docs/master

annotate discussion.roff @ 228:a1468cf505fd

find changesets by author, revision, files, or words in the commit message
Spellchecked.
author markus schnalke <meillo@marmaro.de>
date Sun, 15 Jul 2012 23:58:14 +0200
parents 157c92fc1597
children 96778c1afc3e
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@217 5 It is structured along the goals chosen for the project.
meillo@217 6 A selection of the work undertaken
meillo@217 7 is described.
meillo@217 8 .P
meillo@217 9 This discussion compares the present version of mmh with the state of
meillo@217 10 nmh at the time when the mmh project had started, i.e. fall 2011.
meillo@224 11 Recent changes in nmh are rarely part of the discussion.
meillo@187 12 .P
meillo@187 13 For the reader's convenience, the structure of modern email systems
meillo@217 14 is depicted in the following figure.
meillo@187 15 It illustrates the path a message takes from sender to recipient.
meillo@217 16
meillo@217 17 .sp 1.5
meillo@187 18 .KS
meillo@187 19 .in 2c
meillo@187 20 .so input/mail-agents.pic
meillo@187 21 .KE
meillo@217 22 .sp 1.5
meillo@217 23
meillo@187 24 .LP
meillo@217 25 The ellipses denote mail agents, i.e. different jobs in email processing.
meillo@217 26 These are:
meillo@187 27 .IP "Mail User Agent (MUA)
meillo@217 28 The only program users directly interact with.
meillo@187 29 It includes functions to compose new mail, display received mail,
meillo@187 30 and to manage the mail storage.
meillo@217 31 It is called a \fImail client\fP as well.
meillo@187 32 .IP "Mail Submission Agent (MSA)
meillo@187 33 A special kind of Mail Transfer Agent, used to submit mail into the
meillo@187 34 mail transport system.
meillo@217 35 Often it is also called an MTA.
meillo@187 36 .IP "Mail Transfer Agent (MTA)
meillo@187 37 A node in the mail transport system.
meillo@217 38 It transfers incoming mail to a transport node nearer to the
meillo@217 39 final destination.
meillo@217 40 An MTA may be the final destination itself.
meillo@187 41 .IP "Mail Delivery Agent (MDA)
meillo@217 42 Delivers mail according to a set of rules.
meillo@217 43 Usually, the messages are stored to disk.
meillo@187 44 .IP "Mail Retrieval Agent (MRA)
meillo@217 45 Initiates the transfer of mail from a remote location to the local machine.
meillo@217 46 (The dashed arrow in the figure represents the pull request.)
meillo@217 47 .LP
meillo@217 48 The dashed boxes represent entities that usually reside on single machines.
meillo@217 49 The box on the lower left represents the sender's system.
meillo@187 50 The box on the upper left represents the first mail transfer node.
meillo@224 51 The box on the upper right represents the transfer node responsible
meillo@224 52 for the destination address.
meillo@217 53 The box on the lower right represents the recipient's system.
meillo@187 54 Often, the boxes above the dotted line are servers on the Internet.
meillo@217 55 Many mail clients, including nmh, include all of the components below
meillo@217 56 the dotted line.
meillo@217 57 This is not the case for mmh; it implements the MUA only.
meillo@187 58
meillo@187 59
meillo@187 60
meillo@58 61
meillo@58 62
meillo@58 63
meillo@133 64 .\" --------------------------------------------------------------
meillo@125 65 .H1 "Streamlining
meillo@58 66
meillo@0 67 .P
meillo@217 68 MH once provided a complete email system.
meillo@217 69 The community around nmh tries to keep nmh in similar shape.
meillo@178 70 In fundamental contrast, mmh shall be an MUA only.
meillo@87 71 I believe that the development of all-in-one mail systems is obsolete.
meillo@173 72 Today, email is too complex to be fully covered by a single project.
meillo@173 73 Such a project will not be able to excel in all aspects.
meillo@159 74 Instead, the aspects of email should be covered by multiple projects,
meillo@87 75 which then can be combined to form a complete system.
meillo@169 76 Excellent implementations for the various aspects of email already exist.
meillo@223 77 Just to name three examples: Postfix is a specialized MTA, Procmail
meillo@223 78 is a specialized MDA, and Fetchmail is a specialized MRA.
meillo@89 79 I believe that it is best to use such specialized tools instead of
meillo@217 80 providing the same function once more as a side component.
meillo@58 81 .P
meillo@169 82 Doing something well requires focusing on a small set of specific aspects.
meillo@217 83 Under the assumption that development which is focussed on a particular
meillo@217 84 area produces better results there, specialized projects will be superior
meillo@87 85 in their field of focus.
meillo@87 86 Hence, all-in-one mail system projects \(en no matter if monolithic
meillo@87 87 or modular \(en will never be the best choice in any of the fields.
meillo@217 88 Even in providing the most consistent all-in-one system, they are likely
meillo@217 89 to be beaten by projects that focus exclusively on the creation
meillo@217 90 of a homogeneous system by integrating existing mail components.
meillo@87 91 .P
meillo@217 92 Usually, the limiting resource in the community development of
meillo@217 93 free software is man power.
meillo@217 94 If the development effort is spread over a large development area,
meillo@217 95 it becomes more difficult to compete with the specialists in the
meillo@87 96 various fields.
meillo@87 97 The concrete situation for MH-based mail systems is even tougher,
meillo@169 98 given their small and aged community, concerning both developers and users.
meillo@87 99 .P
meillo@87 100 In consequence, I believe that the available development resources
meillo@100 101 should focus on the point where MH is most unique.
meillo@87 102 This is clearly the user interface \(en the MUA.
meillo@125 103 Peripheral parts should be removed to streamline mmh for the MUA task.
meillo@60 104
meillo@60 105
meillo@100 106 .H2 "Mail Transfer Facilities
meillo@154 107 .Id mail-transfer-facilities
meillo@60 108 .P
meillo@217 109 The removal of the mail transfer facilities, effectively dropping the
meillo@217 110 MSA and MRA, had been the first work task in the mmh project.
meillo@217 111 The desire for this change initiated the creation of the mmh project.
meillo@60 112 .P
meillo@169 113 Focusing on one mail agent role only, is motivated by Eric Allman's
meillo@105 114 experience with Sendmail.
meillo@217 115 He identified the limitation of Sendmail
meillo@217 116 to the MTA task as one reason for its success:
meillo@105 117 .[ [
meillo@105 118 costales sendmail
meillo@105 119 .], p. xviii]
meillo@105 120 .QS
meillo@105 121 Second, I limited myself to the routing function \(en
meillo@110 122 I wouldn't write user agents or delivery back-ends.
meillo@217 123 This was a departure of the dominant thought of the time,
meillo@105 124 in which routing logic, local delivery, and often the network code
meillo@105 125 were incorporated directly into the user agents.
meillo@105 126 .QE
meillo@105 127 .P
meillo@187 128 In nmh, the MSA is called \fIMessage Transfer Service\fP (MTS).
meillo@105 129 This facility, implemented by the
meillo@105 130 .Pn post
meillo@217 131 command, establishes network connections and spoke SMTP to submit
meillo@159 132 messages to be relayed to the outside world.
meillo@217 133 When email transfer changed, this part needed to be changed as well.
meillo@89 134 Encryption and authentication for network connections
meillo@87 135 needed to be supported, hence TLS and SASL were introduced into nmh.
meillo@217 136 This added complexity without improving the core functions.
meillo@217 137 Furthermore, keeping up with recent developments in the field of
meillo@87 138 mail transfer requires development power and specialists.
meillo@217 139 In mmh, this whole facility was simply cut off
meillo@76 140 .Ci f6aa95b724fd8c791164abe7ee5468bf5c34f226
meillo@76 141 .Ci fecd5d34f65597a4dfa16aeabea7d74b191532c3
meillo@217 142 .Ci 156d35f6425bea4c1ed3c4c79783dc613379c65b .
meillo@87 143 Instead, mmh depends on an external MSA.
meillo@217 144 All outgoing mail in mmh goes through the
meillo@60 145 .Pn sendmail
meillo@87 146 command, which almost any MSA provides.
meillo@87 147 If not, a wrapper program can be written.
meillo@87 148 It must read the message from the standard input, extract the
meillo@87 149 recipient addresses from the message header, and hand the message
meillo@87 150 over to the MSA.
meillo@223 151 For example, a wrapper script for qmail would be:
meillo@87 152 .VS
meillo@87 153 #!/bin/sh
meillo@138 154 exec qmail-inject # ignore command line arguments
meillo@87 155 VE
meillo@87 156 The requirement to parse the recipient addresses out of the message header
meillo@217 157 may be removed in the future.
meillo@217 158 Mmh could pass the recipient addresses as command line arguments.
meillo@100 159 This appears to be the better interface.
meillo@60 160 .P
meillo@60 161 To retrieve mail, the
meillo@60 162 .Pn inc
meillo@217 163 command in nmh acts as MRA.
meillo@217 164 It establishes network connections
meillo@217 165 and speaks POP3 to retrieve mail from remote servers.
meillo@76 166 As with mail submission, the network connections required encryption and
meillo@217 167 authentication, thus TLS and SASL were added to nmh.
meillo@169 168 Support for message retrieval through IMAP will soon become necessary
meillo@169 169 additions, too, and likewise for any other changes in mail transfer.
meillo@217 170 But not in mmh because it has dropped the support for retrieving mail
meillo@217 171 from remote locations
meillo@217 172 .Ci ab7b48411962d26439f92f35ed084d3d6275459c .
meillo@76 173 Instead, it depends on an external tool to cover this task.
meillo@169 174 Mmh has two paths for messages to enter mmh's mail storage:
meillo@100 175 (1) Mail can be incorporated with
meillo@60 176 .Pn inc
meillo@87 177 from the system maildrop, or (2) with
meillo@60 178 .Pn rcvstore
meillo@87 179 by reading them, one at a time, from the standard input.
meillo@60 180 .P
meillo@217 181 With the removal of the MSA and MRA, mmh converted from a complete
meillo@217 182 mail system to only an MUA.
meillo@60 183 Now, of course, mmh depends on third-party software.
meillo@87 184 An external MSA is required to transfer mail to the outside world;
meillo@60 185 an external MRA is required to retrieve mail from remote machines.
meillo@217 186 Excellent implementations of such software exist.
meillo@217 187 They likely are superior to the internal versions that were removed.
meillo@217 188 Additionally, the best suiting programs can be chosen freely.
meillo@60 189 .P
meillo@217 190 As it had already been possible to use an external MSA and MRA,
meillo@217 191 why should the internal version not be kept for convenience?
meillo@228 192 Transferred to a different area, the question,
meillo@224 193 whether there is sense in having a fall-back pager in all
meillo@224 194 the command line tools for the cases when
meillo@60 195 .Pn more
meillo@60 196 or
meillo@60 197 .Pn less
meillo@173 198 are not available, appears to be ridiculous.
meillo@100 199 Of course, MSAs and MRAs are more complex than text pagers
meillo@87 200 and not necessarily available but still the concept of orthogonal
meillo@217 201 design holds: ``Write programs that do one thing and do it well''.
meillo@87 202 .[
meillo@87 203 mcilroy unix phil
meillo@87 204 p. 53
meillo@87 205 .]
meillo@87 206 .[
meillo@87 207 mcilroy bstj foreword
meillo@87 208 .]
meillo@87 209 Here, this part of the Unix philosophy was applied not only
meillo@87 210 to the programs but to the project itself.
meillo@87 211 In other words:
meillo@164 212 Develop projects that focus on one thing and do it well.
meillo@169 213 Projects which have grown complex should be split, for the same
meillo@169 214 reasons that programs which have grown complex should be split.
meillo@100 215 If it is conceptionally more elegant to have the MSA and MRA as
meillo@87 216 separate projects then they should be separated.
meillo@217 217 In my opinion, this is the case.
meillo@217 218 The RFCs suggest this separation by clearly distinguishing the
meillo@217 219 different mail handling tasks [RFC\|821].
meillo@217 220 The small interfaces between the mail agents support the
meillo@217 221 separation as well.
meillo@76 222 .P
meillo@217 223 Once, email had been small and simple.
meillo@100 224 At that time,
meillo@60 225 .Pn /bin/mail
meillo@169 226 had covered everything there was to email and still was small and simple.
meillo@100 227 Later, the essential complexity of email increased.
meillo@224 228 (Essential complexity is the complexity defined by the problem itself
meillo@217 229 .[ [
meillo@87 230 brooks no silver bullet
meillo@224 231 .]].)
meillo@217 232 Consequently, email systems grew.
meillo@100 233 RFCs started to introduce the concept of mail agents to separate the
meillo@217 234 various roles because they became more extensive and because
meillo@217 235 new roles appeared.
meillo@217 236 As mail system implementations grew, parts of them were split off.
meillo@169 237 For instance, a POP server was included in the original MH;
meillo@169 238 it was removed in nmh.
meillo@217 239 Now is the time to go one step further and split off the MSA and MRA,
meillo@217 240 as well.
meillo@87 241 Not only does this decrease the code size of the project,
meillo@169 242 more importantly, it unburdens mmh of the whole field of
meillo@217 243 message transfer, with all its implications for the project.
meillo@169 244 There is no more need for concern with changes in network transfer.
meillo@217 245 This independence is gained by depending on external components
meillo@217 246 that cover the field.
meillo@60 247 .P
meillo@217 248 In general, functionality can be added in three different ways:
meillo@171 249 .LI 1
meillo@217 250 By implementing the function in the project itself.
meillo@171 251 .LI 2
meillo@217 252 By depending on a library that provides the function.
meillo@171 253 .LI 3
meillo@217 254 By depending on a program that provides the function.
meillo@171 255 .LP
meillo@169 256 While implementing the function in the project itself leads to the
meillo@169 257 largest increase in code size and requires the most maintenance
meillo@169 258 and development work,
meillo@217 259 it keeps the project's dependence on other software lowest.
meillo@169 260 Using libraries or external programs requires less maintenance work
meillo@217 261 but introduces dependencies on external projects.
meillo@169 262 Programs have the smallest interfaces and provide the best separation,
meillo@87 263 but possibly limit the information exchange.
meillo@169 264 External libraries are more strongly connected than external programs,
meillo@169 265 thus information can be exchanged in a more flexible manner.
meillo@227 266 Obviously, adding code to a project increases the maintenance work.
meillo@217 267 As implementing complex functions in the project itself adds
meillo@217 268 a lot of code, this should be avoided if possible.
meillo@217 269 Thus, the dependencies only change in their character,
meillo@169 270 not in their existence.
meillo@66 271 In mmh, library dependencies on
meillo@66 272 .Pn libsasl2
meillo@66 273 and
meillo@66 274 .Pn libcrypto /\c
meillo@66 275 .Pn libssl
meillo@159 276 were traded against program dependencies on an MSA and an MRA.
meillo@159 277 This also meant trading build-time dependencies against run-time
meillo@87 278 dependencies.
meillo@169 279 Besides providing stronger separation and greater flexibility,
meillo@169 280 program dependencies also allowed
meillo@66 281 over 6\|000 lines of code to be removed from mmh.
meillo@66 282 This made mmh's code base about 12\|% smaller.
meillo@87 283 Reducing the project's code size by such an amount without actually
meillo@87 284 losing functionality is a convincing argument.
meillo@87 285 Actually, as external MSAs and MRAs are likely superior to the
meillo@87 286 project's internal versions, the common user even gains functionality.
meillo@66 287 .P
meillo@169 288 Users of MH should not have problems setting up an external MSA and MRA.
meillo@60 289 Also, the popular MSAs and MRAs have large communities and a lot
meillo@169 290 of available documentation.
meillo@217 291
meillo@223 292 Choices for MSAs range from small forwarders such as \fIssmtp\fP and
meillo@223 293 \fInullmailer\fP, over mid-size MTAs including \fImasqmail\fP and
meillo@223 294 \fIdma\fP, up to full-featured MTAs as for instance \fIPostfix\fP.
meillo@223 295 MRAs are provided for example by \fIfetchmail\fP, \fIgetmail\fP,
meillo@223 296 \fImpop\fP, and \fIfdm\fP.
meillo@60 297
meillo@60 298
meillo@100 299 .H2 "Non-MUA Tools
meillo@60 300 .P
meillo@217 301 One goal of mmh is to remove the tools that do not significantly
meillo@217 302 contribute to the MUA's job.
meillo@217 303 Loosely related and rarely used tools distract from a lean appearance,
meillo@217 304 and require maintenance work without adding much to the core task.
meillo@217 305 By removing these tools, mmh became more streamlined and focused.
meillo@62 306 .BU
meillo@58 307 .Pn conflict
meillo@87 308 was removed
meillo@76 309 .Ci 8b235097cbd11d728c07b966cf131aa7133ce5a9
meillo@217 310 because it is a mail system maintenance tool and not MUA-related.
meillo@87 311 It even checked
meillo@58 312 .Fn /etc/passwd
meillo@58 313 and
meillo@58 314 .Fn /etc/group
meillo@87 315 for consistency, which is completely unrelated to email.
meillo@87 316 A tool like
meillo@87 317 .Pn conflict
meillo@87 318 is surely useful, but it should not be shipped with mmh.
meillo@62 319 .BU
meillo@58 320 .Pn rcvtty
meillo@87 321 was removed
meillo@87 322 .Ci 14767c94b3827be7c867196467ed7aea5f6f49b0
meillo@89 323 because its use case of writing to the user's terminal
meillo@200 324 on reception of mail is obsolete.
meillo@87 325 If users like to be informed of new mail, the shell's
meillo@58 326 .Ev MAILPATH
meillo@87 327 variable or graphical notifications are technically more appealing.
meillo@217 328 Writing to terminals directly is hardly ever desired today.
meillo@169 329 If, though, one prefers this approach, the standard tool
meillo@58 330 .Pn write
meillo@58 331 can be used in a way similar to:
meillo@82 332 .VS
meillo@58 333 scan -file - | write `id -un`
meillo@82 334 VE
meillo@62 335 .BU
meillo@58 336 .Pn viamail
meillo@87 337 was removed
meillo@87 338 .Ci eda72d6a7a7c20ff123043fb7f19c509ea01f932
meillo@87 339 when the new attachment system was activated, because
meillo@58 340 .Pn forw
meillo@76 341 could then cover the task itself.
meillo@62 342 The program
meillo@58 343 .Pn sendfiles
meillo@62 344 was rewritten as a shell script wrapper around
meillo@58 345 .Pn forw .
meillo@76 346 .Ci 0e82199cf3c991a173e0ac8aa776efdb3ded61e6
meillo@62 347 .BU
meillo@58 348 .Pn msgchk
meillo@87 349 was removed
meillo@87 350 .Ci bb9360ead7eb7a3fedcce2eeedfc660014e41dbe ,
meillo@87 351 because it lost its use case when POP support was removed.
meillo@76 352 A call to
meillo@58 353 .Pn msgchk
meillo@87 354 provided hardly more information than:
meillo@82 355 .VS
meillo@58 356 ls -l /var/mail/meillo
meillo@82 357 VE
meillo@217 358 Yet, it distinguished between old and new mail, but
meillo@169 359 these details can be retrieved with
meillo@76 360 .Pn stat (1),
meillo@62 361 too.
meillo@100 362 A small shell script could be written to print the information
meillo@76 363 in a similar way, if truly necessary.
meillo@76 364 As mmh's
meillo@76 365 .Pn inc
meillo@87 366 only incorporates mail from the user's local maildrop,
meillo@62 367 and thus no data transfers over slow networks are involved,
meillo@169 368 there is hardly any need to check for new mail before incorporating it.
meillo@62 369 .BU
meillo@58 370 .Pn msh
meillo@87 371 was removed
meillo@76 372 .Ci 916690191222433a6923a4be54b0d8f6ac01bd02
meillo@87 373 because the tool was in conflict with the philosophy of MH.
meillo@217 374 It provided an interactive shell to access the features of MH.
meillo@217 375 However, it was not just a shell tailored to the needs of mail handling,
meillo@217 376 but one large program that had several MH tools built in.
meillo@217 377 This conflicted with the major feature of MH of being a tool chest.
meillo@76 378 .Pn msh 's
meillo@159 379 main use case had been accessing Bulletin Boards, which have ceased to
meillo@62 380 be popular.
meillo@62 381 .P
meillo@62 382 Removing
meillo@169 383 .Pn msh
meillo@212 384 together with the truly archaic code relics
meillo@58 385 .Pn vmh
meillo@58 386 and
meillo@169 387 .Pn wmh
meillo@62 388 saved more than 7\|000 lines of C code \(en
meillo@66 389 about 15\|% of the project's original source code amount.
meillo@100 390 Having less code \(en with equal readability, of course \(en
meillo@76 391 for the same functionality is an advantage.
meillo@63 392 Less code means less bugs and less maintenance work.
meillo@76 393 As
meillo@63 394 .Pn rcvtty
meillo@63 395 and
meillo@63 396 .Pn msgchk
meillo@87 397 are assumed to be rarely used and can be implemented in different ways,
meillo@87 398 why should one keep them?
meillo@217 399 Removing them streamlined mmh.
meillo@63 400 .Pn viamail 's
meillo@63 401 use case is now partly obsolete and partly covered by
meillo@63 402 .Pn forw ,
meillo@217 403 hence there is no reason to still maintain it.
meillo@63 404 .Pn conflict
meillo@76 405 is not related to the mail client, and
meillo@63 406 .Pn msh
meillo@63 407 conflicts with the basic concept of MH.
meillo@169 408 These two tools might still be useful, but they should not be part of mmh.
meillo@63 409 .P
meillo@220 410 .Id slocal
meillo@169 411 Finally, there is
meillo@217 412 .Pn slocal ,
meillo@217 413 which is an MDA and thus not directly MUA-related.
meillo@217 414 It should be removed from mmh because including it conflicts with
meillo@178 415 the idea that mmh is an MUA only.
meillo@87 416 However,
meillo@76 417 .Pn slocal
meillo@76 418 provides rule-based processing of messages, like filing them into
meillo@76 419 different folders, which is otherwise not available in mmh.
meillo@87 420 Although
meillo@76 421 .Pn slocal
meillo@169 422 neither pulls in dependencies, nor does it include a separate
meillo@154 423 technical area (cf. Sec.
meillo@154 424 .Cf mail-transfer-facilities ),
meillo@169 425 it still accounts for about 1\|000 lines of code that need to be maintained.
meillo@76 426 As
meillo@76 427 .Pn slocal
meillo@76 428 is almost self-standing, it should be split off into a separate project.
meillo@76 429 This would cut the strong connection between the MUA mmh and the MDA
meillo@76 430 .Pn slocal .
meillo@87 431 For anyone not using MH,
meillo@87 432 .Pn slocal
meillo@87 433 would become yet another independent MDA, like
meillo@87 434 .I procmail .
meillo@100 435 Then
meillo@87 436 .Pn slocal
meillo@217 437 could be installed without a complete MH system.
meillo@76 438 Likewise, mmh users could decide to use
meillo@76 439 .I procmail
meillo@217 440 without having a second, unused MDA, i.e.
meillo@87 441 .Pn slocal ,
meillo@76 442 installed.
meillo@100 443 That appears to be conceptionally the best solution.
meillo@76 444 Yet,
meillo@76 445 .Pn slocal
meillo@87 446 is not split off.
