniform way of invoking tools from other tools
meillo@123 4114 can be developed to allow automated testing at compile time.
meillo@145 4115
meillo@145 4116
meillo@145 4117 .ig
meillo@145 4118 XXX consider writing about mhl vs. mhlproc
meillo@145 4119
meillo@145 4120 sbr/showfile.c
meillo@145 4121
meillo@145 4122 23 /*
meillo@145 4123 24 ** If you have your lproc listed as "mhl",
meillo@145 4124 25 ** then really invoked the mhlproc instead
meillo@145 4125 26 ** (which is usually mhl anyway).
meillo@145 4126 27 */
meillo@145 4127
meillo@145 4128 Sat Nov 24 19:09:14 1984 /mtr (agent: Marshall Rose) <uci@udel-dewey>
meillo@145 4129
meillo@145 4130 sbr/showfile.c: if lproc is "mhl", use mhlproc for consistency
meillo@145 4131 (Actually, user should use "lproc: show", "showproc: mhl".)
meillo@145 4132 ..