meillo@100 447 I defer the decision over
meillo@78 448 .Pn slocal
meillo@169 449 out of a need for deeper investigation.
meillo@217 450 In the meanwhile, it remains part of mmh
meillo@217 451 as its continued existence is not significant;
meillo@100 452 .Pn slocal
meillo@100 453 is unrelated to the rest of the project.
meillo@0 454
meillo@58 455
meillo@133 456
meillo@134 457 .H2 "Displaying Messages
meillo@155 458 .Id mhshow
meillo@131 459 .P
meillo@133 460 Since the very beginning, already in the first concept paper,
meillo@227 461 .[
meillo@227 462 original memo rand mh shapiro gaines
meillo@227 463 .]
meillo@58 464 .Pn show
meillo@62 465 had been MH's message display program.
meillo@58 466 .Pn show
meillo@76 467 mapped message numbers and sequences to files and invoked
meillo@58 468 .Pn mhl
meillo@89 469 to have the files formatted.
meillo@173 470 With MIME, this approach was not sufficient anymore.
meillo@217 471 MIME messages can consist of multiple parts.
meillo@217 472 Some parts, like binary attachments or text content in foreign charsets,
meillo@217 473 are not directly displayable.
meillo@58 474 .Pn show 's
meillo@76 475 understanding of messages and
meillo@58 476 .Pn mhl 's
meillo@173 477 display capabilities could not cope with the task any longer.
meillo@62 478 .P
meillo@88 479 Instead of extending these tools, additional tools were written from
meillo@217 480 scratch and were added to the MH tool chest.
meillo@88 481 Doing so is encouraged by the tool chest approach.
meillo@88 482 Modular design is a great advantage for extending a system,
meillo@88 483 as new tools can be added without interfering with existing ones.
meillo@62 484 First, the new MIME features were added in form of the single program
meillo@58 485 .Pn mhn .
meillo@58 486 The command
meillo@82 487 .Cl "mhn -show 42
meillo@217 488 had then shown the message number
meillo@217 489 .Fn 42 ,
meillo@217 490 interpreting MIME.
meillo@58 491 With the 1.0 release of nmh in February 1999, Richard Coleman finished
meillo@58 492 the split of
meillo@58 493 .Pn mhn
meillo@88 494 into a set of specialized tools, which together covered the
meillo@88 495 multiple aspects of MIME.
meillo@88 496 One of them was
meillo@69 497 .Pn mhshow ,
meillo@88 498 which replaced
meillo@88 499 .Cl "mhn -show" .
meillo@88 500 It was capable of displaying MIME messages appropriately.
meillo@62 501 .P
meillo@88 502 From then on, two message display tools were part of nmh,
meillo@76 503 .Pn show
meillo@76 504 and
meillo@76 505 .Pn mhshow .
meillo@88 506 To ease the life of users,
meillo@69 507 .Pn show
meillo@69 508 was extended to automatically hand the job over to
meillo@69 509 .Pn mhshow
meillo@69 510 if displaying the message would be beyond
meillo@69 511 .Pn show 's
meillo@69 512 abilities.
meillo@88 513 In consequence, the user would simply invoke
meillo@69 514 .Pn show
meillo@69 515 (possibly through
meillo@69 516 .Pn next
meillo@69 517 or
meillo@69 518 .Pn prev )
meillo@69 519 and get the message printed with either
meillo@69 520 .Pn show
meillo@69 521 or
meillo@69 522 .Pn mhshow ,
meillo@69 523 whatever was more appropriate.
meillo@69 524 .P
meillo@217 525 Having two similar tools for basically the same task is redundancy.
meillo@217 526 Usually, users do not distinguish between
meillo@88 527 .Pn show
meillo@88 528 and
meillo@88 529 .Pn mhshow
meillo@88 530 in their daily mail reading.
meillo@217 531 Having two separate display programs was therefore unnecessary
meillo@88 532 from a user's point of view.
meillo@88 533 Besides, the development of both programs needed to be in sync,
meillo@76 534 to ensure that the programs behaved in a similar way,
meillo@76 535 because they were used like a single tool.
meillo@76 536 Different behavior would have surprised the user.
meillo@69 537 .P
meillo@69 538 Today, non-MIME messages are rather seen to be a special case of
meillo@100 539 MIME messages, although it is the other way round.
meillo@69 540 As
meillo@69 541 .Pn mhshow
meillo@217 542 already had been able to display non-MIME messages, it appeared natural
meillo@69 543 to drop
meillo@69 544 .Pn show
meillo@69 545 in favor of using
meillo@69 546 .Pn mhshow
meillo@217 547 exclusively
meillo@217 548 .Ci 4c1efddfd499300c7e74263e57d8aa137e84c853 .
meillo@88 549 Removing
meillo@88 550 .Pn show
meillo@217 551 is no loss in function, because
meillo@88 552 .Pn mhshow
meillo@88 553 covers it completely.
meillo@217 554 Yet, the old behavior of
meillo@88 555 .Pn show
meillo@88 556 can still be emulated with the simple command line:
meillo@88 557 .VS
meillo@88 558 mhl `mhpath c`
meillo@88 559 VE
meillo@88 560 .P
meillo@76 561 For convenience,
meillo@76 562 .Pn mhshow
meillo@88 563 was renamed to
meillo@88 564 .Pn show
meillo@88 565 after
meillo@88 566 .Pn show
meillo@88 567 was gone.
meillo@88 568 It is clear that such a rename may confuse future developers when
meillo@88 569 trying to understand the history.
meillo@88 570 Nevertheless, I consider the convenience on the user's side,
meillo@217 571 to outweigh the inconvenience for understanding the evolution
meillo@217 572 of the tools.
meillo@69 573 .P
meillo@88 574 To prepare for the transition,
meillo@69 575 .Pn mhshow
meillo@69 576 was reworked to behave more like
meillo@69 577 .Pn show
meillo@217 578 first (cf. Sec.
meillo@217 579 .Cf mhshow ).
meillo@88 580 Once the tools behaved more alike, the replacing appeared to be
meillo@88 581 even more natural.
meillo@88 582 Today, mmh's new
meillo@69 583 .Pn show
meillo@217 584 has become the one single message display program once again,
meillo@159 585 with the difference
meillo@88 586 that today it handles MIME messages as well as non-MIME messages.
meillo@217 587 The outcomes of the transition are one program less to maintain,
meillo@88 588 no second display program for users to deal with,
meillo@88 589 and less system complexity.
meillo@69 590 .P
meillo@88 591 Still, removing the old
meillo@69 592 .Pn show
meillo@88 593 hurts in one regard: It had been such a simple program.
meillo@159 594 Its lean elegance is missing from the new
meillo@159 595 .Pn show ,
meillo@225 596 but there is no alternative;
meillo@159 597 supporting MIME demands higher essential complexity.
meillo@58 598
meillo@131 599
meillo@133 600
meillo@100 601 .H2 "Configure Options
meillo@58 602 .P
meillo@76 603 Customization is a double-edged sword.
meillo@76 604 It allows better suiting setups, but not for free.
meillo@76 605 There is the cost of code complexity to be able to customize.
meillo@76 606 There is the cost of less tested setups, because there are
meillo@171 607 more possible setups and especially corner cases.
meillo@226 608 Steve Johnson confirms:
meillo@226 609 .[ [
meillo@226 610 eric raymond the art of unix programming
meillo@226 611 .], p. 233]
meillo@226 612 .QS
meillo@226 613 Unless it is done very carefully, the addition of an on/off
meillo@226 614 configuration option can lead to a need to double the amount of
meillo@226 615 testing. Since in practice one never does double the amount of testing,
meillo@226 616 the practical effect is to reduce the amount of testing that any
meillo@226 617 given configuration receives. Ten options leads to 1024 times as much
meillo@226 618 testing, and pretty soon you are talking real reliability problems.
meillo@226 619 .QE
meillo@226 620 .LP
meillo@159 621 Additionally, there is the cost of choice itself.
meillo@76 622 The code complexity directly affects the developers.
meillo@173 623 Less tested code affects both users and developers.
meillo@217 624 The problem of choice affects the users, for once by having to choose
meillo@159 625 but also by more complex interfaces that require more documentation.
meillo@159 626 Whenever options add few advantages but increase the complexity of the
meillo@159 627 system, they should be considered for removal.
meillo@72 628 I have reduced the number of project-specific configure options from
meillo@217 629 15 to 3.
meillo@74 630
meillo@76 631 .U3 "Mail Transfer Facilities
meillo@74 632 .P
meillo@227 633 With the removal of the mail transfer facilities (cf. Sec.
meillo@227 634 .Cf mail-transfer-facilities )
meillo@227 635 five configure options vanished:
meillo@85 636 .P
meillo@85 637 The switches
meillo@85 638 .Sw --with-tls
meillo@85 639 and
meillo@85 640 .Sw --with-cyrus-sasl
meillo@89 641 had activated the support for transfer encryption and authentication.
meillo@217 642 They are not needed anymore.
meillo@85 643 .Ci fecd5d34f65597a4dfa16aeabea7d74b191532c3
meillo@85 644 .Ci 156d35f6425bea4c1ed3c4c79783dc613379c65b
meillo@85 645 .P
meillo@85 646 The configure switch
meillo@85 647 .Sw --enable-pop
meillo@227 648 had activated the message retrieval facility.
meillo@217 649 Whereas the code area that had been conditionally compiled in
meillo@217 650 for TLS and SASL support was small,
meillo@217 651 the conditionally compiled code area for POP support was much larger.
meillo@217 652 The code base had only changed slightly on toggling TLS or SASL
meillo@217 653 support but it had changed much on toggling POP support.
meillo@85 654 The changes in the code base could hardly be overviewed.
meillo@159 655 By having POP support togglable, a second code base had been created,
meillo@85 656 one that needed to be tested.
meillo@85 657 This situation is basically similar for the conditional TLS and SASL
meillo@85 658 code, but there the changes are minor and can yet be overviewed.
meillo@85 659 Still, conditional compilation of a code base creates variations
meillo@85 660 of the original program.
meillo@85 661 More variations require more testing and maintenance work.
meillo@85 662 .P
meillo@227 663 Two other options had only specified default configuration values:
meillo@100 664 .Sw --with-mts
meillo@217 665 defined the default transport service
meillo@217 666 .Ci f6aa95b724fd8c791164abe7ee5468bf5c34f226 .
meillo@85 667 With
meillo@100 668 .Sw --with-smtpservers
meillo@227 669 default SMTP servers could be set
meillo@217 670 .Ci 128545e06224233b7e91fc4c83f8830252fe16c9 .
meillo@164 671 Both of them became irrelevant when the SMTP transport service was removed.
meillo@164 672 In mmh, all messages are handed over to
meillo@164 673 .Pn sendmail
meillo@164 674 for transportation.
meillo@164 675
meillo@72 676
meillo@74 677 .U3 "Backup Prefix
meillo@74 678 .P
meillo@76 679 The backup prefix is the string that was prepended to message
meillo@76 680 filenames to tag them as deleted.
meillo@173 681 By default it had been the comma character (`\fL,\fP').
meillo@78 682 In July 2000, Kimmo Suominen introduced
meillo@78 683 the configure option
meillo@78 684 .Sw --with-hash-backup
meillo@173 685 to change the default to the hash character `\f(CW#\fP'.
meillo@217 686 This choice was probably personal preference, but,
meillo@217 687 being related or not, words that start with the hash character
meillo@78 688 introduce a comment in the Unix shell.
meillo@72 689 Thus, the command line
meillo@72 690 .Cl "rm #13 #15
meillo@72 691 calls
meillo@72 692 .Pn rm
meillo@217 693 without arguments because the first hash character starts a comment
meillo@72 694 that reaches until the end of the line.
meillo@72 695 To delete the backup files,
meillo@72 696 .Cl "rm ./#13 ./#15"
meillo@72 697 needs to be used.
meillo@217 698 Thus, using the hash as backup prefix may be seen as a precaution
meillo@217 699 against backup loss.
meillo@78 700 .P
meillo@159 701 First, I removed the configure option but added the profile entry
meillo@217 702 .Pe Backup-Prefix ,
meillo@217 703 which allowed to specify an arbitrary string as backup prefix
meillo@217 704 .Ci 6c40d481d661d532dd527eaf34cebb6d3f8ed086 .
meillo@76 705 This change did not remove the choice but moved it to a location where
meillo@217 706 it suited better, in my eyes.
meillo@76 707 .P
meillo@217 708 Eventually however, the new trash folder concept
meillo@154 709 (cf. Sec.
meillo@154 710 .Cf trash-folder )
meillo@164 711 removed the need for the backup prefix completely.
meillo@78 712 .Ci 8edc5aaf86f9f77124664f6801bc6c6cdf258173
meillo@133 713 .Ci ca0b3e830b86700d9e5e31b1784de2bdcaf58fc5
meillo@133 714
meillo@76 715
meillo@76 716 .U3 "Editor and Pager
meillo@227 717 .Id editor-pager
meillo@74 718 .P
meillo@74 719 The two configure options
meillo@74 720 .CW --with-editor=EDITOR
meillo@74 721 .CW --with-pager=PAGER
meillo@74 722 were used to specify the default editor and pager at configure time.
meillo@109 723 Doing so at configure time made sense in the eighties,
meillo@76 724 when the set of available editors and pagers varied much across
meillo@76 725 different systems.
meillo@89 726 Today, the situation is more homogeneous.
meillo@74 727 The programs
meillo@74 728 .Pn vi
meillo@74 729 and
meillo@74 730 .Pn more
meillo@76 731 can be expected to be available on every Unix system,
meillo@74 732 as they are specified by POSIX since two decades.
meillo@74 733 (The specifications for
meillo@74 734 .Pn vi
meillo@74 735 and
meillo@74 736 .Pn more
meillo@74 737 appeared in
meillo@74 738 .[
meillo@74 739 posix 1987
meillo@74 740 .]
meillo@74 741 and,
meillo@74 742 .[
meillo@74 743 posix 1992
meillo@74 744 .]
meillo@74 745 respectively.)
meillo@217 746 As a first step, these two tools were hard-coded as defaults
meillo@217 747 .Ci 5d43a99db70c12a673028c7758c20cbe3e13ef5f .
meillo@74 748 Not changed were the
meillo@74 749 .Pe editor
meillo@74 750 and
meillo@74 751 .Pe moreproc
meillo@76 752 profile entries, which allowed the user to override the system defaults.
meillo@217 753 Later, the concept was reworked again to respect the standard
meillo@217 754 environment variables
meillo@74 755 .Ev VISUAL
meillo@74 756 and
meillo@74 757 .Ev PAGER
meillo@76 758 if they are set.
meillo@74 759 Today, mmh determines the editor to use in the following order,
meillo@217 760 taking the first available and non-empty item
meillo@217 761 .Ci f85f4b7ae62e3d05a945dcd46ead51f0a2a89a9b :
meillo@171 762 .LI 1
meillo@74 763 Environment variable
meillo@74 764 .Ev MMHEDITOR
meillo@171 765 .LI 2
meillo@74 766 Profile entry
meillo@74 767 .Pe Editor
meillo@171 768 .LI 3
meillo@74 769 Environment variable
meillo@74 770 .Ev VISUAL
meillo@171 771 .LI 4
meillo@74 772 Environment variable
meillo@74 773 .Ev EDITOR
meillo@171 774 .LI 5
meillo@74 775 Command
meillo@74 776 .Pn vi .
meillo@171 777 .LP
meillo@217 778 The pager to use is determined in a similar order
meillo@217 779 .Ci 0c4214ea2aec6497d0d67b436bbee9bc1d225f1e :
meillo@171 780 .LI 1
meillo@74 781 Environment variable
meillo@74 782 .Ev MMHPAGER
meillo@171 783 .LI 2
meillo@74 784 Profile entry
meillo@74 785 .Pe Pager
meillo@74 786 (replaces
meillo@74 787 .Pe moreproc )
meillo@171 788 .LI 3
meillo@74 789 Environment variable
meillo@74 790 .Ev PAGER
meillo@171 791 .LI 4
meillo@74 792 Command
meillo@74 793 .Pn more .
meillo@171 794 .LP
meillo@76 795 By respecting the
meillo@74 796 .Ev VISUAL /\c
meillo@74 797 .Ev EDITOR
meillo@74 798 and
meillo@74 799 .Ev PAGER
meillo@76 800 environment variables,
meillo@217 801 the new behavior complies with the common style on Unix systems.
meillo@217 802 It is more uniform and clearer for users.
meillo@72 803
meillo@72 804
meillo@76 805 .U3 "ndbm
meillo@72 806 .P
meillo@74 807 .Pn slocal
meillo@217 808 used to depend on the database library
meillo@217 809 .I ndbm .
meillo@217 810 The database is used to store the
meillo@217 811 .Hd Message-ID
meillo@217 812 header field values of all messages delivered.
meillo@217 813 This enabled
meillo@74 814 .Pn slocal
meillo@74 815 to suppress delivering the same message to the same user twice.
meillo@217 816 This features was enabled by the
meillo@74 817 .Sw -suppressdup
meillo@217 818 switch.
meillo@74 819 .P
meillo@217 820 As a variety of versions of the database library exist,
meillo@78 821 .[
meillo@78 822 wolter unix incompat notes dbm
meillo@78 823 .]
meillo@217 824 complicated autoconf code was needed to detect them correctly.
meillo@181 825 Furthermore, the configure switches
meillo@74 826 .Sw --with-ndbm=ARG
meillo@74 827 and
meillo@74 828 .Sw --with-ndbmheader=ARG
meillo@74 829 were added to help with difficult setups that would
meillo@217 830 not be detected automatically or not correctly.
meillo@74 831 .P
meillo@74 832 By removing the suppress duplicates feature of
meillo@74 833 .Pn slocal ,
meillo@74 834 the dependency on
meillo@74 835 .I ndbm
meillo@217 836 vanished and 120 lines of complex autoconf code could be saved
meillo@217 837 .Ci ecd6d6a20cb7a1507e3a20d6c4cb3a1cf14c6bbf .
meillo@217 838 The change removed functionality but that is considered minor to the
meillo@217 839 improvement of dropping the dependency and the complex autoconf code.
meillo@225 840
meillo@72 841
meillo@217 842 .U3 "MH-E Support
meillo@72 843 .P
meillo@74 844 The configure option
meillo@74 845 .Sw --disable-mhe
meillo@217 846 was removed when the MH-E support was reworked.
meillo@217 847 MH-E is the Emacs front-end to MH.
meillo@76 848 It requires MH to provide minor additional functions.
meillo@76 849 The
meillo@76 850 .Sw --disable-mhe
meillo@217 851 configure option had switched off these extensions.
meillo@217 852 After removing the support for old versions of MH-E,
meillo@74 853 only the
meillo@74 854 .Sw -build
meillo@76 855 switches of
meillo@74 856 .Pn forw
meillo@74 857 and
meillo@74 858 .Pn repl
meillo@217 859 are left to be MH-E extensions.
meillo@76 860 They are now always built in because they add little code and complexity.
meillo@76 861 In consequence, the
meillo@74 862 .Sw --disable-mhe
meillo@76 863 configure option was removed
meillo@217 864 .Ci a7ce7b4a580d77b6c2c4d980812beb589aa4c643 .
meillo@217 865 Dropping the option also removed a variant of the code base
meillo@217 866 that would have needed to be tested.
meillo@217 867 This change was undertaken in January 2012 in nmh and
meillo@217 868 thereafter merged into mmh.
meillo@217 869
meillo@72 870
meillo@74 871 .U3 "Masquerading
meillo@72 872 .P
meillo@74 873 The configure option
meillo@74 874 .Sw --enable-masquerade
meillo@76 875 could take up to three arguments:
meillo@217 876 .Ar draft_from ,
meillo@217 877 .Ar mmailid ,
meillo@217 878 and
meillo@217 879 .Ar username_extension .
meillo@74 880 They activated different types of address masquerading.
meillo@74 881 All of them were implemented in the SMTP-speaking
meillo@74 882 .Pn post
meillo@217 883 command.
meillo@76 884 Address masquerading is an MTA's task and mmh does not cover
meillo@76 885 this field anymore.
meillo@76 886 Hence, true masquerading needs to be implemented in the external MTA.
meillo@74 887 .P
meillo@74 888 The
meillo@74 889 .I mmailid
meillo@74 890 masquerading type is the oldest one of the three and the only one
meillo@74 891 available in the original MH.
meillo@74 892 It provided a
meillo@74 893 .I username
meillo@74 894 to
meillo@74 895 .I fakeusername
meillo@217 896 mapping, based on the
meillo@217 897 .Fn passwd 's
meillo@217 898 GECOS field.
meillo@217 899 Nmh's man page
meillo@181 900 .Mp mh-tailor (5)
meillo@74 901 described the use case as being the following:
meillo@98 902 .QS
meillo@74 903 This is useful if you want the messages you send to always
meillo@74 904 appear to come from the name of an MTA alias rather than your
meillo@74 905 actual account name. For instance, many organizations set up
meillo@74 906 `First.Last' sendmail aliases for all users. If this is
meillo@74 907 the case, the GECOS field for each user should look like:
meillo@74 908 ``First [Middle] Last <First.Last>''
meillo@98 909 .QE
meillo@74 910 .P
meillo@74 911 As mmh sends outgoing mail via the local MTA only,
meillo@76 912 the best location to do such global rewrites is there.
meillo@74 913 Besides, the MTA is conceptionally the right location because it
meillo@74 914 does the reverse mapping for incoming mail (aliasing), too.
meillo@181 915 Furthermore, masquerading set up there is readily available for all
meillo@74 916 mail software on the system.
meillo@76 917 Hence, mmailid masquerading was removed.
meillo@74 918 .Ci 0836c8000ccb34b59410ef1c15b1b7feac70ce5f
meillo@74 919 .P
meillo@74 920 The
meillo@74 921 .I username_extension
meillo@76 922 masquerading type did not replace the username but would append a suffix,
meillo@76 923 specified by the
meillo@74 924 .Ev USERNAME_EXTENSION
meillo@76 925 environment variable, to it.
meillo@76 926 This provided support for the
meillo@74 927 .I user-extension
meillo@217 928 feature of qmail
meillo@217 929 .[ [
meillo@217 930 sill qmail handbook
meillo@217 931 .], p. 141]
meillo@217 932 and the similar
meillo@74 933 .I "plussed user
meillo@217 934 processing of Sendmail.
meillo@217 935 .[ [
meillo@217 936 sendmail costales
meillo@217 937 .], p. 476]
meillo@217 938 The decision to remove this username_extension masquerading
meillo@217 939 was motivated by the fact that
meillo@74 940 .Pn spost
meillo@217 941 had not supported it yet.
meillo@217 942 Username extensions can be used in mmh, but less convenient.
meillo@225 943 .\" XXX In the format file: %(getenv USERNAME_EXTENSION)
meillo@217 944 .Ci 2abae0bfd0ad5bf898461e50aa4b466d641f23d9
meillo@74 945 .P
meillo@74 946 The
meillo@74 947 .I draft_from
meillo@74 948 masquerading type instructed
meillo@74 949 .Pn post
meillo@84 950 to use the value of the
meillo@84 951 .Hd From
meillo@84 952 header field as SMTP envelope sender.
meillo@76 953 Sender addresses could be replaced completely.
meillo@76 954 Mmh offers a kind of masquerading similar in effect, but
meillo@74 955 with technical differences.
meillo@76 956 As mmh does not transfer messages itself, the local MTA has final control
meillo@217 957 over the sender's address.
meillo@217 958 Any masquerading mmh introduces may be reverted by the MTA.
meillo@76 959 In times of pedantic spam checking, an MTA will take care to use
meillo@76 960 sensible envelope sender addresses to keep its own reputation up.
meillo@84 961 Nonetheless, the MUA can set the
meillo@84 962 .Hd From
meillo@217 963 header field and thereby propose a sender address to the MTA.
meillo@74 964 The MTA may then decide to take that one or generate the canonical sender
meillo@74 965 address for use as envelope sender address.
meillo@217 966 .Ci b14ea6073f77b4359aaf3fddd0e105989db9
meillo@74 967 .P
meillo@74 968 In mmh, the MTA will always extract the recipient and sender from the
meillo@84 969 message header (\c
meillo@74 970 .Pn sendmail 's
meillo@74 971 .Sw -t
meillo@74 972 switch).
meillo@84 973 The
meillo@84 974 .Hd From
meillo@84 975 header field of the draft may be set arbitrary by the user.
meillo@74 976 If it is missing, the canonical sender address will be generated by the MTA.
meillo@74 977
meillo@74 978 .U3 "Remaining Options
meillo@74 979 .P
meillo@74 980 Two configure options remain in mmh.
meillo@74 981 One is the locking method to use:
meillo@74 982 .Sw --with-locking=[dot|fcntl|flock|lockf] .
meillo@217 983 The idea of removing all methods except the portable
meillo@217 984 .I "dot locking
meillo@76 985 and having that one as the default is appealing, but this change
meillo@76 986 requires deeper technical investigation into the topic.
meillo@76 987 The other option,
meillo@74 988 .Sw --enable-debug ,
meillo@217 989 compiles the programs with debugging symbols.
meillo@74 990 This option is likely to stay.
meillo@72 991
meillo@72 992
meillo@58 993
meillo@63 994
meillo@100 995 .H2 "Command Line Switches
meillo@58 996 .P
meillo@227 997 The command line switches of MH tools are similar in style to the
meillo@227 998 switches in the X Window System.
meillo@227 999 They consist of a single dash (`\fL-\fP') followed by a word.
meillo@217 1000 For example
meillo@217 1001 .Cl -truncate .
meillo@212 1002 To ease typing, the word can be abbreviated, given the remaining
meillo@217 1003 prefix is unambiguous.
meillo@212 1004 If no other switch starts with the letter `t', then any of
meillo@212 1005 .Cl "-truncate" ,
meillo@212 1006 .Cl "-trunc" ,
meillo@212 1007 .Cl "-tr" ,
meillo@212 1008 and
meillo@212 1009 .Cl "-t
meillo@212 1010 is equal.
meillo@212 1011 As a result, switches can neither be grouped (as in
meillo@212 1012 .Cl "ls -ltr" )
meillo@212 1013 nor can switch arguments be appended directly to the switch (as in
meillo@212 1014 .Cl "sendmail -q30m" ).
meillo@212 1015 Many switches have negating counter-parts, which start with `no'.
meillo@212 1016 For example
meillo@212 1017 .Cl "-notruncate
meillo@212 1018 inverts the
meillo@212 1019 .Cl "-truncate
meillo@212 1020 switch.
meillo@212 1021 They exist to override the effect of default switches in the profile.
meillo@93 1022 Every program in mmh has two generic switches:
meillo@93 1023 .Sw -help ,
meillo@93 1024 to print a short message on how to use the program, and
meillo@159 1025 .Sw -Version
meillo@164 1026 (with capital `V'), to tell what version of mmh the program belongs to.
meillo@93 1027 .P
meillo@93 1028 Switches change the behavior of programs.
meillo@93 1029 Programs that do one thing in one way require no switches.
meillo@93 1030 In most cases, doing something in exactly one way is too limiting.
meillo@217 1031 If one task should be accomplished in various ways,
meillo@217 1032 switches are a good approach to alter the behavior of a program.
meillo@93 1033 Changing the behavior of programs provides flexibility and customization
meillo@217 1034 to users, but at the same time it complicates the code,
meillo@217 1035 the documentation, and the usage of the program.
meillo@93 1036 Therefore, the number of switches should be kept small.
meillo@217 1037 A small set of well-chosen switches is best.
meillo@217 1038 Usually, the number of switches increases over time.
meillo@93 1039 Already in 1985, Rose and Romine have identified this as a major
meillo@93 1040 problem of MH:
meillo@93 1041 .[ [
meillo@93 1042 rose romine real work
meillo@93 1043 .], p. 12]
meillo@98 1044 .QS
meillo@93 1045 A complaint often heard about systems which undergo substantial development
meillo@93 1046 by many people over a number of years, is that more and more options are
meillo@93 1047 introduced which add little to the functionality but greatly increase the
meillo@93 1048 amount of information a user needs to know in order to get useful work done.
meillo@93 1049 This is usually referred to as creeping featurism.
meillo@93 1050 .QP
meillo@93 1051 Unfortunately MH, having undergone six years of off-and-on development by
meillo@93 1052 ten or so well-meaning programmers (the present authors included),
meillo@93 1053 suffers mightily from this.
meillo@98 1054 .QE
meillo@93 1055 .P
meillo@217 1056 Being reluctant to adding new switches (or \fIoptions\fP,
meillo@217 1057 as Rose and Romine call them) is one part of a counter-action,
meillo@97 1058 the other part is removing hardly used switches.
meillo@217 1059 Nmh's tools have lots of switches already implemented.
meillo@217 1060 Hence, cleaning up by removing some of them was the more important part
meillo@97 1061 of the counter-action.
meillo@93 1062 Removing existing functionality is always difficult because it
meillo@93 1063 breaks programs that use these functions.
meillo@93 1064 Also, for every obsolete feature, there'll always be someone who still
meillo@93 1065 uses it and thus opposes its removal.
meillo@93 1066 This puts the developer into the position,
meillo@93 1067 where sensible improvements to style are regarded as destructive acts.
meillo@97 1068 Yet, living with the featurism is far worse, in my eyes, because
meillo@97 1069 future needs will demand adding further features,
meillo@93 1070 worsening the situation more and more.
meillo@93 1071 Rose and Romine added in a footnote,
meillo@93 1072 ``[...]
meillo@93 1073 .Pn send
meillo@217 1074 will no doubt acquire an endless number of switches in the years to come''
meillo@217 1075 .[ [
meillo@217 1076 rose romine real work
meillo@217 1077 .], p. 12].
meillo@97 1078 Although clearly humorous, the comment points to the nature of the problem.
meillo@97 1079 Refusing to add any new switches would encounter the problem at its root,
meillo@97 1080 but this is not practical.
meillo@97 1081 New needs will require new switches and it would be unwise to block
meillo@97 1082 them strictly.
meillo@97 1083 Nevertheless, removing obsolete switches still is an effective approach
meillo@97 1084 to deal with the problem.
meillo@97 1085 Working on an experimental branch without an established user base,
meillo@97 1086 eased my work because I did not offend users when I removed existing
meillo@110 1087 functions.
meillo@93 1088 .P
meillo@93 1089 Rose and Romine counted 24 visible and 9 more hidden switches for
meillo@93 1090 .Pn send .
meillo@97 1091 In nmh, they increased up to 32 visible and 12 hidden ones.
meillo@182 1092 At the time of writing, no more than 4 visible switches and 1 hidden switch
meillo@97 1093 have remained in mmh's
meillo@97 1094 .Pn send .
meillo@217 1095 These numbers include the two generic switches,
meillo@182 1096 .Sw -help
meillo@182 1097 and
meillo@183 1098 .Sw -Version .
meillo@217 1099 .P
meillo@183 1100 Hidden switches are ones not documented.
meillo@183 1101 In mmh, 12 tools have hidden switches.
meillo@183 1102 9 of them are
meillo@183 1103 .Sw -debug
meillo@183 1104 switches, the other 6 provide special interfaces for internal use.
meillo@93 1105 .P
meillo@217 1106 The following figure displays the number of switches for each of the tools
meillo@159 1107 that is available in both nmh and mmh.
meillo@100 1108 The tools are sorted by the number of switches they had in nmh.
meillo@217 1109 Both visible and hidden switches were counted,
meillo@97 1110 but not the generic help and version switches.
meillo@93 1111 Whereas in the beginning of the project, the average tool had 11 switches,
meillo@93 1112 now it has no more than 5 \(en only half as many.
meillo@93 1113 If the `no' switches and similar inverse variant are folded onto
meillo@100 1114 their counter-parts, the average tool had 8 switches in pre-mmh times and
meillo@100 1115 has 4 now.
meillo@93 1116 The total number of functional switches in mmh dropped from 465
meillo@182 1117 to 233.
meillo@58 1118
meillo@93 1119 .KS
meillo@93 1120 .in 1c
meillo@93 1121 .so input/switches.grap
meillo@93 1122 .KE
meillo@58 1123
meillo@93 1124 .P
meillo@93 1125 A part of the switches vanished after functions were removed.
meillo@93 1126 This was the case for network mail transfer, for instance.
meillo@97 1127 Sometimes, however, the work flow was the other way:
meillo@97 1128 I looked through the
meillo@97 1129 .Mp mh-chart (7)
meillo@97 1130 man page to identify the tools with apparently too many switches.
meillo@217 1131 Then I considered the benefit of each switch by examining
meillo@217 1132 the tool's man page and source code, aided by literature research
meillo@217 1133 and testing.
meillo@97 1134
meillo@58 1135
meillo@93 1136 .U3 "Draft Folder Facility
meillo@93 1137 .P
meillo@100 1138 A change early in the project was the complete transition from
meillo@217 1139 the single draft message to the draft folder facility
meillo@227 1140 .Ci 337338b404931f06f0db2119c9e145e8ca5a9860
meillo@227 1141 (cf. Sec.
meillo@227 1142 .Cf draft-folder ).
meillo@109 1143 The draft folder facility was introduced in the mid-eighties, when
meillo@100 1144 Rose and Romine called it a ``relatively new feature''.
meillo@93 1145 .[
meillo@93 1146 rose romine real work
meillo@93 1147 .]
meillo@217 1148 Since then, the facility was included, inactive by default.
meillo@217 1149 By making it permanently active and by related rework of the tools, the
meillo@93 1150 .Sw -[no]draftfolder ,
meillo@93 1151 and
meillo@93 1152 .Sw -draftmessage
meillo@217 1153 switches could be removed from
meillo@93 1154 .Pn comp ,
meillo@93 1155 .Pn repl ,
meillo@93 1156 .Pn forw ,
meillo@93 1157 .Pn dist ,
meillo@93 1158 .Pn whatnow ,
meillo@93 1159 and
meillo@217 1160 .Pn send
meillo@217 1161 .Ci 337338b404931f06f0db2119c9e145e8ca5a9860 .
meillo@217 1162 The only flexibility lost with this change is having multiple
meillo@97 1163 draft folders within one profile.
meillo@97 1164 I consider this a theoretical problem only.
meillo@159 1165 At the same time, the
meillo@93 1166 .Sw -draft
meillo@93 1167 switch of
meillo@93 1168 .Pn anno ,
meillo@93 1169 .Pn refile ,
meillo@93 1170 and
meillo@93 1171 .Pn send
meillo@93 1172 was removed.
meillo@159 1173 The special treatment of \fIthe\fP draft message became irrelevant after
meillo@217 1174 the rework of the draft system
meillo@159 1175 (cf. Sec.
meillo@217 1176 .Cf draft-folder ).
meillo@93 1177
meillo@95 1178
meillo@102 1179 .U3 "In Place Editing
meillo@93 1180 .P
meillo@93 1181 .Pn anno
meillo@93 1182 had the switches
meillo@93 1183 .Sw -[no]inplace
meillo@100 1184 to either annotate the message in place and thus preserve hard links,
meillo@217 1185 or annotate a copy to replace the original message.
meillo@217 1186 The latter approach broke hard links.
meillo@97 1187 Following the assumption that linked messages should truly be the
meillo@217 1188 same message and annotating it should not break the link, the
meillo@93 1189 .Sw -[no]inplace
meillo@93 1190 switches were removed and the previous default
meillo@93 1191 .Sw -inplace
meillo@217 1192 was made the definitive behavior
meillo@217 1193 .Ci c8195849d2e366c569271abb0f5f60f4ebf0b4d0 .
meillo@93 1194 The
meillo@93 1195 .Sw -[no]inplace
meillo@93 1196 switches of
meillo@93 1197 .Pn repl ,
meillo@93 1198 .Pn forw ,
meillo@93 1199 and
meillo@93 1200 .Pn dist
meillo@217 1201 could be removed, as well, as they were simply passed through to
meillo@93 1202 .Pn anno .
meillo@93 1203 .P
meillo@93 1204 .Pn burst
meillo@93 1205 also had
meillo@93 1206 .Sw -[no]inplace
meillo@217 1207 switches, but with a different meaning.
meillo@95 1208 With
meillo@95 1209 .Sw -inplace ,
meillo@95 1210 the digest had been replaced by the table of contents (i.e. the
meillo@110 1211 introduction text) and the burst messages were placed right
meillo@95 1212 after this message, renumbering all following messages.
meillo@95 1213 Also, any trailing text of the digest was lost, though,
meillo@95 1214 in practice, it usually consists of an end-of-digest marker only.
meillo@217 1215 Nonetheless, this behavior appeared less elegant than the
meillo@95 1216 .Sw -noinplace
meillo@95 1217 behavior, which already had been the default.
meillo@95 1218 Nmh's
meillo@95 1219 .Mp burst (1)
meillo@95 1220 man page reads:
meillo@98 1221 .QS
meillo@164 1222 If
meillo@164 1223 .Sw -noinplace
meillo@164 1224 is given, each digest is preserved, no table
meillo@93 1225 of contents is produced, and the messages contained within
meillo@93 1226 the digest are placed at the end of the folder. Other messages
meillo@93 1227 are not tampered with in any way.
meillo@98 1228 .QE
meillo@95 1229 .LP
meillo@93 1230 The decision to drop the
meillo@93 1231 .Sw -inplace
meillo@95 1232 behavior was supported by the code complexity and the possible data loss
meillo@95 1233 it caused.
meillo@93 1234 .Sw -noinplace
meillo@95 1235 was chosen to be the definitive behavior.
meillo@97 1236 .Ci 68a686adeb39223a5e1ad35e4a24890ec053679d
meillo@93 1237
meillo@95 1238
meillo@95 1239 .U3 "Forms and Format Strings
meillo@93 1240 .P
meillo@95 1241 Historically, the tools that had
meillo@95 1242 .Sw -form
meillo@95 1243 switches to supply a form file had
meillo@95 1244 .Sw -format
meillo@95 1245 switches as well to supply the contents of a form file as a string
meillo@95 1246 on the command line directly.
meillo@95 1247 In consequence, the following two lines equaled:
meillo@95 1248 .VS
meillo@95 1249 scan -form scan.mailx
meillo@217 1250 scan -format "`cat /path/to/scan.mailx`"
meillo@95 1251 VE
meillo@95 1252 The
meillo@95 1253 .Sw -format
meillo@95 1254 switches were dropped in favor for extending the
meillo@95 1255 .Sw -form
meillo@217 1256 switches
meillo@217 1257 .Ci f51956be123db66b00138f80464d06f030dbb88d .
meillo@217 1258 If their argument starts with an equal sign (`\fL=\fP'),
meillo@95 1259 then the rest of the argument is taken as a format string,
meillo@95 1260 otherwise the arguments is treated as the name of a format file.
meillo@95 1261 Thus, now the following two lines equal:
meillo@95 1262 .VS
meillo@95 1263 scan -form scan.mailx
meillo@217 1264 scan -form "=`cat /path/to/scan.mailx`"
meillo@95 1265 VE
meillo@95 1266 This rework removed the prefix collision between
meillo@95 1267 .Sw -form
meillo@95 1268 and
meillo@95 1269 .Sw -format .
meillo@217 1270 Typing `\fL-fo\fP' is sufficient to specify form file or format string.
meillo@95 1271 .P
meillo@95 1272 The different meaning of
meillo@95 1273 .Sw -format
meillo@95 1274 for
meillo@217 1275 .Pn forw
meillo@217 1276 and
meillo@95 1277 .Pn repl
meillo@95 1278 was removed in mmh.
meillo@95 1279 .Pn forw
meillo@95 1280 was completely switched to MIME-type forwarding, thus removing the
meillo@217 1281 .Sw -[no]format
meillo@217 1282 .Ci 6e271608b7b9c23771523f88d23a4d3593010cf1 .
meillo@95 1283 For
meillo@95 1284 .Pn repl ,
meillo@95 1285 the
meillo@95 1286 .Sw -[no]format
meillo@95 1287 switches were reworked to
meillo@95 1288 .Sw -[no]filter
meillo@217 1289 switches
meillo@217 1290 .Ci 67411b1f95d6ec987b4c732459e1ba8a8ac192c6 .
meillo@95 1291 The
meillo@95 1292 .Sw -format
meillo@95 1293 switches of
meillo@95 1294 .Pn send
meillo@95 1295 and
meillo@95 1296 .Pn post ,
meillo@217 1297 which had a third meaning, were removed likewise
meillo@217 1298 .Ci f3cb7cde0e6f10451b6848678d95860d512224b9 .
meillo@95 1299 Eventually, the ambiguity of the
meillo@95 1300 .Sw -format
meillo@217 1301 switches is resolved by not having such switches anymore in mmh.
meillo@95 1302
meillo@95 1303
meillo@95 1304 .U3 "MIME Tools
meillo@95 1305 .P
meillo@217 1306 The MIME tools, which once were part of
meillo@100 1307 .Pn mhn
meillo@164 1308 (whatever that stood for),
meillo@95 1309 had several switches that added little practical value to the programs.
meillo@95 1310 The
meillo@95 1311 .Sw -[no]realsize
meillo@95 1312 switches of
meillo@95 1313 .Pn mhbuild
meillo@95 1314 and
meillo@95 1315 .Pn mhlist
meillo@217 1316 were removed
meillo@217 1317 .Ci 8d8f1c3abc586c005c904e52c4adbfe694d2201c .
meillo@217 1318 Real size calculations are done always now because nmh's
meillo@159 1319 .Mp mhbuild (1)
meillo@217 1320 man page states that
meillo@217 1321 ``This provides an accurate count at the expense of a small delay''
meillo@228 1322 with the small delay not being noticeable on modern systems.
meillo@95 1323 .P
meillo@95 1324 The
meillo@95 1325 .Sw -[no]check
meillo@95 1326 switches were removed together with the support for
meillo@95 1327 .Hd Content-MD5
meillo@217 1328 header fields [RFC\|1864]
meillo@154 1329 (cf. Sec.
meillo@154 1330 .Cf content-md5 )
meillo@217 1331 .Ci 31dc797eb5178970d68962ca8939da3fd9a8efda .
meillo@95 1332 .P
meillo@95 1333 The
meillo@95 1334 .Sw -[no]ebcdicsafe
meillo@95 1335 and
meillo@95 1336 .Sw -[no]rfc934mode
meillo@95 1337 switches of
meillo@95 1338 .Pn mhbuild
meillo@217 1339 were removed because they are considered obsolete
meillo@97 1340 .Ci 01a3480928da485b4d6109d36d751dfa71799d58
meillo@217 1341 .Ci 3363e2624dce0eb8164cf8b3f1ab385c8ff72e88 .
meillo@95 1342 .P
meillo@95 1343 Content caching of external MIME parts, activated with the
meillo@95 1344 .Sw -rcache
meillo@95 1345 and
meillo@95 1346 .Sw -wcache
meillo@217 1347 switches was completely removed
meillo@217 1348 .Ci d1fefd9f614e4dc3cda16da6c69133c1b2005269 .
meillo@97 1349 External MIME parts are rare today, having a caching facility
meillo@159 1350 for them appears to be unnecessary.
meillo@95 1351 .P
meillo@95 1352 In pre-MIME times,
meillo@95 1353 .Pn mhl
meillo@95 1354 had covered many tasks that are part of MIME handling today.
meillo@95 1355 Therefore,
meillo@95 1356 .Pn mhl
meillo@95 1357 could be simplified to a large extend, reducing the number of its
meillo@217 1358 switches from 21 to 6
meillo@97 1359 .Ci 350ad6d3542a07639213cf2a4fe524e829c1e7b6
meillo@217 1360 .Ci 0e46503be3c855bddaeae3843e1b659279c35d70 .
meillo@95 1361
meillo@95 1362
meillo@95 1363
meillo@95 1364
meillo@95 1365 .U3 "Header Printing
meillo@95 1366 .P
meillo@95 1367 .Pn folder 's
meillo@95 1368 data output is self-explaining enough that
meillo@159 1369 displaying the header line makes little sense.
meillo@95 1370 Hence, the
meillo@95 1371 .Sw -[no]header
meillo@217 1372 switch was removed and headers are never printed
meillo@217 1373 .Ci 601cc73d1fa05ce96faa728f036d6c51b91701c7 .
meillo@95 1374 .P
meillo@95 1375 In
meillo@95 1376 .Pn mhlist ,
meillo@95 1377 the
meillo@95 1378 .Sw -[no]header
meillo@217 1379 switches were removed, as well
meillo@217 1380 .Ci b24f96523aaf60e44e04a3ffb1d22e69a13a602f .
meillo@217 1381 In this case, the headers are printed always because the output
meillo@217 1382 is not self-explaining.
meillo@95 1383 .P
meillo@95 1384 .Pn scan
meillo@95 1385 also had
meillo@95 1386 .Sw -[no]header
meillo@95 1387 switches.
meillo@217 1388 Printing this header had been sensible until the introduction of
meillo@217 1389 format strings made it impossible to display column headings.
meillo@95 1390 Only the folder name and the current date remained to be printed.
meillo@217 1391 As this information can be perfectly generated with
meillo@95 1392 .Pn folder
meillo@95 1393 and
meillo@95 1394 .Pn date ,
meillo@217 1395 the switches were removed
meillo@217 1396 .Ci c477dc5d1d03fa6d9a8ab3dd3508c63cbddc044e .
meillo@95 1397 .P
meillo@95 1398 By removing all
meillo@95 1399 .Sw -header
meillo@95 1400 switches, the collision with
meillo@95 1401 .Sw -help
meillo@95 1402 on the first two letters was resolved.
meillo@95 1403 Currently,
meillo@95 1404 .Sw -h
meillo@95 1405 evaluates to
meillo@95 1406 .Sw -help
meillo@95 1407 for all tools of mmh.
meillo@95 1408
meillo@95 1409
meillo@139 1410 .U3 "Suppressing Edits or the Invocation of the WhatNow Shell
meillo@95 1411 .P
meillo@95 1412 The
meillo@95 1413 .Sw -noedit
meillo@100 1414 switch of
meillo@95 1415 .Pn comp ,
meillo@95 1416 .Pn repl ,
meillo@95 1417 .Pn forw ,
meillo@95 1418 .Pn dist ,
meillo@95 1419 and
meillo@95 1420 .Pn whatnow
meillo@217 1421 was removed and replaced by specifying
meillo@95 1422 .Sw -editor
meillo@217 1423 with an empty argument
meillo@217 1424 .Ci 75fca31a5b9d5c1a99c74ab14c94438d8852fba9 .
meillo@95 1425 (Specifying
meillo@159 1426 .Cl "-editor /bin/true
meillo@217 1427 is nearly the same. It differs only in setting the previous editor.)
meillo@95 1428 .P
meillo@95 1429 The more important change is the removal of the
meillo@95 1430 .Sw -nowhatnowproc
meillo@217 1431 switch
meillo@217 1432 .Ci ee4f43cf2ef0084ec698e4e87159a94c01940622 .
meillo@217 1433 This switch had once introduced an awkward behavior,
meillo@217 1434 as explained in nmh's man page for
meillo@95 1435 .Mp comp (1):
meillo@98 1436 .QS
meillo@164 1437 The
meillo@164 1438 .Sw -editor
meillo@164 1439 .Ar editor
meillo@164 1440 switch indicates the editor to use for
meillo@164 1441 the initial edit. Upon exiting from the editor,
meillo@164 1442 .Pn comp
meillo@164 1443 will invoke the
meillo@164 1444 .Pn whatnow
meillo@164 1445 program. See
meillo@164 1446 .Mp whatnow (1)
meillo@164 1447 for a discussion of available options.
meillo@164 1448 The invocation of this program can be
meillo@164 1449 inhibited by using the
meillo@164 1450 .Sw -nowhatnowproc
meillo@164 1451 switch. (In truth of fact, it is the
meillo@164 1452 .Pn whatnow
meillo@164 1453 program which starts the initial edit.
meillo@164 1454 Hence,
meillo@164 1455 .Sw -nowhatnowproc
meillo@164 1456 will prevent any edit from occurring.)
meillo@98 1457 .QE
meillo@95 1458 .P
meillo@95 1459 Effectively, the
meillo@95 1460 .Sw -nowhatnowproc
meillo@217 1461 switch caused only only a draft message to be created.
meillo@95 1462 As
meillo@159 1463 .Cl "-whatnowproc /bin/true
meillo@217 1464 does the same, the
meillo@95 1465 .Sw -nowhatnowproc
meillo@95 1466 switch was removed for being redundant.
meillo@95 1467
meillo@95 1468
meillo@95 1469
meillo@95 1470 .U3 "Various
meillo@95 1471 .BU
meillo@139 1472 With the removal of MMDF maildrop format support,
meillo@139 1473 .Pn packf
meillo@139 1474 and
meillo@139 1475 .Pn rcvpack
meillo@139 1476 no longer needed their
meillo@139 1477 .Sw -mbox
meillo@139 1478 and
meillo@139 1479 .Sw -mmdf
meillo@139 1480 switches.
meillo@217 1481 The behavior of
meillo@139 1482 .Sw -mbox
meillo@217 1483 is the sole behavior now
meillo@217 1484 .Ci 3916ab66ad5d183705ac12357621ea8661afd3c0 .
meillo@171 1485 Further rework in both tools made the
meillo@139 1486 .Sw -file
meillo@217 1487 switch unnecessary
meillo@217 1488 .Ci ca1023716d4c2ab890696f3e41fa0d94267a940e .
meillo@139 1489
meillo@139 1490 .BU
meillo@217 1491 Mmh's tools do no longer clear the screen (\c
meillo@139 1492 .Pn scan 's
meillo@139 1493 and
meillo@139 1494 .Pn mhl 's
meillo@139 1495 .Sw -[no]clear
meillo@139 1496 switches
meillo@139 1497 .Ci e57b17343dcb3ff373ef4dd089fbe778f0c7c270
meillo@139 1498 .Ci 943765e7ac5693ae177fd8d2b5a2440e53ce816e ).
meillo@217 1499 Neither does
meillo@139 1500 .Pn mhl
meillo@139 1501 ring the bell (\c
meillo@139 1502 .Sw -[no]bell
meillo@139 1503 .Ci e11983f44e59d8de236affa5b0d0d3067c192e24 )
meillo@217 1504 nor does it page the output itself (\c
meillo@139 1505 .Sw -length
meillo@139 1506 .Ci 5b9d883db0318ed2b84bb82dee880d7381f99188 ).
meillo@139 1507 Generally, the pager to use is no longer specified with the
meillo@139 1508 .Sw -[no]moreproc
meillo@139 1509 command line switches for
meillo@139 1510 .Pn mhl
meillo@139 1511 and
meillo@139 1512 .Pn show /\c
meillo@217 1513 .Pn mhshow
meillo@217 1514 .Ci 39e87a75b5c2d3572ec72e717720b44af291e88a .
meillo@139 1515
meillo@139 1516 .BU
meillo@96 1517 In order to avoid prefix collisions among switch names, the
meillo@95 1518 .Sw -version
meillo@95 1519 switch was renamed to
meillo@95 1520 .Sw -Version
meillo@217 1521 (with capital `V')
meillo@217 1522 .Ci 32b2354dbaf4bf934936eb5b102a4a3d2fdd209a .
meillo@95 1523 Every program has the
meillo@95 1524 .Sw -version
meillo@95 1525 switch but its first three letters collided with the
meillo@95 1526 .Sw -verbose
meillo@95 1527 switch, present in many programs.
meillo@95 1528 The rename solved this problem once for all.
meillo@95 1529 Although this rename breaks a basic interface, having the
meillo@95 1530 .Sw -V
meillo@95 1531 abbreviation to display the version information, isn't all too bad.
meillo@139 1532
meillo@95 1533 .BU
meillo@95 1534 .Sw -[no]preserve
meillo@95 1535 of
meillo@95 1536 .Pn refile
meillo@168 1537 was removed
meillo@168 1538 .Ci 8edc5aaf86f9f77124664f6801bc6c6cdf258173
meillo@168 1539 because what use was it anyway?
meillo@168 1540 Quoting nmh's man page
meillo@165 1541 .Mp refile (1):
meillo@98 1542 .QS
meillo@95 1543 Normally when a message is refiled, for each destination
meillo@95 1544 folder it is assigned the number which is one above the current
meillo@95 1545 highest message number in that folder. Use of the
meillo@164 1546 .Sw -preserv
meillo@164 1547 [sic!] switch will override this message renaming, and try
meillo@95 1548 to preserve the number of the message. If a conflict for a
meillo@164 1549 particular folder occurs when using the
meillo@164 1550 .Sw -preserve
meillo@164 1551 switch, then
meillo@164 1552 .Pn refile
meillo@164 1553 will use the next available message number which
meillo@95 1554 is above the message number you wish to preserve.
meillo@98 1555 .QE
meillo@139 1556
meillo@95 1557 .BU
meillo@95 1558 The removal of the
meillo@95 1559 .Sw -[no]reverse
meillo@95 1560 switches of
meillo@95 1561 .Pn scan
meillo@97 1562 .Ci 8edc5aaf86f9f77124664f6801bc6c6cdf258173
meillo@217 1563 is a bug fix.
meillo@217 1564 This is supported by the comments
meillo@95 1565 ``\-[no]reverse under #ifdef BERK (I really HATE this)''
meillo@95 1566 by Rose and
meillo@95 1567 ``Lists messages in reverse order with the `\-reverse' switch.
meillo@217 1568 This should be considered a bug'' by Romine in the changelogs.
meillo@217 1569 The question remains why neither Rose nor Romine have fixed this
meillo@217 1570 bug in the eighties when they wrote these comments.
meillo@93 1571
meillo@93 1572
meillo@93 1573
meillo@102 1574
meillo@102 1575
meillo@95 1576
meillo@95 1577
meillo@133 1578 .\" --------------------------------------------------------------
meillo@74 1579 .H1 "Modernizing
meillo@102 1580 .P
meillo@164 1581 In the more than thirty years of MH's existence, its code base was
meillo@159 1582 increasingly extended.
meillo@118 1583 New features entered the project and became alternatives to the
meillo@118 1584 existing behavior.
meillo@217 1585 Relics from several decades have gathered in the code base
meillo@118 1586 but seldom obsolete features were dropped.
meillo@118 1587 This section describes the removing of old code
meillo@118 1588 and the modernizing of the default setup.
meillo@118 1589 It focuses on the functional aspect only;
meillo@154 1590 the non-functional aspects of code style are discussed in Sec.
meillo@154 1591 .Cf code-style .
meillo@58 1592
meillo@58 1593
meillo@212 1594 .H2 "Code Relics
meillo@0 1595 .P
meillo@217 1596 My position regarding the removal of obsolete code
meillo@159 1597 is much more revolutional than the nmh community appreciates.
meillo@217 1598 Working on an experimental version, I was able to quickly drop
meillo@217 1599 functionality that I considered ancient.
meillo@104 1600 The need for consensus with peers would have slowed this process down.
meillo@104 1601 Without the need to justify my decisions, I was able to rush forward.
meillo@217 1602 .P
meillo@110 1603 In December 2011, Paul Vixie motivated the nmh developers to just
meillo@104 1604 do the work:
meillo@104 1605 .[
meillo@104 1606 paul vixie edginess nmh-workers
meillo@104 1607 .]
meillo@104 1608 .QS
meillo@104 1609 let's stop walking on egg shells with this code base. there's no need to
meillo@104 1610 discuss whether to keep using vfork, just note in [sic!] passing, [...]
meillo@104 1611 we don't need a separate branch for removing vmh
meillo@104 1612 or ridding ourselves of #ifdef's or removing posix replacement functions
meillo@164 1613 or depending on pure ansi/posix ``libc''.
meillo@104 1614 .QP
meillo@164 1615 these things should each be a day or two of work and the ``main branch''
meillo@104 1616 should just be modern. [...]
meillo@104 1617 let's push forward, aggressively.
meillo@104 1618 .QE
meillo@104 1619 .LP
meillo@104 1620 I did so already in the months before.
meillo@104 1621 I pushed forward.
meillo@104 1622 I simply dropped the cruft.
meillo@104 1623 .P
meillo@104 1624 The decision to drop a feature was based on literature research and
meillo@159 1625 careful thinking, but whether having had contact with this particular
meillo@104 1626 feature within my own computer life served as a rule of thumb.
meillo@159 1627 I explained my reasons in the commit messages
meillo@109 1628 in the version control system.
meillo@104 1629 Hence, others can comprehend my view and argue for undoing the change
meillo@104 1630 if I have missed an important aspect.
meillo@109 1631 I was quick in dropping parts.
meillo@179 1632 I rather include falsely dropped parts again, than going at a slower pace.
meillo@179 1633 Mmh is experimental work; it requires tough decisions.
meillo@12 1634
meillo@102 1635
meillo@217 1636 .U3 "Process Forking
meillo@12 1637 .P
meillo@109 1638 Being a tool chest, MH creates many processes.
meillo@104 1639 In earlier times
meillo@104 1640 .Fu fork()
meillo@104 1641 had been an expensive system call, because the process's image needed
meillo@159 1642 to be completely duplicated at once.
meillo@200 1643 This expensive work was especially unnecessary in the commonly occurring
meillo@159 1644 case wherein the image is replaced by a call to
meillo@104 1645 .Fu exec()
meillo@104 1646 right after having forked the child process.
meillo@104 1647 The
meillo@104 1648 .Fu vfork()
meillo@104 1649 system call was invented to speed up this particular case.
meillo@104 1650 It completely omits the duplication of the image.
meillo@104 1651 On old systems this resulted in significant speed ups.
meillo@104 1652 Therefore MH used
meillo@104 1653 .Fu vfork()
meillo@104 1654 whenever possible.
meillo@12 1655 .P
meillo@104 1656 Modern memory management units support copy-on-write semantics, which make
meillo@104 1657 .Fu fork()
meillo@104 1658 almost as fast as
meillo@104 1659 .Fu vfork() .
meillo@104 1660 The man page of
meillo@104 1661 .Mp vfork (2)
meillo@104 1662 in FreeBSD 8.0 states:
meillo@104 1663 .QS
meillo@104 1664 This system call will be eliminated when proper system sharing mechanisms
meillo@104 1665 are implemented. Users should not depend on the memory sharing semantics
meillo@104 1666 of vfork() as it will, in that case, be made synonymous to fork(2).
meillo@104 1667 .QE
meillo@104 1668 .LP
meillo@104 1669 Vixie supports the removal with the note that ``the last
meillo@104 1670 system on which fork was so slow that an mh user would notice it, was
meillo@104 1671 Eunice. that was 1987''.
meillo@104 1672 .[
meillo@104 1673 nmh-workers vixie edginess
meillo@104 1674 .]
meillo@104 1675 I replaced all calls to
meillo@104 1676 .Fu vfork()
meillo@104 1677 with calls to
meillo@217 1678 .Fu fork()
meillo@217 1679 .Ci 40821f5c1316e9205a08375e7075909cc9968e7d .
meillo@104 1680 .P
meillo@104 1681 Related to the costs of
meillo@104 1682 .Fu fork()
meillo@104 1683 is the probability of its success.
meillo@109 1684 In the eighties, on heavy loaded systems, calls to
meillo@104 1685 .Fu fork()
meillo@104 1686 were prone to failure.
meillo@104 1687 Hence, many of the
meillo@104 1688 .Fu fork()
meillo@104 1689 calls in the code were wrapped into loops to retry the
meillo@104 1690 .Fu fork()
meillo@217 1691 several times, to increase the chances to succeed eventually.
meillo@109 1692 On modern systems, a failing
meillo@104 1693 .Fu fork()
meillo@109 1694 call is unusual.
meillo@104 1695 Hence, in the rare case when
meillo@104 1696 .Fu fork()
meillo@217 1697 fails, mmh programs simply abort
meillo@217 1698 .Ci 5fbf37ee68e018998ada61eeab73e035b26834b6 .
meillo@12 1699
meillo@12 1700
meillo@109 1701 .U3 "Header Fields
meillo@104 1702 .BU
meillo@84 1703 The
meillo@84 1704 .Hd Encrypted
meillo@104 1705 header field was introduced by RFC\|822,
meillo@109 1706 but already marked as legacy in RFC\|2822.
meillo@109 1707 Today, OpenPGP provides the basis for standardized exchange of encrypted
meillo@104 1708 messages [RFC\|4880, RFC\|3156].
meillo@109 1709 Hence, the support for
meillo@104 1710 .Hd Encrypted
meillo@217 1711 header fields is removed in mmh
meillo@217 1712 .Ci 064527f7b57ab050e5af13e15ad99aeeab125857 .
meillo@104 1713 .BU
meillo@159 1714 The native support for
meillo@84 1715 .Hd Face
meillo@217 1716 header fields has been removed, as well
meillo@217 1717 .Ci 8e5be81f784682822f5e868c1bf3c8624682bd23 .
meillo@104 1718 This feature is similar to the
meillo@84 1719 .Hd X-Face
meillo@84 1720 header field in its intent,
meillo@21 1721 but takes a different approach to store the image.
meillo@84 1722 Instead of encoding the image data directly into the header field,
meillo@109 1723 it contains the hostname and UDP port where the image
meillo@109 1724 date can be retrieved.
meillo@159 1725 There is even a third Face system,
meillo@109 1726 which is the successor of
meillo@109 1727 .Hd X-Face ,
meillo@109 1728 although it re-uses the
meillo@104 1729 .Hd Face
meillo@217 1730 header field name.
meillo@109 1731 It was invented in 2005 and supports colored PNG images.
meillo@104 1732 None of the Face systems described here is popular today.
meillo@104 1733 Hence, mmh has no direct support for them.
meillo@104 1734 .BU
meillo@154 1735 .Id content-md5
meillo@104 1736 The
meillo@104 1737 .Hd Content-MD5
meillo@104 1738 header field was introduced by RFC\|1864.
meillo@104 1739 It provides detection of data corruption during the transfer.
meillo@104 1740 But it can not ensure verbatim end-to-end delivery of the contents
meillo@104 1741 [RFC\|1864].
meillo@104 1742 The proper approach to verify content integrity in an
meillo@227 1743 end-to-end relationship is the use of digital signatures [RFC\|4880].
meillo@104 1744 On the other hand, transfer protocols should detect corruption during
meillo@109 1745 the transmission.
meillo@109 1746 The TCP includes a checksum field therefore.
meillo@104 1747 These two approaches in combinations render the
meillo@104 1748 .Hd Content-MD5
meillo@104 1749 header field superfluous.
meillo@109 1750 Not a single one out of 4\|200 messages from two decades
meillo@217 1751 in the nmh-workers mailing list archive
meillo@217 1752 .[
meillo@217 1753 nmh-workers mailing list archive website
meillo@217 1754 .]
meillo@217 1755 contains a
meillo@104 1756 .Hd Content-MD5
meillo@104 1757 header field.
meillo@104 1758 Neither did any of the 60\|000 messages in my personal mail storage.
meillo@217 1759 Removing the support for this header field
meillo@217 1760 .Ci 31dc797eb5178970d68962ca8939da3fd9a8efda ,
meillo@104 1761 removed the last place where MD5 computation was needed.
meillo@104 1762 Hence, the MD5 code could be removed as well.
meillo@104 1763 Over 500 lines of code vanished by this one change.
meillo@104 1764
meillo@104 1765
meillo@104 1766 .U3 "MMDF maildrop support
meillo@21 1767 .P
meillo@217 1768 This type of maildrop format is conceptionally similar to the mbox format,
meillo@139 1769 but uses a different message delimiter (`\fL\\1\\1\\1\\1\fP',
meillo@139 1770 commonly written as `\fL^A^A^A^A\fP', instead of `\fLFrom\0\fP').
meillo@104 1771 Mbox is the de-facto standard maildrop format on Unix,
meillo@159 1772 whereas the MMDF maildrop format is now forgotten.
meillo@217 1773 Mbox remains as the only packed mailbox format, supported in mmh.
meillo@104 1774 .P
meillo@109 1775 The simplifications within the code were moderate.
meillo@109 1776 Mainly, the reading and writing of MMDF mailbox files was removed.
meillo@109 1777 But also, switches of
meillo@109 1778 .Pn packf
meillo@104 1779 and
meillo@109 1780 .Pn rcvpack
meillo@217 1781 could be removed
meillo@217 1782 .Ci 3916ab66ad5d183705ac12357621ea8661afd3c0 .
meillo@109 1783 In the message parsing function
meillo@109 1784 .Fn sbr/m_getfld.c ,
meillo@217 1785 knowledge of MMDF packed mail boxes was removed
meillo@217 1786 .Ci 684ec30d81e1223a282764452f4902ed4ad1c754 .
meillo@109 1787 Further code structure simplifications may be possible there,
meillo@109 1788 because only one single packed mailbox format is left to be supported.
meillo@104 1789 I have not worked on them yet because
meillo@104 1790 .Fu m_getfld()
meillo@104 1791 is heavily optimized and thus dangerous to touch.
meillo@104 1792 The risk of damaging the intricate workings of the optimized code is
meillo@104 1793 too high.
meillo@104 1794
meillo@12 1795
meillo@101 1796 .U3 "Prompter's Control Keys
meillo@20 1797 .P
meillo@20 1798 The program
meillo@20 1799 .Pn prompter
meillo@104 1800 queries the user to fill in a message form.
meillo@217 1801 When used as
meillo@104 1802 .Cl "comp -editor prompter" ,
meillo@20 1803 the resulting behavior is similar to
meillo@20 1804 .Pn mailx .
meillo@51 1805 Apparently,
meillo@20 1806 .Pn prompter
meillo@173 1807 had not been touched lately.
meillo@104 1808 Otherwise it's hardly explainable why it
meillo@20 1809 still offered the switches
meillo@84 1810 .Sw -erase
meillo@84 1811 .Ar chr
meillo@20 1812 and
meillo@84 1813 .Sw -kill
meillo@84 1814 .Ar chr
meillo@20 1815 to name the characters for command line editing.
meillo@21 1816 The times when this had been necessary are long time gone.
meillo@20 1817 Today these things work out-of-the-box, and if not, are configured
meillo@20 1818 with the standard tool
meillo@20 1819 .Pn stty .
meillo@104 1820 The switches are removed now
meillo@104 1821 .Ci 0bd9750710cdbab80cfb4036dd87af20afe1552f .
meillo@20 1822
meillo@104 1823
meillo@109 1824 .U3 "Hardcopy Terminal Support
meillo@21 1825 .P
meillo@109 1826 More of a funny anecdote is a check for being connected to a
meillo@109 1827 hardcopy terminal.
meillo@159 1828 It remained in the code until spring 2012, when I finally removed it
meillo@104 1829 .Ci b7764c4a6b71d37918a97594d866258f154017ca .
meillo@21 1830 .P
meillo@109 1831 The check only prevented a pager to be placed between the printing
meillo@104 1832 program (\c
meillo@104 1833 .Pn mhl )
meillo@104 1834 and the terminal.
meillo@109 1835 In nmh, this could have been ensured statically with the
meillo@104 1836 .Sw -nomoreproc
meillo@109 1837 at the command line, too.
meillo@121 1838 In mmh, setting the profile entry
meillo@104 1839 .Pe Pager
meillo@104 1840 or the environment variable
meillo@104 1841 .Ev PAGER
meillo@104 1842 to
meillo@109 1843 .Pn cat
meillo@159 1844 is sufficient.
meillo@104 1845
meillo@104 1846
meillo@21 1847
meillo@12 1848
meillo@58 1849 .H2 "Attachments
meillo@22 1850 .P
meillo@101 1851 The mind model of email attachments is unrelated to MIME.
meillo@217 1852 Although the MIME RFCs [RFC\|2045\(en2049] define the technical
meillo@217 1853 requirements for having attachments, they do not mention the term.
meillo@101 1854 Instead of attachments, MIME talks about ``multi-part message bodies''
meillo@101 1855 [RFC\|2045], a more general concept.
meillo@101 1856 Multi-part messages are messages
meillo@101 1857 ``in which one or more different
meillo@101 1858 sets of data are combined in a single body''
meillo@101 1859 [RFC\|2046].
meillo@101 1860 MIME keeps its descriptions generic;
meillo@101 1861 it does not imply specific usage models.
meillo@217 1862 Today, one usage model is prevalent: attachments.
meillo@101 1863 The idea is having a main text document with files of arbitrary kind
meillo@101 1864 attached to it.
meillo@101 1865 In MIME terms, this is a multi-part message having a text part first
meillo@110 1866 and parts of arbitrary type following.
meillo@101 1867 .P
meillo@101 1868 MH's MIME support is a direct implementation of the RFCs.
meillo@101 1869 The perception of the topic described in the RFCs is clearly visible
meillo@101 1870 in MH's implementation.
meillo@159 1871 As a result,
meillo@159 1872 MH had all the MIME features but no idea of attachments.
meillo@173 1873 But users do not need all the MIME features,
meillo@109 1874 they want convenient attachment handling.
meillo@109 1875
meillo@102 1876
meillo@102 1877 .U3 "Composing MIME Messages
meillo@102 1878 .P
meillo@102 1879 In order to improve the situation on the message composing side,
meillo@217 1880 Jon Steinhart had added an attachment system to nmh in 2002
meillo@217 1881 .Ci 7480dbc14bc90f2d872d434205c0784704213252 .
meillo@102 1882 In the file
meillo@102 1883 .Fn docs/README-ATTACHMENTS ,
meillo@217 1884 he described his motivation to do so:
meillo@101 1885 .QS
meillo@159 1886 Although nmh contains the necessary functionality for MIME message
meillo@159 1887 handing [sic!], the interface to this functionality is pretty obtuse.
meillo@101 1888 There's no way that I'm ever going to convince my partner to write
meillo@101 1889 .Pn mhbuild
meillo@101 1890 composition files!
meillo@101 1891 .QE
meillo@102 1892 .LP
meillo@102 1893 With this change, the mind model of attachments entered nmh.
meillo@102 1894 In the same document:
meillo@101 1895 .QS
meillo@101 1896 These changes simplify the task of managing attachments on draft files.
meillo@101 1897 They allow attachments to be added, listed, and deleted.
meillo@101 1898 MIME messages are automatically created when drafts with attachments
meillo@101 1899 are sent.
meillo@101 1900 .QE
meillo@102 1901 .LP
meillo@217 1902 Unfortunately, the attachment system, like every new facilities in nmh,
meillo@110 1903 was inactive by default.
meillo@101 1904 .P
meillo@217 1905 During my time in Argentina, I tried to improve the attachment system.
meillo@217 1906 But, after long discussions my patch died as a proposal on the
meillo@217 1907 mailing list because of great opposition in the nmh community.
meillo@101 1908 .[
meillo@101 1909 nmh-workers attachment proposal
meillo@101 1910 .]
meillo@217 1911 In January 2012, I extended the patch and applied it to mmh
meillo@217 1912 .Ci 8ff284ff9167eff8f5349481529332d59ed913b1 .
meillo@102 1913 In mmh, the attachment system is active by default.
meillo@102 1914 Instead of command line switches, the
meillo@102 1915 .Pe Attachment-Header
meillo@102 1916 profile entry is used to specify
meillo@102 1917 the name of the attachment header field.
meillo@102 1918 It is pre-defined to
meillo@102 1919 .Hd Attach .
meillo@101 1920 .P
meillo@159 1921 To add an attachment to a draft, a header line needs to be added:
meillo@101 1922 .VS
meillo@101 1923 To: bob
meillo@101 1924 Subject: The file you wanted
meillo@101 1925 Attach: /path/to/the/file-bob-wanted
meillo@101 1926 --------
meillo@101 1927 Here it is.
meillo@101 1928 VE
meillo@101 1929 The header field can be added to the draft manually in the editor,
meillo@102 1930 or by using the `attach' command at the WhatNow prompt, or
meillo@102 1931 non-interactively with
meillo@101 1932 .Pn anno :
meillo@101 1933 .VS
meillo@102 1934 anno -append -nodate -component Attach -text /path/to/attachment
meillo@101 1935 VE
meillo@102 1936 Drafts with attachment headers are converted to MIME automatically by
meillo@102 1937 .Pn send .
meillo@102 1938 The conversion to MIME is invisible to the user.
meillo@159 1939 The draft stored in the draft folder is always in source form with
meillo@101 1940 attachment headers.
meillo@179 1941 If the MIMEification fails (e.g. because the file to attach
meillo@179 1942 is not accessible) the original draft is not changed.
meillo@101 1943 .P
meillo@102 1944 The attachment system handles the forwarding of messages, too.
meillo@173 1945 If the attachment header value starts with a plus character (`\fL+\fP'),
meillo@101 1946 like in
meillo@101 1947 .Cl "Attach: +bob 30 42" ,
meillo@159 1948 the given messages in the specified folder will be attached.
meillo@101 1949 This allowed to simplify
meillo@217 1950 .Pn forw
meillo@217 1951 .Ci f41f04cf4ceca7355232cf7413e59afafccc9550 .
meillo@101 1952 .P
meillo@101 1953 Closely related to attachments is non-ASCII text content,
meillo@217 1954 because it requires MIME as well.
meillo@102 1955 In nmh, the user needed to call `mime' at the WhatNow prompt
meillo@101 1956 to have the draft converted to MIME.
meillo@102 1957 This was necessary whenever the draft contained non-ASCII characters.
meillo@101 1958 If the user did not call `mime', a broken message would be sent.
meillo@101 1959 Therefore, the
meillo@101 1960 .Pe automimeproc
meillo@101 1961 profile entry could be specified to have the `mime' command invoked
meillo@102 1962 automatically each time.
meillo@179 1963 Unfortunately, this approach conflicted with the attachment system
meillo@101 1964 because the draft would already be in MIME format at the time
meillo@101 1965 when the attachment system wanted to MIMEify it.
meillo@102 1966 To use nmh's attachment system, `mime' must not be called at the
meillo@102 1967 WhatNow prompt and
meillo@101 1968 .Pe automimeproc
meillo@102 1969 must not be set in the profile.
meillo@101 1970 But then the case of non-ASCII text without attachment headers was
meillo@101 1971 not caught.
meillo@102 1972 All in all, the solution was complex and irritating.
meillo@168 1973 My patch from December 2010
meillo@168 1974 .[
meillo@168 1975 nmh-workers attachment proposal
meillo@168 1976 .]
meillo@168 1977 would have simplified the situation.
meillo@102 1978 .P
meillo@101 1979 Mmh's current solution is even more elaborate.
meillo@101 1980 Any necessary MIMEification is done automatically.
meillo@101 1981 There is no `mime' command at the WhatNow prompt anymore.
meillo@102 1982 The draft will be converted automatically to MIME when either an
meillo@102 1983 attachment header or non-ASCII text is present.
meillo@173 1984 Furthermore, the hash character (`\fL#\fP') is not special any more
meillo@159 1985 at line beginnings in the draft message.
meillo@159 1986 Users need not concern themselves with the whole topic at all.
meillo@101 1987 .P
meillo@102 1988 Although the new approach does not anymore support arbitrary MIME
meillo@102 1989 compositions directly, the full power of
meillo@101 1990 .Pn mhbuild
meillo@101 1991 can still be accessed.
meillo@217 1992 Given no attachment headers are included, users can create
meillo@101 1993 .Pn mhbuild
meillo@102 1994 composition drafts like in nmh.
meillo@217 1995 Then, at the WhatNow prompt, they can invoke
meillo@101 1996 .Cl "edit mhbuild
meillo@217 1997 to convert the draft to MIME.
meillo@217 1998 Because the resulting draft neither contains non-ASCII characters
meillo@102 1999 nor has it attachment headers, the attachment system will not touch it.
meillo@101 2000 .P
meillo@159 2001 The approach taken in mmh is tailored towards today's most common case:
meillo@159 2002 a text part, possibly with attachments.
meillo@159 2003 This case was simplified.
meillo@102 2004
meillo@112 2005
meillo@102 2006 .U3 "MIME Type Guessing
meillo@102 2007 .P
meillo@159 2008 From the programmer's point of view, the use of
meillo@101 2009 .Pn mhbuild
meillo@159 2010 composition drafts had one notable advantage over attachment headers:
meillo@159 2011 The user provides the appropriate MIME types for files to include.
meillo@217 2012 The new attachment system needs to find out the correct MIME type itself.
meillo@217 2013 This is a difficult task.
meillo@102 2014 Determining the correct MIME type of content is partly mechanical,
meillo@102 2015 partly intelligent work.
meillo@102 2016 Forcing the user to find out the correct MIME type,
meillo@102 2017 forces him to do partly mechanical work.
meillo@179 2018 Letting the computer do the work can lead to bad choices for difficult
meillo@102 2019 content.
meillo@217 2020 For mmh, the latter option was chosen to spare the user the work
meillo@217 2021 .Ci 3baec236a39c5c89a9bda8dbd988d643a21decc6 .
meillo@102 2022 .P
meillo@102 2023 Determining the MIME type by the suffix of the file name is a dumb
meillo@102 2024 approach, yet it is simple to implement and provides good results
meillo@102 2025 for the common cases.
meillo@217 2026 If no MIME type can be determined, text content is sent as `text/plain',
meillo@217 2027 anything else under the generic fall-back type `application/octet-stream'.
meillo@102 2028 Mmh implements this approach in the
meillo@102 2029 .Pn print-mimetype
meillo@217 2030 script
meillo@217 2031 .Ci 4b5944268ea0da7bb30598a27857304758ea9b44 .
meillo@102 2032 .P
meillo@112 2033 A far better, though less portable, approach is the use of
meillo@102 2034 .Pn file .
meillo@102 2035 This standard tool tries to determine the type of files.
meillo@102 2036 Unfortunately, its capabilities and accuracy varies from system to system.
meillo@102 2037 Additionally, its output was only intended for human beings,
meillo@102 2038 but not to be used by programs.
meillo@102 2039 Nevertheless, modern versions of GNU
meillo@102 2040 .Pn file ,
meillo@217 2041 which are prevalent on the popular GNU/Linux systems,
meillo@159 2042 provide MIME type output in machine-readable form.
meillo@217 2043 Although this solution is system-dependent,
meillo@102 2044 it solves the difficult problem well.
meillo@102 2045 On systems where GNU
meillo@102 2046 .Pn file ,
meillo@102 2047 version 5.04 or higher, is available it should be used.
meillo@102 2048 One needs to specify the following profile entry to do so:
meillo@102 2049 .VS
meillo@102 2050 Mime-Type-Query: file -b --mime
meillo@102 2051 VE
meillo@102 2052 .LP
meillo@102 2053 Other versions of
meillo@102 2054 .Pn file
meillo@217 2055 might possibly be usable with wrapper scripts that reformat the output.
meillo@102 2056 The diversity among
meillo@102 2057 .Pn file
meillo@102 2058 implementations is great; one needs to check the local variant.
meillo@102 2059 .P
meillo@102 2060 It is not possible in mmh to override the automatic MIME type guessing
meillo@102 2061 for a specific file.
meillo@159 2062 To do so, either the user would need to know in advance for which file
meillo@217 2063 the automatic guessing fails or the system would require interaction.
meillo@102 2064 I consider both cases impractical.
meillo@102 2065 The existing solution should be sufficient.
meillo@102 2066 If not, the user may always fall back to
meillo@102 2067 .Pn mhbuild
meillo@217 2068 composition drafts and bypass the attachment system.
meillo@101 2069
meillo@102 2070
meillo@102 2071 .U3 "Storing Attachments
meillo@102 2072 .P
meillo@169 2073 Extracting MIME parts of a message and storing them to disk is performed by
meillo@108 2074 .Pn mhstore .
meillo@108 2075 The program has two operation modes,
meillo@108 2076 .Sw -auto
meillo@108 2077 and
meillo@108 2078 .Sw -noauto .
meillo@108 2079 With the former one, each part is stored under the filename given in the
meillo@108 2080 MIME part's meta information, if available.
meillo@108 2081 This naming information is usually available for modern attachments.
meillo@108 2082 If no filename is available, this MIME part is stored as if
meillo@108 2083 .Sw -noauto
meillo@108 2084 would have been specified.
meillo@108 2085 In the
meillo@108 2086 .Sw -noauto
meillo@108 2087 mode, the parts are processed according to rules, defined by
meillo@108 2088 .Pe mhstore-store-*
meillo@108 2089 profile entries.
meillo@108 2090 These rules define generic filename templates for storing
meillo@108 2091 or commands to post-process the contents in arbitrary ways.
meillo@108 2092 If no matching rule is available the part is stored under a generic
meillo@108 2093 filename, built from message number, MIME part number, and MIME type.
meillo@108 2094 .P
meillo@108 2095 The
meillo@108 2096 .Sw -noauto
meillo@108 2097 mode had been the default in nmh because it was considered safe,
meillo@108 2098 in contrast to the
meillo@108 2099 .Sw -auto
meillo@108 2100 mode.
meillo@108 2101 In mmh,
meillo@108 2102 .Sw -auto
meillo@108 2103 is not dangerous anymore.
meillo@108 2104 Two changes were necessary:
meillo@171 2105 .LI 1
meillo@108 2106 Any directory path is removed from the proposed filename.
meillo@108 2107 Thus, the files are always stored in the expected directory.
meillo@108 2108 .Ci 41b6eadbcecf63c9a66aa5e582011987494abefb
meillo@171 2109 .LI 2
meillo@108 2110 Tar files are not extracted automatically any more.
meillo@108 2111 Thus, the rest of the file system will not be touched.
meillo@108 2112 .Ci 94c80042eae3383c812d9552089953f9846b1bb6
meillo@217 2113 .P
meillo@217 2114 In mmh, the result of
meillo@108 2115 .Cl "mhstore -auto
meillo@110 2116 can be foreseen from the output of
meillo@108 2117 .Cl "mhlist -verbose" .
meillo@217 2118 Although the
meillo@108 2119 .Sw -noauto
meillo@217 2120 mode is considered to be more powerful, it is less convenient and
meillo@108 2121 .Sw -auto
meillo@217 2122 is safe now.
meillo@217 2123 Additionally, storing attachments under their original name
meillo@217 2124 is intuitive.
meillo@108 2125 Hence,
meillo@108 2126 .Sw -auto
meillo@217 2127 serves better as the default option
meillo@217 2128 .Ci 3410b680416c49a7617491af38bc1929855a331d .
meillo@108 2129 .P
meillo@108 2130 Files are stored into the directory given by the
meillo@108 2131 .Pe Nmh-Storage
meillo@108 2132 profile entry, if set, or
meillo@108 2133 into the current working directory, otherwise.
meillo@108 2134 Storing to different directories is only possible with
meillo@108 2135 .Pe mhstore-store-*
meillo@108 2136 profile entries.
meillo@108 2137 .P
meillo@217 2138 Still existing files get overwritten silently in both modes.
meillo@108 2139 This can be considered a bug.
meillo@108 2140 Yet, each other behavior has its draw-backs, too.
meillo@108 2141 Refusing to replace files requires adding a
meillo@108 2142 .Sw -force
meillo@217 2143 switch.
meillo@108 2144 Users will likely need to invoke
meillo@108 2145 .Pn mhstore
meillo@108 2146 a second time with
meillo@159 2147 .Sw -force .
meillo@159 2148 Eventually, only the user can decide in the specific case.
meillo@108 2149 This requires interaction, which I like to avoid if possible.
meillo@108 2150 Appending a unique suffix to the filename is another bad option.
meillo@108 2151 For now, the behavior remains as it is.
meillo@108 2152 .P
meillo@108 2153 In mmh, only MIME parts of type message are special in
meillo@108 2154 .Pn mhstore 's
meillo@108 2155 .Sw -auto
meillo@108 2156 mode.
meillo@108 2157 Instead of storing message/rfc822 parts as files to disk,
meillo@108 2158 they are stored as messages into the current mail folder.
meillo@159 2159 The same applies to message/partial, although the parts are
meillo@159 2160 automatically reassembled beforehand.
meillo@159 2161 MIME parts of type message/external-body are not automatically retrieved
meillo@159 2162 anymore.
meillo@159 2163 Instead, information on how to retrieve them is output.
meillo@217 2164 Not supporting this rare case saved nearly one thousand lines of code
meillo@217 2165 .Ci 55e1d8c654ee0f7c45b9361ce34617983b454c32 .
meillo@217 2166 The MIME type `application/octet-stream; type=tar' is not special anymore.
meillo@217 2167 The automatically extracting of such MIME parts had been the
meillo@217 2168 dangerous part of the
meillo@108 2169 .Sw -auto
meillo@217 2170 mode
meillo@217 2171 .Ci 94c80042eae3383c812d9552089953f9846b1bb6 .
meillo@108 2172
meillo@102 2173
meillo@102 2174
meillo@102 2175 .U3 "Showing MIME Messages
meillo@227 2176 .Id showing-mime-msgs
meillo@102 2177 .P
meillo@114 2178 The program
meillo@114 2179 .Pn mhshow
meillo@217 2180 was written to display MIME messages.
meillo@114 2181 It implemented the conceptional view of the MIME RFCs.
meillo@114 2182 Nmh's
meillo@114 2183 .Pn mhshow
meillo@217 2184 handles each MIME part independently, presenting them separately
meillo@114 2185 to the user.
meillo@114 2186 This does not match today's understanding of email attachments,
meillo@114 2187 where displaying a message is seen to be a single, integrated operation.
meillo@114 2188 Today, email messages are expected to consist of a main text part
meillo@114 2189 plus possibly attachments.
meillo@217 2190 They are no more seen to be arbitrary MIME hierarchies with
meillo@114 2191 information on how to display the individual parts.
meillo@114 2192 I adjusted
meillo@114 2193 .Pn mhshow 's
meillo@114 2194 behavior to the modern view on the topic.
meillo@114 2195 .P
meillo@224 2196 One should note that this section completely ignores the original
meillo@114 2197 .Pn show
meillo@114 2198 program, because it was not capable to display MIME messages
meillo@227 2199 and is no longer part of mmh
meillo@227 2200 (cf. Sec.
meillo@227 2201 .Cf mhshow ).
meillo@114 2202 Although
meillo@114 2203 .Pn mhshow
meillo@114 2204 was renamed to
meillo@114 2205 .Pn show
meillo@114 2206 in mmh, this section uses the name
meillo@114 2207 .Pn mhshow ,
meillo@224 2208 in order to avoid confusion.
meillo@114 2209 .P
meillo@114 2210 In mmh, the basic idea is that
meillo@114 2211 .Pn mhshow
meillo@114 2212 should display a message in one single pager session.
meillo@114 2213 Therefore,
meillo@114 2214 .Pn mhshow
meillo@114 2215 invokes a pager session for all its output,
meillo@217 2216 whenever it prints to a terminal
meillo@217 2217 .Ci a4197ea6ffc5c1550e8b52d5a654bcaaaee04a4e .
meillo@114 2218 In consequence,
meillo@114 2219 .Pn mhl
meillo@217 2220 does no more invoke a pager
meillo@217 2221 .Ci 0e46503be3c855bddaeae3843e1b659279c35d70 .
meillo@114 2222 With
meillo@114 2223 .Pn mhshow
meillo@114 2224 replacing the original
meillo@114 2225 .Pn show ,
meillo@217 2226 the output of
meillo@114 2227 .Pn mhl
meillo@217 2228 no longer goes to the terminal directly, but through
meillo@114 2229 .Pn mhshow .
meillo@114 2230 Hence,
meillo@114 2231 .Pn mhl
meillo@114 2232 does not need to invoke a pager.
meillo@114 2233 The one and only job of
meillo@114 2234 .Pn mhl
meillo@114 2235 is to format messages or parts of them.
meillo@114 2236 The only place in mmh, where a pager is invoked is
meillo@114 2237 .Pn mhshow .
meillo@114 2238 .P
meillo@227 2239 Only text content is displayed.
meillo@227 2240 Other kinds of attachments are ignored.
meillo@217 2241 Non-text content needs to be converted to text by appropriate
meillo@217 2242 .Pe mhshow-show-*
meillo@217 2243 profile entries before, if this is possible and wanted.
meillo@217 2244 A common example for this are PDF files.
meillo@227 2245 .P
meillo@227 2246 MIME parts are always displayed serially.
meillo@114 2247 The request to display the MIME type `multipart/parallel' in parallel
meillo@114 2248 is ignored.
meillo@217 2249 It is simply treated as `multipart/mixed'
meillo@217 2250 .Ci d0581ba306a7299113a346f9b4c46ce97bc4cef6 .
meillo@217 2251 This was already possible to requested with the, now removed,
meillo@114 2252 .Sw -serialonly
meillo@114 2253 switch of
meillo@114 2254 .Pn mhshow .
meillo@179 2255 As MIME parts are always processed exclusively, i.e. serially,
meillo@217 2256 the `\fL%e\fP' escape in
meillo@114 2257 .Pe mhshow-show-*
meillo@217 2258 profile entries became useless and was thus removed
meillo@217 2259 .Ci a20d405db09b7ccca74d3e8c57550883da49e1ae .
meillo@114 2260 For parallel display, the attachments need to be stored to disk first.
meillo@114 2261 .P
meillo@114 2262 To display text content in foreign charsets, they need to be converted
meillo@114 2263 to the native charset.
meillo@114 2264 Therefore,
meillo@114 2265 .Pe mhshow-charset-*
meillo@217 2266 profile entries were needed.
meillo@169 2267 In mmh, the conversion is performed automatically by piping the
meillo@169 2268 text through the
meillo@114 2269 .Pn iconv
meillo@217 2270 command, if necessary
meillo@217 2271 .Ci 2433122c20baccb10b70b49c04c6b0497b5b3b60 .
meillo@114 2272 Custom
meillo@114 2273 .Pe mhshow-show-*
meillo@114 2274 rules for textual content might need a
meillo@114 2275 .Cl "iconv -f %c %f |
meillo@114 2276 prefix to have the text converted to the native charset.
meillo@114 2277 .P
meillo@121 2278 Although the conversion of foreign charsets to the native one
meillo@114 2279 has improved, it is not consistent enough.
meillo@114 2280 Further work needs to be done and
meillo@114 2281 the basic concepts in this field need to be re-thought.
meillo@114 2282 Though, the default setup of mmh displays message in foreign charsets
meillo@114 2283 correctly without the need to configure anything.
meillo@114 2284
meillo@114 2285
meillo@114 2286 .ig
meillo@114 2287
meillo@114 2288 .P
meillo@114 2289 mhshow/mhstore: Removed support for retrieving message/external-body parts.
meillo@173 2290 These tools will not download the contents automatically anymore. Instead,
meillo@114 2291 they print the information needed to get the contents. If someone should
meillo@114 2292 really receive one of those rare message/external-body messages, he can
meillo@114 2293 do the job manually. We save nearly a thousand lines of code. That's worth
meillo@114 2294 it!
meillo@114 2295 (The profile entry `nmh-access-ftp' and sbr/ruserpass.c for reading
meillo@114 2296 ~/.netrc are gone now.)
meillo@114 2297 .Ci 55e1d8c654ee0f7c45b9361ce34617983b454c32
meillo@114 2298
meillo@114 2299 ..
meillo@102 2300
meillo@58 2301
meillo@58 2302
meillo@166 2303 .H2 "Signing and Encrypting
meillo@22 2304 .P
meillo@166 2305 Nmh offers no direct support for digital signatures and message encryption.
meillo@157 2306 This functionality needed to be added through third-party software.
meillo@217 2307 In mmh, the functionality is included because it
meillo@217 2308 is a part of modern email and is likely wanted by users of mmh.
meillo@217 2309 A fresh mmh installation supports signing and encrypting
meillo@157 2310 out-of-the-box.
meillo@157 2311 Therefore, Neil Rickert's
meillo@157 2312 .Pn mhsign
meillo@157 2313 and
meillo@157 2314 .Pn mhpgp
meillo@157 2315 scripts
meillo@157 2316 .[
meillo@157 2317 neil rickert mhsign mhpgp
meillo@157 2318 .]
meillo@217 2319 were included
meillo@177 2320 .Ci f45cdc98117a84f071759462c7ae212f4bc5ab2e
meillo@177 2321 .Ci 58cf09aa36e9f7f352a127158bbf1c5678bc6ed8 .
meillo@177 2322 The scripts fit well because they are lightweight and
meillo@177 2323 similar of style to the existing tools.
meillo@217 2324 Additionally, no licensing difficulties appeared
meillo@157 2325 as they are part of the public domain.
meillo@112 2326 .P
meillo@157 2327 .Pn mhsign
meillo@157 2328 handles the signing and encrypting part.
meillo@157 2329 It comprises about 250 lines of shell code and interfaces between
meillo@223 2330 .Pn gnupg
meillo@217 2331 and the MH system.
meillo@177 2332 It was meant to be invoked manually at the WhatNow prompt, but in mmh,
meillo@157 2333 .Pn send
meillo@177 2334 invokes
meillo@217 2335 .Pn mhsign
meillo@177 2336 automatically
meillo@177 2337 .Ci c7b5e1df086bcc37ff40163ee67571f076cf6683 .
meillo@177 2338 Special header fields were introduced to request this action.
meillo@157 2339 If a draft contains the
meillo@157 2340 .Hd Sign
meillo@157 2341 header field,
meillo@157 2342 .Pn send
meillo@177 2343 will initiate the signing.
meillo@217 2344 The signing key is either chosen automatically or it is specified by the
meillo@157 2345 .Pe Pgpkey
meillo@157 2346 profile entry.
meillo@157 2347 .Pn send
meillo@217 2348 always create signatures using the PGP/MIME standard [RFC\|4880],
meillo@217 2349 but by invoking
meillo@217 2350 .Pn mhsign
meillo@217 2351 manually, old-style non-MIME signatures can be created as well.
meillo@177 2352 To encrypt an outgoing message, the draft needs to contain an
meillo@157 2353 .Hd Enc
meillo@157 2354 header field.
meillo@177 2355 Public keys of all recipients are searched for in the gnupg keyring and
meillo@177 2356 in a file called
meillo@177 2357 .Fn pgpkeys ,
meillo@177 2358 which contains exceptions and overrides.
meillo@157 2359 Unless public keys are found for all recipients,
meillo@177 2360 .Pn mhsign
meillo@177 2361 will refuse to encrypt it.
meillo@157 2362 Currently, messages with hidden (BCC) recipients can not be encrypted.
meillo@171 2363 This work is pending because it requires a structurally more complex
meillo@171 2364 approach.
meillo@157 2365 .P
meillo@177 2366 .Pn mhpgp
meillo@177 2367 is the companion to
meillo@177 2368 .Pn mhsign .
meillo@177 2369 It verifies signatures and decrypts messages.
meillo@217 2370 Encrypted messages can be either temporarily decrypted and displayed
meillo@177 2371 or permanently decrypted and stored into the current folder.
meillo@177 2372 Currently,
meillo@177 2373 .Pn mhpgp
meillo@177 2374 needs to be invoked manually.
meillo@177 2375 The integration into
meillo@177 2376 .Pn show
meillo@177 2377 and
meillo@177 2378 .Pn mhstore
meillo@217 2379 to verify signatures and decrypt messages as needed
meillo@217 2380 is planned but not yet realized.
meillo@177 2381 .P
meillo@217 2382 Both scripts were written for nmh.
meillo@217 2383 Hence they needed to be adjust
meillo@177 2384 according to the differences between nmh and mmh.
meillo@177 2385 For instance, they use the backup prefix no longer.
meillo@181 2386 Furthermore, compatibility support for old PGP features was dropped.
meillo@177 2387 .P
meillo@157 2388 The integrated message signing and encrypting support is one of the
meillo@157 2389 most recent features in mmh.
meillo@217 2390 It has not had the time to mature.
meillo@177 2391 User feedback and personal experience need to be accumulated to
meillo@177 2392 direct the further development of the facility.
meillo@217 2393 Already it seems to be worthwhile to consider adding
meillo@157 2394 .Sw -[no]sign
meillo@157 2395 and
meillo@157 2396 .Sw -[no]enc
meillo@157 2397 switches to
meillo@157 2398 .Pn send ,
meillo@177 2399 to be able to override the corresponding header fields.
meillo@177 2400 A profile entry:
meillo@157 2401 .VS
meillo@157 2402 send: -sign
meillo@157 2403 VE
meillo@177 2404 would then activate signing for all outgoing messages.
meillo@177 2405 With the present approach, a
meillo@177 2406 .Hd Send
meillo@177 2407 header component needs to be added to each draft template
meillo@177 2408 to achieve the same result.
meillo@177 2409 Adding the switches would ease the work greatly and keep the
meillo@177 2410 template files clean.
meillo@157 2411
meillo@58 2412
meillo@58 2413
meillo@102 2414
meillo@133 2415 .H2 "Draft and Trash Folder
meillo@131 2416 .P
meillo@58 2417
meillo@131 2418 .U3 "Draft Folder
meillo@154 2419 .Id draft-folder
meillo@131 2420 .P
meillo@131 2421 In the beginning, MH had the concept of a draft message.
meillo@217 2422 This was a file named
meillo@131 2423 .Fn draft
meillo@217 2424 in the MH directory, which was treated special.
meillo@131 2425 On composing a message, this draft file was used.
meillo@131 2426 When starting to compose another message before the former one was sent,
meillo@131 2427 the user had to decide among:
meillo@171 2428 .LI 1
meillo@168 2429 Using the old draft to finish and send it before starting with a new one.
meillo@171 2430 .LI 2
meillo@168 2431 Discarding the old draft and replacing it with a new one.
meillo@171 2432 .LI 3
meillo@168 2433 Preserving the old draft by refiling it to a folder.
meillo@171 2434 .LP
meillo@217 2435 Working on multiple drafts was only possible in alternation.
meillo@217 2436 For that, the current draft needed to be refiled to a folder and
meillo@168 2437 another one re-used for editing.
meillo@131 2438 Working on multiple drafts at the same time was impossible.
meillo@131 2439 The usual approach of switching to a different MH context did not
meillo@168 2440 help anything.
meillo@131 2441 .P
meillo@131 2442 The draft folder facility exists to
meillo@131 2443 allow true parallel editing of drafts, in a straight forward way.
meillo@131 2444 It was introduced by Marshall T. Rose, already in 1984.
meillo@131 2445 Similar to other new features, the draft folder was inactive by default.
meillo@131 2446 Even in nmh, the highly useful draft folder was not available
meillo@131 2447 out-of-the-box.
meillo@131 2448 At least, Richard Coleman added the man page
meillo@131 2449 .Mp mh-draft (5)
meillo@131 2450 to better document the feature.
meillo@131 2451 .P
meillo@131 2452 Not using the draft folder facility has the single advantage of having
meillo@131 2453 the draft file at a static location.
meillo@131 2454 This is simple in simple cases but the concept does not scale for more
meillo@131 2455 complex cases.
meillo@217 2456 The concept of the draft message is too limited for the problem
meillo@217 2457 it tries to solve.
meillo@131 2458 Therefore the draft folder was introduced.
meillo@131 2459 It is the more powerful and more natural concept.
meillo@131 2460 The draft folder is a folder like any other folder in MH.
meillo@131 2461 Its messages can be listed like any other messages.
meillo@131 2462 A draft message is no longer a special case.
meillo@131 2463 Tools do not need special switches to work on the draft message.
meillo@171 2464 Hence corner cases were removed.
meillo@131 2465 .P
meillo@131 2466 The trivial part of the work was activating the draft folder with a
meillo@131 2467 default name.
meillo@131 2468 I chose the name
meillo@217 2469 .Fn +drafts ,
meillo@131 2470 for obvious reasons.
meillo@131 2471 In consequence, the command line switches
meillo@131 2472 .Sw -draftfolder
meillo@131 2473 and
meillo@131 2474 .Sw -draftmessage
meillo@131 2475 could be removed.
meillo@131 2476 More difficult but also more improving was updating the tools to the
meillo@131 2477 new concept.
meillo@131 2478 For nearly three decades, the tools needed to support two draft handling
meillo@131 2479 approaches.
meillo@217 2480 By fully switching to the draft folder, the tools could be
meillo@217 2481 simplified by dropping the awkward draft message handling code.
meillo@131 2482 .Sw -draft
meillo@131 2483 switches were removed because operating on a draft message is no longer
meillo@131 2484 special.
meillo@131 2485 It became indistinguishable to operating on any other message.
meillo@168 2486 .Ci 337338b404931f06f0db2119c9e145e8ca5a9860
meillo@168 2487 .P
meillo@168 2488 There is no more need to query the user for draft handling
meillo@168 2489 .Ci 2d48b455c303a807041c35e4248955f8bec59eeb .
meillo@131 2490 It is always possible to add another new draft.
meillo@131 2491 Refiling drafts is without difference to refiling other messages.
meillo@168 2492 All of these special cases are gone.
meillo@131 2493 Yet, one draft-related switch remained.
meillo@131 2494 .Pn comp
meillo@131 2495 still has
meillo@131 2496 .Sw -[no]use
meillo@131 2497 for switching between two modes:
meillo@171 2498 .LI 1
meillo@217 2499 Modifying an existing draft, with
meillo@217 2500 .Sw -use .
meillo@171 2501 .LI 2
meillo@217 2502 Composing a new draft, possibly taking some existing message as template,
meillo@217 2503 with
meillo@217 2504 .Sw -nouse ,
meillo@217 2505 the default.
meillo@171 2506 .LP
meillo@131 2507 In either case, the behavior of
meillo@131 2508 .Pn comp
meillo@131 2509 is deterministic.
meillo@131 2510 .P
meillo@131 2511 .Pn send
meillo@131 2512 now operates on the current message in the draft folder by default.
meillo@131 2513 As message and folder can both be overridden by specifying them on
meillo@131 2514 the command line, it is possible to send any message in the mail storage
meillo@131 2515 by simply specifying its number and folder.
meillo@131 2516 In contrast to the other tools,
meillo@131 2517 .Pn send
meillo@131 2518 takes the draft folder as its default folder.
meillo@131 2519 .P
meillo@131 2520 Dropping the draft message concept in favor for the draft folder concept,
meillo@217 2521 replaced special cases with regular cases.
meillo@131 2522 This simplified the source code of the tools, as well as the concepts.
meillo@131 2523 In mmh, draft management does not break with the MH concepts
meillo@131 2524 but applies them.
meillo@133 2525 .Cl "scan +drafts" ,
meillo@133 2526 for instance, is a truly natural request.
meillo@217 2527 .P
meillo@169 2528 Most of the work was already performed by Rose in the eighties.
meillo@133 2529 The original improvement of mmh is dropping the old draft message approach
meillo@217 2530 and thus simplifying the tools, the documentation,
meillo@217 2531 and the system as a whole.
meillo@131 2532 Although my part in the draft handling improvement was small,
meillo@217 2533 it was important.
meillo@131 2534
meillo@131 2535
meillo@131 2536 .U3 "Trash Folder
meillo@154 2537 .Id trash-folder
meillo@131 2538 .P
meillo@131 2539 Similar to the situation for drafts is the situation for removed messages.
meillo@131 2540 Historically, a message was ``deleted'' by prepending a specific
meillo@173 2541 \fIbackup prefix\fP, usually the comma character,
meillo@173 2542 to the file name.
meillo@164 2543 The specific file would then be ignored by MH because only files with
meillo@164 2544 names consisting of digits only are treated as messages.
meillo@131 2545 Although files remained in the file system,
meillo@168 2546 the messages were no longer visible in MH.
meillo@168 2547 To truly delete them, a maintenance job was needed.
meillo@168 2548 Usually a cron job was installed to delete them after a grace time.
meillo@131 2549 For instance:
meillo@131 2550 .VS
meillo@131 2551 find $HOME/Mail -type f -name ',*' -ctime +7 -delete
meillo@131 2552 VE
meillo@168 2553 In such a setup, the original message could be restored
meillo@131 2554 within the grace time interval by stripping the
meillo@217 2555 backup prefix from the file name \(en usually but not always.
meillo@168 2556 If the last message of a folder with six messages (\fL1-6\fP) was removed,
meillo@131 2557 message
meillo@131 2558 .Fn 6 ,
meillo@168 2559 became file
meillo@131 2560 .Fn ,6 .
meillo@168 2561 If then a new message entered the same folder, it would be named with
meillo@168 2562 the number one above the highest existing message number.
meillo@168 2563 In this case the message would be named
meillo@217 2564 .Fn 6 ,
meillo@217 2565 reusing the number.
meillo@168 2566 If this new message would be removed as well,
meillo@217 2567 then the backup of the former message becomes overwritten.
meillo@168 2568 Hence, the ability to restore removed messages did not only depend on
meillo@181 2569 the sweeping cron job but also on the removing of further messages.
meillo@131 2570 It is undesirable to have such obscure and complex mechanisms.
meillo@168 2571 The user should be given a small set of clear assertions, such as
meillo@131 2572 ``Removed files are restorable within a seven-day grace time.''
meillo@131 2573 With the addition ``... unless a message with the same name in the
meillo@131 2574 same folder is removed before.'' the statement becomes complex.
meillo@217 2575 A user will hardly be able to keep track of all removals to know
meillo@131 2576 if the assertion still holds true for a specific file.
meillo@164 2577 In practice, the real mechanism is unclear to the user.
meillo@131 2578 .P
meillo@217 2579 Furthermore, the backup files were scattered within the whole mail storage.
meillo@217 2580 This complicated managing them.
meillo@217 2581 It was possible with the help of
meillo@131 2582 .Pn find ,
meillo@217 2583 but everything is more convenient
meillo@217 2584 if the deleted messages are collected in one place.
meillo@131 2585 .P
meillo@131 2586 The profile entry
meillo@131 2587 .Pe rmmproc
meillo@131 2588 (previously named
meillo@131 2589 .Pe Delete-Prog )
meillo@131 2590 was introduced very early to improve the situation.
meillo@164 2591 It could be set to any command, which would be executed to remove
meillo@131 2592 the specified messages.
meillo@217 2593 This had overridden the default action, described above.
meillo@217 2594 Refiling the to-be-removed files to a trash folder was the usual example.
meillo@131 2595 Nmh's man page
meillo@131 2596 .Mp rmm (1)
meillo@131 2597 proposes to set the
meillo@131 2598 .Pe rmmproc
meillo@131 2599 to
meillo@131 2600 .Cl "refile +d
meillo@217 2601 to move messages to the trash folder
meillo@217 2602 .Fn +d
meillo@131 2603 instead of renaming them with the backup prefix.
meillo@217 2604 The man page additionally proposes the expunge command
meillo@131 2605 .Cl "rm `mhpath +d all`
meillo@168 2606 to empty the trash folder.
meillo@131 2607 .P
meillo@217 2608 Removing messages in such a way has advantages:
meillo@217 2609 .LI 1
meillo@131 2610 The mail storage is prevented from being cluttered with removed messages
meillo@131 2611 because they are all collected in one place.
meillo@131 2612 Existing and removed messages are thus separated more strictly.
meillo@217 2613 .LI 2
meillo@131 2614 No backup files are silently overwritten.
meillo@217 2615 .LI 3
meillo@217 2616 Most important, however, removed messages are kept in the MH domain.
meillo@131 2617 Messages in the trash folder can be listed like those in any other folder.
meillo@131 2618 Deleted messages can be displayed like any other messages.
meillo@169 2619 .Pn refile
meillo@169 2620 can restore deleted messages.
meillo@131 2621 All operations on deleted files are still covered by the MH tools.
meillo@131 2622 The trash folder is just like any other folder in the mail storage.
meillo@131 2623 .P
meillo@131 2624 Similar to the draft folder case, I dropped the old backup prefix approach
meillo@131 2625 in favor for replacing it by the better suiting trash folder system.
meillo@131 2626 Hence,
meillo@131 2627 .Pn rmm
meillo@131 2628 calls
meillo@131 2629 .Pn refile
meillo@131 2630 to move the to-be-removed message to the trash folder,
meillo@131 2631 .Fn +trash
meillo@131 2632 by default.
meillo@164 2633 To sweep it clean, the user can use
meillo@131 2634 .Cl "rmm -unlink +trash a" ,
meillo@131 2635 where the
meillo@131 2636 .Sw -unlink
meillo@131 2637 switch causes the files to be unlinked.
meillo@168 2638 .Ci 8edc5aaf86f9f77124664f6801bc6c6cdf258173
meillo@168 2639 .Ci ca0b3e830b86700d9e5e31b1784de2bdcaf58fc5
meillo@131 2640 .P
meillo@217 2641 Dropping the legacy approach and converting to the new approach
meillo@217 2642 completely, simplified the code base.
meillo@131 2643 The relationship between
meillo@131 2644 .Pn rmm
meillo@131 2645 and
meillo@131 2646 .Pn refile
meillo@131 2647 was inverted.
meillo@131 2648 In mmh,
meillo@131 2649 .Pn rmm
meillo@131 2650 invokes
meillo@217 2651 .Pn refile .
meillo@217 2652 That used to be the other way round.
meillo@131 2653 Yet, the relationship is simpler now.
meillo@164 2654 Loops, like described in nmh's man page for
meillo@131 2655 .Mp refile (1),
meillo@164 2656 can no longer occur:
meillo@131 2657 .QS
meillo@131 2658 Since
meillo@131 2659 .Pn refile
meillo@131 2660 uses your
meillo@131 2661 .Pe rmmproc
meillo@131 2662 to delete the message, the
meillo@131 2663 .Pe rmmproc
meillo@131 2664 must NOT call
meillo@131 2665 .Pn refile
meillo@131 2666 without specifying
meillo@131 2667 .Sw -normmproc
meillo@131 2668 or you will create an infinite loop.
meillo@131 2669 .QE
meillo@131 2670 .LP
meillo@131 2671 .Pn rmm
meillo@131 2672 either unlinks a message with
meillo@131 2673 .Fu unlink()
meillo@131 2674 or invokes
meillo@131 2675 .Pn refile
meillo@131 2676 to move it to the trash folder.
meillo@131 2677 .Pn refile
meillo@131 2678 does not invoke any tools.
meillo@131 2679 .P
meillo@136 2680 By generalizing the message removal in the way that it became covered
meillo@136 2681 by the MH concepts made the whole system more powerful.
meillo@131 2682
meillo@131 2683
meillo@131 2684
meillo@131 2685
meillo@131 2686
meillo@133 2687 .H2 "Modern Defaults
meillo@133 2688 .P
meillo@133 2689 Nmh has a bunch of convenience-improving features inactive by default,
meillo@217 2690 although one can expect every new user to want them active.
meillo@133 2691 The reason they are inactive by default is the wish to stay compatible
meillo@133 2692 with old versions.
meillo@217 2693 But what are old versions?
meillo@136 2694 Still, the highly useful draft folder facility has not been activated
meillo@136 2695 by default although it was introduced over twenty-five years ago.
meillo@133 2696 .[
meillo@133 2697 rose romine real work
meillo@133 2698 .]
meillo@136 2699 The community seems not to care.
meillo@217 2700 .P
meillo@217 2701 In nmh, new users are required to first build up
meillo@217 2702 a profile before they can access the modern features.
meillo@136 2703 Without an extensive profile, the setup is hardly usable
meillo@133 2704 for modern emailing.
meillo@133 2705 The point is not the customization of the setup,
meillo@136 2706 but the need to activate generally useful facilities.
meillo@217 2707 Yet, the real problem lies less in enabling the features,
meillo@217 2708 as this is straight forward as soon as one knows what he wants.
meillo@168 2709 The real problem is that new users need deep insight into the project
meillo@217 2710 to discover the available but inactive features.
meillo@133 2711 To give an example, I needed one year of using nmh
meillo@133 2712 before I became aware of the existence of the attachment system.
meillo@133 2713 One could argue that this fact disqualifies my reading of the
meillo@133 2714 documentation.
meillo@133 2715 If I would have installed nmh from source back then, I could agree.
meillo@217 2716 Yet, I had used a pre-packaged version and had expected that it would
meillo@133 2717 just work.
meillo@133 2718 Nevertheless, I had been convinced by the concepts of MH already
meillo@133 2719 and I am a software developer,
meillo@133 2720 still I required a lot of time to discover the cool features.
meillo@133 2721 How can we expect users to be even more advanced than me,
meillo@217 2722 just to enable them to use MH in a convenient and modern way?
meillo@133 2723 Unless they are strongly convinced of the concepts, they will fail.
meillo@133 2724 I have seen friends of me giving up disappointed
meillo@133 2725 before they truly used the system,
meillo@133 2726 although they had been motivated in the beginning.
meillo@217 2727 New users suffer hard enough to get used to the tool chest approach,
meillo@179 2728 we developers should spare them further inconveniences.
meillo@133 2729 .P
meillo@136 2730 Maintaining compatibility for its own sake is bad,
meillo@217 2731 because the code base will collect more and more compatibility code.
meillo@200 2732 Sticking to the compatibility code means remaining limited;
meillo@168 2733 whereas adjusting to the changes renders the compatibility unnecessary.
meillo@217 2734 Keeping unused alternatives in the code for longer than a short
meillo@217 2735 grace time is a bad choice as they likely
meillo@217 2736 gather bugs by not being constantly tested.
meillo@136 2737 Also, the increased code size and the greater number of conditions
meillo@136 2738 increase the maintenance costs.
meillo@133 2739 If any MH implementation would be the back-end of widespread
meillo@133 2740 email clients with large user bases, compatibility would be more
meillo@133 2741 important.
meillo@133 2742 Yet, it appears as if this is not the case.
meillo@133 2743 Hence, compatibility is hardly important for technical reasons.
meillo@217 2744 Its importance originates from personal reasons rather.
meillo@133 2745 Nmh's user base is small and old.
meillo@217 2746 Changing the interfaces causes inconvenience to long-term users of MH.
meillo@217 2747 It forces them to change their many years old MH configurations.
meillo@168 2748 I do understand this aspect, but by sticking to the old users,
meillo@217 2749 new users are kept from entering the world of MH.
meillo@217 2750 But the future lies in new users.
meillo@168 2751 In consequence, mmh invites new users by providing a convenient
meillo@168 2752 and modern setup, readily usable out-of-the-box.
meillo@133 2753 .P
meillo@136 2754 In mmh, all modern features are active by default and many previous
meillo@217 2755 approaches are removed or only accessible in a manual way.
meillo@136 2756 New default features include:
meillo@133 2757 .BU
meillo@133 2758 The attachment system (\c
meillo@217 2759 .Hd Attach )
meillo@217 2760 .Ci 8ff284ff9167eff8f5349481529332d59ed913b1 .
meillo@133 2761 .BU
meillo@133 2762 The draft folder facility (\c
meillo@217 2763 .Fn +drafts )
meillo@217 2764 .Ci 337338b404931f06f0db2119c9e145e8ca5a9860 .
meillo@133 2765 .BU
meillo@133 2766 The unseen sequence (`u')
meillo@133 2767 .Ci c2360569e1d8d3678e294eb7c1354cb8bf7501c1
meillo@217 2768 and the sequence negation prefix (`!')
meillo@217 2769 .Ci db74c2bd004b2dc9bf8086a6d8bf773ac051f3cc .
meillo@133 2770 .BU
meillo@217 2771 Quoting the original message in the reply
meillo@217 2772 .Ci 67411b1f95d6ec987b4c732459e1ba8a8ac192c6 .
meillo@133 2773 .BU
meillo@217 2774 Forwarding messages using MIME
meillo@217 2775 .Ci 6e271608b7b9c23771523f88d23a4d3593010cf1 .
meillo@171 2776 .LP
meillo@217 2777 An mmh setup with a profile that defines only the path to the
meillo@136 2778 mail storage, is already convenient to use.
meillo@217 2779 Again, Paul Vixie's supports the direction I took:
meillo@136 2780 ``the `main branch' should just be modern''.
meillo@136 2781 .[
meillo@136 2782 paul vixie edginess nmh-workers
meillo@136 2783 .]
meillo@131 2784
meillo@133 2785
meillo@133 2786
meillo@133 2787
meillo@133 2788
meillo@133 2789 .\" --------------------------------------------------------------
meillo@131 2790 .H1 "Styling
meillo@22 2791 .P
meillo@118 2792 Kernighan and Pike have emphasized the importance of style in the
meillo@219 2793 preface of \fPThe Practice of Programming\fP:
meillo@118 2794 .[ [
meillo@118 2795 kernighan pike practice of programming
meillo@118 2796 .], p. x]
meillo@118 2797 .QS
meillo@118 2798 Chapter 1 discusses programming style.
meillo@219 2799 Good style is so important to good programming that we have chosen
meillo@118 2800 to cover it first.
meillo@118 2801 .QE
meillo@168 2802 This section covers changes in mmh that were guided by the desire
meillo@118 2803 to improve on style.
meillo@219 2804 Many of them follow the advice given in the quoted book.
meillo@118 2805
meillo@118 2806
meillo@127 2807
meillo@127 2808
meillo@127 2809 .H2 "Code Style
meillo@154 2810 .Id code-style
meillo@118 2811 .P
meillo@118 2812 .U3 "Indentation Style
meillo@118 2813 .P
meillo@219 2814 Indentation styles are the holy cow of programming.
meillo@219 2815 Kernighan and Pike write:
meillo@118 2816 .[ [
meillo@118 2817 kernighan pike practice of programming
meillo@118 2818 .], p. 10]
meillo@118 2819 .QS
meillo@118 2820 Programmers have always argued about the layout of programs,
meillo@118 2821 but the specific style is much less important than its consistent
meillo@118 2822 application.
meillo@121 2823 Pick one style, preferably ours, use it consistently, and don't waste
meillo@118 2824 time arguing.
meillo@118 2825 .QE
meillo@118 2826 .P
meillo@118 2827 I agree that the constant application is most important,
meillo@118 2828 but I believe that some styles have advantages over others.
meillo@118 2829 For instance the indentation with tab characters only.
meillo@219 2830 The number of tabs corresponds to the nesting level \(en
meillo@118 2831 one tab, one level.
meillo@219 2832 Tab characters provide flexible visual appearance because developers
meillo@228 2833 can adjust their width as preferred.
meillo@219 2834 There is no more need to check for the correct mixture of
meillo@219 2835 tabs and spaces.
meillo@219 2836 Two simple rules ensure the integrity and flexibility of the visual
meillo@219 2837 appearance:
meillo@219 2838 .LI 1
meillo@219 2839 Leading whitespace must consist of tabs only.
meillo@219 2840 .LI 2
meillo@219 2841 All other whitespace should be spaces.
meillo@219 2842 .LP
meillo@121 2843 Although reformatting existing code should be avoided, I did it.
meillo@200 2844 I did not waste time arguing; I just reformatted the code.
meillo@118 2845 .Ci a485ed478abbd599d8c9aab48934e7a26733ecb1
meillo@118 2846
meillo@118 2847 .U3 "Comments
meillo@118 2848 .P
meillo@219 2849 Kernighan and Pike demand: ``Don't belabor the obvious''.
meillo@118 2850 .[ [
meillo@118 2851 kernighan pike practice of programming
meillo@118 2852 .], p. 23]
meillo@219 2853 Following the advice, I removed unnecessary comments.
meillo@219 2854 For instance, I removed all comments in the following code excerpt
meillo@219 2855 .Ci 426543622b377fc5d091455cba685e114b6df674 :
meillo@118 2856 .VS
meillo@120 2857 context_replace(curfolder, folder); /* update current folder */
meillo@120 2858 seq_setcur(mp, mp->lowsel); /* update current message */
meillo@120 2859 seq_save(mp); /* synchronize message sequences */
meillo@120 2860 folder_free(mp); /* free folder/message structure */
meillo@120 2861 context_save(); /* save the context file */
meillo@120 2862
meillo@120 2863 [...]
meillo@120 2864
meillo@120 2865 int c; /* current character */
meillo@120 2866 char *cp; /* miscellaneous character pointer */
meillo@120 2867
meillo@120 2868 [...]
meillo@120 2869
meillo@120 2870 /* NUL-terminate the field */
meillo@120 2871 *cp = '\0';
meillo@118 2872 VE
meillo@118 2873 .P
meillo@219 2874 The information in each of the comments was present in the code
meillo@219 2875 statements already, except for the NUL-termination, which became
meillo@219 2876 obvious from the context.
meillo@136 2877
meillo@118 2878
meillo@118 2879 .U3 "Names
meillo@118 2880 .P
meillo@219 2881 Regarding this topic, Kernighan and Pike suggest:
meillo@118 2882 ``Use active names for functions''.
meillo@118 2883 .[ [
meillo@118 2884 kernighan pike practice of programming
meillo@118 2885 .], p. 4]
meillo@118 2886 One application of this rule was the rename of
meillo@118 2887 .Fu check_charset()
meillo@118 2888 to
meillo@219 2889 .Fu is_native_charset()
meillo@219 2890 .Ci 8d77b48284c58c135a6b2787e721597346ab056d .
meillo@219 2891 The same change additionally fixed a violation of ``Be accurate'',
meillo@181 2892 .[ [
meillo@181 2893 kernighan pike practice of programming
meillo@181 2894 .], p. 4]
meillo@219 2895 as the code did not match the expectation the function suggested.
meillo@219 2896 It did not compare charset names but prefixes of them only.
meillo@219 2897 In case the native charset was `ISO-8859-1', then
meillo@219 2898 .VS
meillo@219 2899 check_charset("ISO-8859-11", strlen("ISO-8859-11"))
meillo@219 2900 VE
meillo@219 2901 had returned true although the upper halves of the code pages
meillo@219 2902 are different.
meillo@118 2903 .P
meillo@118 2904 More important than using active names is using descriptive names.
meillo@145 2905 .VS
meillo@145 2906 m_unknown(in); /* the MAGIC invocation... */
meillo@145 2907 VE
meillo@145 2908 Renaming the obscure
meillo@118 2909 .Fu m_unknown()
meillo@219 2910 function was a delightful event, although it made the code less funny
meillo@219 2911 .Ci 611d68d19204d7cbf5bd585391249cb5bafca846 .
meillo@118 2912 .P
meillo@118 2913 Magic numbers are generally considered bad style.
meillo@118 2914 Obviously, Kernighan and Pike agree:
meillo@118 2915 ``Give names to magic numbers''.
meillo@118 2916 .[ [
meillo@118 2917 kernighan pike practice of programming
meillo@118 2918 .], p. 19]
meillo@219 2919 .P
meillo@219 2920 The argument
meillo@219 2921 .CW outnum
meillo@219 2922 of the function
meillo@219 2923 .Fu scan()
meillo@219 2924 in
meillo@219 2925 .Fn uip/scansbr.c
meillo@219 2926 holds the number of the message to be created.
meillo@219 2927 As well it encodes program logic with negative numbers and zero.
meillo@219 2928 This led to obscure code.
meillo@219 2929 I clarified the code by introducing two variables that extracted
meillo@219 2930 the hidden information:
meillo@219 2931 .VS
meillo@219 2932 int incing = (outnum > 0);
meillo@219 2933 int ismbox = (outnum != 0);
meillo@219 2934 VE
meillo@219 2935 The readable names are thus used in conditions;
meillo@219 2936 the variable
meillo@219 2937 .CW outnum
meillo@219 2938 is used only to extract ordinary message numbers
meillo@219 2939 .Ci b8b075c77be7794f3ae9ff0e8cedb12b48fd139f .
meillo@219 2940 .P
meillo@219 2941 Through the clarity improvement of the change detours in the program
meillo@219 2942 logic of related code parts became apparent.
meillo@219 2943 The implementation was simplified.
meillo@219 2944 This possibility to improve had been invisible before
meillo@219 2945 .Ci aa60b0ab5e804f8befa890c0a6df0e3143ce0723 .
meillo@219 2946 .P
meillo@219 2947 The names just described were a first step, yet the situation
meillo@219 2948 was further improved by giving names to the magic values of
meillo@219 2949 .CW outnum :
meillo@118 2950 .VS
meillo@118 2951 #define SCN_MBOX (-1)
meillo@118 2952 #define SCN_FOLD 0
meillo@118 2953 VE
meillo@219 2954 The two variables were updated thereafter as well:
meillo@219 2955 .VS
meillo@219 2956 int incing = (outnum != SCN_MBOX && outnum != SCN_FOLD);
meillo@219 2957 int scanfolder = (outnum == SCN_FOLD);
meillo@219 2958 VE
meillo@219 2959 Furthermore,
meillo@219 2960 .CW ismbox
meillo@219 2961 was replaced by
meillo@219 2962 .CW scanfolder
meillo@219 2963 because that matched better to the program logic.
meillo@118 2964 .Ci 7ffb36d28e517a6f3a10272056fc127592ab1c19
meillo@219 2965
meillo@118 2966
meillo@133 2967
meillo@133 2968
meillo@133 2969 .H2 "Structural Rework
meillo@133 2970 .P
meillo@219 2971 Although the stylistic changes described already improve the
meillo@219 2972 readability of the source code, all of them were changes ``in the small''.
meillo@219 2973 Structural changes, in contrast, affect much larger code areas.
meillo@219 2974 They are more difficult to accomplish but lead to larger improvements,
meillo@219 2975 especially as they often influence the outer shape of the tools as well.
meillo@118 2976 .P
meillo@118 2977 At the end of their chapter on style,
meillo@118 2978 Kernighan and Pike ask: ``But why worry about style?''
meillo@181 2979 .[ [
meillo@181 2980 kernighan pike practice of programming
meillo@219 2981 .], p. 28].
meillo@219 2982 Following are two examples of structural rework that demonstrate
meillo@136 2983 why style is important in the first place.
meillo@136 2984
meillo@136 2985
meillo@136 2986 .U3 "Rework of \f(CWanno\fP
meillo@118 2987 .P
meillo@120 2988 Until 2002,
meillo@120 2989 .Pn anno
meillo@219 2990 had six functional command line switches:
meillo@120 2991 .Sw -component
meillo@120 2992 and
meillo@120 2993 .Sw -text ,
meillo@219 2994 each with an argument,
meillo@120 2995 and the two pairs of flags,
meillo@120 2996 .Sw -[no]date
meillo@120 2997 and
meillo@120 2998 .Sw -[no]inplace .
meillo@120 2999 Then Jon Steinhart introduced his attachment system.
meillo@120 3000 In need for more advanced annotation handling, he extended
meillo@120 3001 .Pn anno .
meillo@120 3002 He added five more switches:
meillo@120 3003 .Sw -draft ,
meillo@120 3004 .Sw -list ,
meillo@120 3005 .Sw -delete ,
meillo@120 3006 .Sw -append ,
meillo@120 3007 and
meillo@120 3008 .Sw -number ,
meillo@219 3009 the last one taking an argument
meillo@219 3010 .Ci 7480dbc14bc90f2d872d434205c0784704213252 .
meillo@120 3011 Later,
meillo@120 3012 .Sw -[no]preserve
meillo@219 3013 was added as well
meillo@219 3014 .Ci d9b1d57351d104d7ec1a5621f090657dcce8cb7f .
meillo@120 3015 Then, the Synopsis section of the man page
meillo@120 3016 .Mp anno (1)
meillo@120 3017 read:
meillo@120 3018 .VS
meillo@219 3019 anno [+folder] [msgs] [-component f(CIfieldfP] [-inplace | -noinplace]
meillo@120 3020 [-date | -nodate] [-draft] [-append] [-list] [-delete]
meillo@219 3021 [-number [f(CInumfP|fPallfP]] [-preserve | -nopreserve] [-version]
meillo@219 3022 [-help] [-text f(CIbodyfP]
meillo@120 3023 VE
meillo@120 3024 .LP
meillo@120 3025 The implementation followed the same structure.
meillo@120 3026 Problems became visible when
meillo@120 3027 .Cl "anno -list -number 42
meillo@219 3028 worked on the current message instead of on message number 42,
meillo@120 3029 and
meillo@120 3030 .Cl "anno -list -number l:5
meillo@124 3031 did not work on the last five messages but failed with the mysterious
meillo@120 3032 error message: ``anno: missing argument to -list''.
meillo@121 3033 Yet, the invocation matched the specification in the man page.
meillo@120 3034 There, the correct use of
meillo@120 3035 .Sw -number
meillo@120 3036 was defined as being
meillo@120 3037 .Cl "[-number [num|all]]
meillo@120 3038 and the textual description for the combination with
meillo@120 3039 .Sw -list
meillo@120 3040 read:
meillo@120 3041 .QS
meillo@164 3042 The
meillo@164 3043 .Sw -list
meillo@164 3044 option produces a listing of the field bodies for
meillo@120 3045 header fields with names matching the specified component,
meillo@164 3046 one per line. The listing is numbered, starting at 1, if the
meillo@164 3047 .Sw -number
meillo@164 3048 option is also used.
meillo@120 3049 .QE
meillo@120 3050 .LP
meillo@120 3051 The problem was manifold.
meillo@120 3052 Semantically, the argument to the
meillo@120 3053 .Sw -number
meillo@120 3054 switch is only necessary in combination with
meillo@120 3055 .Sw -delete ,
meillo@120 3056 but not with
meillo@120 3057 .Sw -list .
meillo@219 3058 The code, however, required a numeric argument in any case.
meillo@219 3059 If the argument was missing or non-numeric,
meillo@219 3060 .Pn anno
meillo@219 3061 aborted with an error message that additionally had an off-by-one error.
meillo@219 3062 It printed the name of the switch one before the concerned one.
meillo@120 3063 .P
meillo@219 3064 Trying to fix these problems on the surface would not have solved them.
meillo@219 3065 They originate from a discrepance between the
meillo@120 3066 structure of the problem and the structure implemented in the program.
meillo@219 3067 Such structural differences can only be solved by adjusting the
meillo@219 3068 structure of the implementation to the structure of the problem.
meillo@120 3069 .P
meillo@219 3070 Steinhart had added the new
meillo@120 3071 .Sw -list
meillo@120 3072 and
meillo@120 3073 .Sw -delete
meillo@219 3074 switches in a style similar to the other switches though
meillo@219 3075 they are of structural different type.
meillo@120 3076 Semantically,
meillo@120 3077 .Sw -list
meillo@120 3078 and
meillo@120 3079 .Sw -delete
meillo@219 3080 introduce operation modes.
meillo@120 3081 Historically,
meillo@120 3082 .Pn anno
meillo@120 3083 had only one operation mode: adding header fields.
meillo@219 3084 With the extension, two more modes were added:
meillo@120 3085 listing and deleting header fields.
meillo@120 3086 The structure of the code changes did not pay respect to this
meillo@219 3087 fundamental change.
meillo@120 3088 Neither the implementation nor the documentation did clearly
meillo@219 3089 declare the exclusive operation modes as such.
meillo@120 3090 Having identified the problem, I solved it by putting structure into
meillo@120 3091 .Pn anno
meillo@219 3092 and its documentation
meillo@219 3093 .Ci d54c8db8bdf01e8381890f7729bc0ef4a055ea11 .
meillo@120 3094 .P
meillo@173 3095 The difference is visible in both the code and the documentation.
meillo@219 3096 For instance in the following code excerpt:
meillo@120 3097 .VS
meillo@120 3098 int delete = -2; /* delete header element if set */
meillo@120 3099 int list = 0; /* list header elements if set */
meillo@120 3100 [...]
meillo@121 3101 case DELETESW: /* delete annotations */
meillo@121 3102 delete = 0;
meillo@121 3103 continue;
meillo@121 3104 case LISTSW: /* produce a listing */
meillo@121 3105 list = 1;
meillo@121 3106 continue;
meillo@120 3107 VE
meillo@121 3108 .LP
meillo@219 3109 which was replaced by:
meillo@120 3110 .VS
meillo@120 3111 static enum { MODE_ADD, MODE_DEL, MODE_LIST } mode = MODE_ADD;
meillo@120 3112 [...]
meillo@121 3113 case DELETESW: /* delete annotations */
meillo@121 3114 mode = MODE_DEL;
meillo@121 3115 continue;
meillo@121 3116 case LISTSW: /* produce a listing */
meillo@121 3117 mode = MODE_LIST;
meillo@121 3118 continue;
meillo@120 3119 VE
meillo@120 3120 .LP
meillo@121 3121 The replacement code does not only reflect the problem's structure better,
meillo@121 3122 it is easier to understand as well.
meillo@121 3123 The same applies to the documentation.
meillo@120 3124 The man page was completely reorganized to propagate the same structure.
meillo@219 3125 This is already visible in the Synopsis section:
meillo@120 3126 .VS
meillo@219 3127 anno [+folder] [msgs] [-component f(CIfieldfP] [-text fPbodyfP]
meillo@120 3128 [-append] [-date | -nodate] [-preserve | -nopreserve]
meillo@120 3129 [-Version] [-help]
meillo@120 3130
meillo@219 3131 anno -delete [+folder] [msgs] [-component fPfieldfP] [-text
meillo@219 3132 fPbodyfP] [-number fPnum fP| fPall fP] [-preserve | -nopreserve]
meillo@120 3133 [-Version] [-help]
meillo@120 3134
meillo@219 3135 anno -list [+folder] [msgs] [-component fPfieldfP] [-number]
meillo@120 3136 [-Version] [-help]
meillo@120 3137 VE
meillo@118 3138
meillo@58 3139
meillo@58 3140
meillo@133 3141 .U3 "Path Conversion
meillo@133 3142 .P
meillo@134 3143 Four kinds of path names can appear in MH:
meillo@171 3144 .LI 1
meillo@134 3145 Absolute Unix directory paths, like
meillo@134 3146 .Fn /etc/passwd .
meillo@171 3147 .LI 2
meillo@134 3148 Relative Unix directory paths, like
meillo@134 3149 .Fn ./foo/bar .
meillo@171 3150 .LI 3
meillo@134 3151 Absolute MH folder paths, like
meillo@219 3152 .Fn +projects/mmh .
meillo@171 3153 .LI 4
meillo@134 3154 Relative MH folder paths, like
meillo@134 3155 .Fn @subfolder .
meillo@171 3156 .LP
meillo@219 3157 Relative MH folder paths, are hardly documented
meillo@219 3158 although they are useful for large mail storages.
meillo@134 3159 The current mail folder is specified as `\c
meillo@134 3160 .Fn @ ',
meillo@134 3161 just like the current directory is specified as `\c
meillo@134 3162 .Fn . '.
meillo@134 3163 .P
meillo@134 3164 To allow MH tools to understand all four notations,
meillo@219 3165 they need to be able to convert between them.
meillo@134 3166 In nmh, these path name conversion functions were located in the files
meillo@134 3167 .Fn sbr/path.c
meillo@134 3168 (``return a pathname'') and
meillo@134 3169 .Fn sbr/m_maildir.c
meillo@134 3170 (``get the path for the mail directory'').
meillo@134 3171 The seven functions in the two files were documented with no more
meillo@134 3172 than two comments, which described obvious information.
meillo@219 3173 The signatures of the four exported functions did not explain their
meillo@219 3174 semantics:
meillo@219 3175 .LI 1
meillo@219 3176 .CW "char *path(char *, int);
meillo@219 3177 .LI 2
meillo@219 3178 .CW "char *pluspath(char *);
meillo@219 3179 .LI 3
meillo@219 3180 .CW "char *m_mailpath(char *);
meillo@219 3181 .LI 4
meillo@219 3182 .CW "char *m_maildir(char *);
meillo@134 3183 .P
meillo@219 3184 My investigations provided the following descriptions:
meillo@171 3185 .LI 1
meillo@134 3186 The second parameter of
meillo@134 3187 .Fu path()
meillo@219 3188 defines the type as which the path given in the first parameter should
meillo@219 3189 be treated.
meillo@134 3190 Directory paths are converted to absolute directory paths.
meillo@134 3191 Folder paths are converted to absolute folder paths.
meillo@173 3192 Folder paths must not include a leading `\fL@\fP' character.
meillo@134 3193 Leading plus characters are preserved.
meillo@134 3194 The result is a pointer to newly allocated memory.
meillo@171 3195 .LI 2
meillo@134 3196 .Fu pluspath()
meillo@134 3197 is a convenience-wrapper to
meillo@134 3198 .Fu path() ,
meillo@134 3199 to convert folder paths only.
meillo@134 3200 This function can not be used for directory paths.
meillo@134 3201 An empty string parameter causes a buffer overflow.
meillo@171 3202 .LI 3
meillo@134 3203 .Fu m_mailpath()
meillo@134 3204 converts directory paths to absolute directory paths.
meillo@173 3205 The characters `\fL+\fP' or `\fL@\fP' at the beginning of the path name are
meillo@134 3206 treated literal, i.e. as the first character of a relative directory path.
meillo@134 3207 Hence, this function can not be used for folder paths.
meillo@219 3208 In any case, the result is an absolute directory path,
meillo@219 3209 returned as a pointer to newly allocated memory.
meillo@171 3210 .LI 4
meillo@134 3211 .Fu m_maildir()
meillo@134 3212 returns the parameter unchanged if it is an absolute directory path
meillo@173 3213 or begins with the entry `\fL.\fP' or `\fL..\fP'.
meillo@134 3214 All other strings are prepended with the current working directory.
meillo@219 3215 Hence, this function can not be used for folder paths.
meillo@134 3216 The result is either an absolute directory path or a relative
meillo@219 3217 directory path, starting with dot or dot-dot.
meillo@134 3218 In contrast to the other functions, the result is a pointer to
meillo@134 3219 static memory.
meillo@134 3220 .P
meillo@134 3221 The situation was obscure, irritating, error-prone, and non-orthogonal.
meillo@219 3222 Additionally, no clear terminology was used to name the different
meillo@219 3223 kinds of path names.
meillo@219 3224 Sometimes, the names were even misleading, much as the first argument of
meillo@134 3225 .Fu m_mailpath() ,
meillo@219 3226 which was named
meillo@219 3227 .CW folder ,
meillo@219 3228 although
meillo@134 3229 .Fu m_mailpath()
meillo@219 3230 could not be used with MH folder arguments.
meillo@134 3231 .P
meillo@219 3232 I clarified the path name conversion by complete rework.
meillo@134 3233 First of all, the terminology needed to be defined.
meillo@134 3234 A path name is either in the Unix domain, then it is called
meillo@227 3235 \fIdirectory path\fP or it is in the MH domain,
meillo@227 3236 then it is called \fIfolder path\fP.
meillo@134 3237 The two terms need to be used with strict distinction.
meillo@134 3238 Second, I exploited the concept of path type indicators.
meillo@219 3239 By requiring every path name to start with a distinct type identifier,
meillo@219 3240 the conversion between the types could be fully automated.
meillo@227 3241 This allows the tools to accept path names of any type from the user.
meillo@134 3242 Therefore, it was necessary to require relative directory paths to be
meillo@134 3243 prefixed with a dot character.
meillo@134 3244 In consequence, the dot character could no longer be an alias for the
meillo@227 3245 current message
meillo@227 3246 .Ci cff0e16925e7edbd25b8b9d6d4fbdf03e0e60c01 .
meillo@134 3247 Third, I created three new functions to replace the previous mess:
meillo@171 3248 .LI 1
meillo@134 3249 .Fu expandfol()
meillo@219 3250 converts folder paths to absolute folder paths.
meillo@134 3251 Directory paths are simply passed through.
meillo@134 3252 This function is to be used for folder paths only, thus the name.
meillo@134 3253 The result is a pointer to static memory.
meillo@171 3254 .LI 2
meillo@134 3255 .Fu expanddir()
meillo@134 3256 converts directory paths to absolute directory paths.
meillo@134 3257 Folder paths are treated as relative directory paths.
meillo@134 3258 This function is to be used for directory paths only, thus the name.
meillo@134 3259 The result is a pointer to static memory.
meillo@171 3260 .LI 3
meillo@134 3261 .Fu toabsdir()
meillo@134 3262 converts any type of path to an absolute directory path.
meillo@134 3263 This is the function of choice for path conversion.
meillo@134 3264 Absolute directory paths are the most general representation of a
meillo@134 3265 path name.
meillo@134 3266 The result is a pointer to static memory.
meillo@134 3267 .P
meillo@134 3268 The new functions have names that indicate their use.
meillo@134 3269 Two of the functions convert relative to absolute path names of the
meillo@134 3270 same type.
meillo@134 3271 The third function converts any path name type to the most general one,
meillo@134 3272 the absolute directory path.
meillo@134 3273 All of the functions return pointers to static memory.
meillo@219 3274 The file
meillo@219 3275 .Fn sbr/path.c
meillo@219 3276 contains the implementation of the functions;
meillo@134 3277 .Fn sbr/m_maildir.c
meillo@219 3278 was removed.
meillo@168 3279 .Ci d39e2c447b0d163a5a63f480b23d06edb7a73aa0
meillo@134 3280 .P
meillo@134 3281 Along with the path conversion rework, I also replaced
meillo@134 3282 .Fu getfolder(FDEF)
meillo@134 3283 with
meillo@134 3284 .Fu getdeffol()
meillo@134 3285 and
meillo@134 3286 .Fu getfolder(FCUR)
meillo@134 3287 with
meillo@134 3288 .Fu getcurfol() ,
meillo@219 3289 which only wraps
meillo@219 3290 .Fu expandfol(""@"")
meillo@219 3291 for convenience.
meillo@134 3292 This code was moved from
meillo@134 3293 .Fn sbr/getfolder.c
meillo@219 3294 into
meillo@219 3295 .Fn sbr/path.c
meillo@219 3296 as well.
meillo@168 3297 .Ci d39e2c447b0d163a5a63f480b23d06edb7a73aa0
meillo@134 3298 .P
meillo@134 3299 The related function
meillo@134 3300 .Fu etcpath()
meillo@219 3301 is now included in
meillo@134 3302 .Fn sbr/path.c ,
meillo@168 3303 too
meillo@168 3304 .Ci b4c29794c12099556151d93a860ee51badae2e35 .
meillo@134 3305 Previously, it had been located in
meillo@219 3306 .Fn config/config.c .
meillo@134 3307 .P
meillo@219 3308 Now,
meillo@134 3309 .Fn sbr/path.c
meillo@219 3310 contains all path handling code.
meillo@219 3311 Besides being less code, its readability is highly improved.
meillo@219 3312 The functions follow a common style and are well documented.
meillo@133 3313
meillo@133 3314
meillo@133 3315
meillo@133 3316
meillo@133 3317 .H2 "Profile Reading
meillo@133 3318 .P
meillo@220 3319 The MH profile contains the configuration of a user-specific MH setup.
meillo@220 3320 MH tools read the profile right after starting up
meillo@220 3321 because it contains the location of the user's mail storage
meillo@138 3322 and similar settings that influence the whole setup.
meillo@220 3323 Furthermore, the profile contains the default switches for the tools
meillo@220 3324 as well.
meillo@220 3325 The context file is read along with the profile.
meillo@138 3326 .P
meillo@138 3327 For historic reasons, some MH tools did not read the profile and context.
meillo@138 3328 Among them were
meillo@138 3329 .Pn post /\c
meillo@138 3330 .Pn spost ,
meillo@138 3331 .Pn mhmail ,
meillo@138 3332 and
meillo@138 3333 .Pn slocal .
meillo@138 3334 The reason why these tools ignored the profile were not clearly stated.
meillo@220 3335 During a discussion on the nmh-workers mailing list,
meillo@181 3336 David Levine posted an explanation, quoting John Romine:
meillo@138 3337 .[
meillo@138 3338 nmh-workers levine post profile
meillo@138 3339 .]
meillo@220 3340
meillo@138 3341 .QS
meillo@138 3342 I asked John Romine and here's what he had to say, which
meillo@138 3343 agrees and provides an example that convinces me:
meillo@138 3344 .QS
meillo@164 3345 My take on this is that
meillo@164 3346 .Pn post
meillo@164 3347 should not be called by users directly, and it doesn't read the
meillo@164 3348 .Fn .mh_profile
meillo@138 3349 (only front-end UI programs read the profile).
meillo@138 3350 .QP
meillo@164 3351 For example, there can be contexts where
meillo@164 3352 .Pn post
meillo@164 3353 is called by a helper program (like `\c
meillo@164 3354 .Pn mhmail ')
meillo@164 3355 which may be run by a non-MH user.
meillo@164 3356 We don't want this to prompt the user to create an MH profile, etc.
meillo@138 3357 .QP
meillo@164 3358 My suggestion would be to have
meillo@164 3359 .Pn send
meillo@164 3360 pass a (hidden) `\c
meillo@164 3361 .Sw -fileproc
meillo@164 3362 .Ar proc '
meillo@164 3363 option to
meillo@164 3364 .Pn post
meillo@164 3365 if needed.
meillo@164 3366 You could also
meillo@164 3367 use an environment variable (I think
meillo@164 3368 .Pn send /\c
meillo@164 3369 .Pn whatnow
meillo@164 3370 do this).
meillo@138 3371 .QE
meillo@220 3372 .sp \n(PDu
meillo@164 3373 I think that's the way to go.
meillo@164 3374 My personal preference is to use a command line option,
meillo@164 3375 not an environment variable.
meillo@138 3376 .QE
meillo@220 3377
meillo@138 3378 .P
meillo@220 3379 To solve the problem that
meillo@138 3380 .Pn post
meillo@220 3381 does not honor the
meillo@138 3382 .Pe fileproc
meillo@138 3383 profile entry,
meillo@138 3384 the community roughly agreed that a switch
meillo@138 3385 .Sw -fileproc
meillo@138 3386 should be added to
meillo@138 3387 .Pn post
meillo@138 3388 to be able to pass a different fileproc.
meillo@138 3389 I strongly disagree with this approach because it does not solve
meillo@138 3390 the problem; it only removes a single symptom.
meillo@220 3391 The actual problem is that
meillo@138 3392 .Pn post
meillo@220 3393 does not behave as expected,
meillo@220 3394 though all programs should behave as expected.
meillo@220 3395 Clear and general concepts are a precondition for this.
meillo@220 3396 Thus, there should be no separation into ``front-end UI programs''
meillo@220 3397 and ones that ``should not be called by users directly''.
meillo@220 3398 The real solution is having all MH tools read the profile.
meillo@138 3399 .P
meillo@220 3400 But the problem has a further aspect,
meillo@220 3401 which originates from
meillo@220 3402 .Pn mhmail
meillo@220 3403 mainly.
meillo@138 3404 .Pn mhmail
meillo@138 3405 was intended to be a replacement for
meillo@138 3406 .Pn mailx
meillo@138 3407 on systems with MH installations.
meillo@220 3408 In difference to
meillo@220 3409 .Pn mailx ,
meillo@138 3410 .Pn mhmail
meillo@220 3411 used MH's
meillo@138 3412 .Pn post
meillo@220 3413 to send the message.
meillo@220 3414 The idea was that using
meillo@138 3415 .Pn mhmail
meillo@220 3416 should not be influenced whether the user had
meillo@138 3417 MH set up for himself or not.
meillo@220 3418 Therefore
meillo@138 3419 .Pn mhmail
meillo@220 3420 had not read the profile.
meillo@138 3421 As
meillo@138 3422 .Pn mhmail
meillo@138 3423 used
meillo@138 3424 .Pn post ,
meillo@138 3425 .Pn post
meillo@220 3426 was not allowed to read the profile neither.
meillo@138 3427 This is the reason for the actual problem.
meillo@220 3428 Yet, this was not considered much of a problem because
meillo@138 3429 .Pn post
meillo@138 3430 was not intended to be used by users directly.
meillo@220 3431 To invoke
meillo@220 3432 .Pn post ,
meillo@138 3433 .Pn send
meillo@220 3434 was used an a front-end.
meillo@138 3435 .Pn send
meillo@138 3436 read the profile and passed all relevant values on the command line to
meillo@138 3437 .Pn post
meillo@138 3438 \(en an awkward solution.
meillo@138 3439 .P
meillo@138 3440 The important insight is that
meillo@138 3441 .Pn mhmail
meillo@220 3442 is a wolf in sheep's clothing.
meillo@220 3443 This alien tool broke the concepts because it was treated like
meillo@220 3444 a normal MH tool.
meillo@138 3445 Instead it should have been treated accordingly to its foreign style.
meillo@220 3446 .P
meillo@220 3447 The solution is not to prevent the tools from reading the profile but
meillo@220 3448 to instruct them to read a different profile.
meillo@138 3449 .Pn mhmail
meillo@220 3450 could have set up a well-defined profile and caused the following
meillo@138 3451 .Pn post
meillo@220 3452 to use this profile by exporting an environment variable.
meillo@220 3453 With this approach, no special cases would have been introduced
meillo@220 3454 and no surprises would have been caused.
meillo@220 3455 By writing a wrapper program to provide a clean temporary profile,
meillo@220 3456 the concept could have been generalized orthogonally to the whole
meillo@220 3457 MH tool chest.
meillo@220 3458 .P
meillo@220 3459 In mmh, the wish to have
meillo@220 3460 .Pn mhmail
meillo@220 3461 as a replacement for
meillo@220 3462 .Pn mailx
meillo@220 3463 is considered obsolete.
meillo@220 3464 Mmh's
meillo@220 3465 .Pn mhmail
meillo@220 3466 does no longer cover this use-case
meillo@220 3467 .Ci d36e56e695fe1c482c7920644bfbb6386ac9edb0 .
meillo@220 3468 Currently,
meillo@220 3469 .Pn mhmail
meillo@220 3470 is in a transition state
meillo@220 3471 .Ci 32d4f9daaa70519be3072479232ff7be0500d009 .
meillo@220 3472 It may become a front-end to
meillo@220 3473 .Pn comp ,
meillo@220 3474 which provides an alternative interface which can be more convenient
meillo@220 3475 in some cases.
meillo@220 3476 This would convert
meillo@220 3477 .Pn mhmail
meillo@220 3478 into an ordinary MH tool.
meillo@220 3479 If, however, this idea does not convince, then
meillo@220 3480 .Pn mhmail
meillo@220 3481 will be removed.
meillo@220 3482 .P
meillo@220 3483 In the mmh tool chest, every program reads the profile.
meillo@220 3484 (\c
meillo@220 3485 .Pn slocal
meillo@220 3486 is not considered part of the mmh tool chest (cf. Sec.
meillo@220 3487 .Cf slocal ).)
meillo@220 3488 Mmh has no
meillo@220 3489 .Pn post
meillo@220 3490 program, but it has
meillo@220 3491 .Pn spost ,
meillo@220 3492 which now does read the profile
meillo@220 3493 .Ci 3e017a7abbdf69bf0dff7a4073275961eda1ded8 .
meillo@220 3494 Following this change,
meillo@220 3495 .Pn send
meillo@220 3496 and
meillo@220 3497 .Pn spost
meillo@220 3498 can be considered for merging.
meillo@220 3499 Besides
meillo@220 3500 .Pn send ,
meillo@220 3501 .Pn spost
meillo@220 3502 is only invoked directly by the to-be-changed
meillo@220 3503 .Pn mhmail
meillo@220 3504 implementation and by
meillo@220 3505 .Pn rcvdist ,
meillo@220 3506 which requires rework anyway.
meillo@220 3507
meillo@220 3508 .P
meillo@220 3509 Jeffrey Honig quoted Marshall T. Rose explaining the decision that
meillo@220 3510 .Pn post
meillo@220 3511 ignores the profile:
meillo@138 3512 .[
meillo@197 3513 nmh-workers honig post profile
meillo@138 3514 .]
meillo@138 3515 .QS
meillo@138 3516 when you run mh commands in a script, you want all the defaults to be
meillo@138 3517 what the man page says.
meillo@138 3518 when you run a command by hand, then you want your own defaults...
meillo@138 3519 .QE
meillo@138 3520 .LP
meillo@228 3521 The explanation neither matches the problem concerned exactly
meillo@220 3522 nor is the interpretation clear.
meillo@220 3523 If the described desire addresses the technical level,
meillo@228 3524 then it conflicts fundamentally with the Unix philosophy,
meillo@220 3525 precisely because the indistinquishability of human and script
meillo@220 3526 input is the main reason for the huge software leverage in Unix.
meillo@220 3527 If, however, the described desire addresses the user's view,
meillo@220 3528 then different technical solutions are more appropriate.
meillo@220 3529 The two cases can be regarded simply as two different MH setups.
meillo@220 3530 Hence, mapping the problem of different behavior between interactive and
meillo@220 3531 automated use on the concept of switching between different profiles,
meillo@220 3532 marks it already solved.
meillo@133 3533
meillo@133 3534
meillo@127 3535
meillo@121 3536 .H2 "Standard Libraries
meillo@22 3537 .P
meillo@121 3538 MH is one decade older than the POSIX and ANSI C standards.
meillo@121 3539 Hence, MH included own implementations of functions
meillo@220 3540 that were neither standardized nor widely available, back then.
meillo@220 3541 Today, twenty years after POSIX and ANSI C were published,
meillo@220 3542 developers can expect that systems comply with these standards.
meillo@121 3543 In consequence, MH-specific replacements for standard functions
meillo@121 3544 can and should be dropped.
meillo@220 3545 Kernighan and Pike advise: ``Use standard libraries''.
meillo@121 3546 .[ [
meillo@121 3547 kernighan pike practice of programming
meillo@121 3548 .], p. 196]
meillo@121 3549 Actually, MH had followed this advice in history,
meillo@220 3550 but it had not adjusted to more recent changes in this field.
meillo@121 3551 The
meillo@121 3552 .Fu snprintf()
meillo@121 3553 function, for instance, was standardized with C99 and is available
meillo@121 3554 almost everywhere because of its high usefulness.
meillo@220 3555 Thus, the project's own implementation of
meillo@121 3556 .Fu snprintf()
meillo@123 3557 was dropped in March 2012 in favor for using the one of the
meillo@220 3558 standard library
meillo@220 3559 .Ci 0052f1024deb0a0a2fc2e5bacf93d45a5a9c9b32 .
meillo@123 3560 Such decisions limit the portability of mmh
meillo@173 3561 if systems do not support these standardized and widespread functions.
meillo@123 3562 This compromise is made because mmh focuses on the future.
meillo@121 3563 .P
meillo@220 3564 As I am still in my twenties, have no programming experience from
meillo@220 3565 past decades.
meillo@220 3566 I have not followed the evolution of C through time.
meillo@220 3567 I have not suffered from the the Unix wars.
meillo@121 3568 I have not longed for standardization.
meillo@121 3569 All my programming experience is from a time when ANSI C and POSIX
meillo@121 3570 were well established already.
meillo@220 3571 Thus, I needed to learn about the history in retrospective.
meillo@121 3572 I have only read a lot of books about the (good) old times.
meillo@220 3573 This put me in a difficult position when working with old code.
meillo@123 3574 I need to freshly acquire knowledge about old code constructs and ancient
meillo@123 3575 programming styles, whereas older programmers know these things by
meillo@123 3576 heart from their own experience.
meillo@123 3577 Being aware of the situation, I rather let people with more historic
meillo@220 3578 experience do the transition from ancient code constructs to
meillo@220 3579 standardized ones.
meillo@121 3580 Lyndon Nerenberg covered large parts of this task for the nmh project.
meillo@121 3581 He converted project-specific functions to POSIX replacements,
meillo@121 3582 also removing the conditionals compilation of now standardized features.
meillo@220 3583 Ken Hornstein and David Levine had their part in this work, as well.
meillo@220 3584 Often, I only pulled the changes over from nmh into mmh.
meillo@220 3585 These changes include many commits, among them:
meillo@121 3586 .Ci 768b5edd9623b7238e12ec8dfc409b82a1ed9e2d
meillo@121 3587 .Ci 0052f1024deb0a0a2fc2e5bacf93d45a5a9c9b32 .
meillo@102 3588 .P
meillo@220 3589 Nevertheless, I worked on the task as well, tidying up the
meillo@220 3590 \fIMH standard library\fP,
meillo@220 3591 .Fn libmh.a .
meillo@220 3592 It is located in the
meillo@123 3593 .Fn sbr
meillo@220 3594 (``subroutines'') directory in the source tree and
meillo@220 3595 includes functions that mmh tools usually need.
meillo@123 3596 Among them are MH-specific functions for profile, context, sequence,
meillo@123 3597 and folder handling, but as well
meillo@123 3598 MH-independent functions, such as auxiliary string functions,
meillo@123 3599 portability interfaces and error-checking wrappers for critical
meillo@123 3600 functions of the standard library.
meillo@220 3601 .BU
meillo@123 3602 I have replaced the
meillo@121 3603 .Fu atooi()
meillo@121 3604 function with calls to
meillo@220 3605 .Fu strtoul() ,
meillo@220 3606 setting the third parameter, the base, to eight.
meillo@123 3607 .Fu strtoul()
meillo@220 3608 is part of C89 and thus considered safe to use
meillo@220 3609 .Ci c490c51b3c0f8871b6953bd0c74551404f840a74 .
meillo@220 3610 .BU
meillo@121 3611 I did remove project-included fallback implementations of
meillo@121 3612 .Fu memmove()
meillo@121 3613 and
meillo@220 3614 .Fu strerror()
meillo@220 3615 .Ci b067ff5c465a5d243ce5a19e562085a9a1a97215 ,
meillo@121 3616 although Peter Maydell had re-included them into nmh in 2008
meillo@121 3617 to support SunOS 4.
meillo@121 3618 Nevertheless, these functions are part of ANSI C.
meillo@121 3619 Systems that do not even provide full ANSI C support should not
meillo@121 3620 put a load on mmh.
meillo@220 3621 .BU
meillo@121 3622 The
meillo@121 3623 .Fu copy()
meillo@180 3624 function copies the string in parameter one to the location in
meillo@180 3625 parameter two.
meillo@121 3626 In contrast to
meillo@121 3627 .Fu strcpy() ,
meillo@121 3628 it returns a pointer to the terminating null-byte in the destination area.
meillo@123 3629 The code was adjusted to replace
meillo@121 3630 .Fu copy()
meillo@123 3631 with
meillo@121 3632 .Fu strcpy() ,
meillo@121 3633 except within
meillo@121 3634 .Fu concat() ,
meillo@121 3635 where
meillo@121 3636 .Fu copy()
meillo@123 3637 was more convenient.
meillo@123 3638 Therefore, the definition of
meillo@121 3639 .Fu copy()
meillo@123 3640 was moved into the source file of
meillo@121 3641 .Fu concat()
meillo@220 3642 and its visibility it limited to that
meillo@220 3643 .Ci 552fd7253e5ee9e554c5c7a8248a6322aa4363bb .
meillo@220 3644 .BU
meillo@121 3645 The function
meillo@121 3646 .Fu r1bindex()
meillo@121 3647 had been a generalized version of
meillo@121 3648 .Fu basename()
meillo@121 3649 with minor differences.
meillo@121 3650 As all calls to
meillo@121 3651 .Fu r1bindex()
meillo@220 3652 had the slash (`\fL/\fP') as delimiter anyway,
meillo@121 3653 replacing
meillo@121 3654 .Fu r1bindex()
meillo@121 3655 with the more specific and better-named function
meillo@121 3656 .Fu basename()
meillo@121 3657 became desirable.
meillo@121 3658 Unfortunately, many of the 54 calls to
meillo@121 3659 .Fu r1bindex()
meillo@123 3660 depended on a special behavior,
meillo@121 3661 which differed from the POSIX specification for
meillo@121 3662 .Fu basename() .
meillo@121 3663 Hence,
meillo@121 3664 .Fu r1bindex()
meillo@121 3665 was kept but renamed to
meillo@123 3666 .Fu mhbasename() ,
meillo@220 3667 setting the delimiter to the slash
meillo@220 3668 .Ci 240013872c392fe644bd4f79382d9f5314b4ea60 .
meillo@121 3669 For possible uses of
meillo@121 3670 .Fu r1bindex()
meillo@121 3671 with a different delimiter,
meillo@121 3672 the ANSI C function
meillo@121 3673 .Fu strrchr()
meillo@121 3674 provides the core functionality.
meillo@220 3675 .BU
meillo@121 3676 The
meillo@121 3677 .Fu ssequal()
meillo@121 3678 function \(en apparently for ``substring equal'' \(en
meillo@121 3679 was renamed to
meillo@121 3680 .Fu isprefix() ,
meillo@220 3681 because this is what it actually checked
meillo@220 3682 .Ci c20b4fa14515c7ab388ce35411d89a7a92300711.
meillo@220 3683 Its source file had included both of the following comments, no joke.
meillo@220 3684 .in -\n(PIu
meillo@121 3685 .VS
meillo@121 3686 /*
meillo@121 3687 * THIS CODE DOES NOT WORK AS ADVERTISED.
meillo@121 3688 * It is actually checking if s1 is a PREFIX of s2.
meillo@121 3689 * All calls to this function need to be checked to see
meillo@121 3690 * if that needs to be changed. Prefix checking is cheaper, so
meillo@121 3691 * should be kept if it's sufficient.
meillo@121 3692 */
meillo@121 3693
meillo@121 3694 /*
meillo@121 3695 * Check if s1 is a substring of s2.
meillo@121 3696 * If yes, then return 1, else return 0.
meillo@121 3697 */
meillo@121 3698 VE
meillo@220 3699 .in +\n(PIu
meillo@220 3700 Eventually, the function was completely replaced with calls to
meillo@220 3701 .Fu strncmp()
meillo@220 3702 .Ci b0b1dd37ff515578cf7cba51625189eb34a196cb .
meillo@121 3703
meillo@102 3704
meillo@102 3705
meillo@102 3706
meillo@133 3707
meillo@133 3708 .H2 "User Data Locations
meillo@133 3709 .P
meillo@133 3710 In nmh, a personal setup consists of the MH profile and the MH directory.
meillo@133 3711 The profile is a file named
meillo@133 3712 .Fn \&.mh_profile
meillo@133 3713 in the user's home directory.
meillo@133 3714 It contains the static configuration.
meillo@133 3715 It also contains the location of the MH directory in the profile entry
meillo@133 3716 .Pe Path .
meillo@133 3717 The MH directory contains the mail storage and is the first
meillo@220 3718 place to search for form files, scan formats, and similar
meillo@133 3719 configuration files.
meillo@133 3720 The location of the MH directory can be chosen freely by the user.
meillo@220 3721 The usual name is a directory named
meillo@133 3722 .Fn Mail
meillo@220 3723 in the user's home directory.
meillo@133 3724 .P
meillo@200 3725 The way MH data is split between profile and MH directory is a legacy.
meillo@133 3726 It is only sensible in a situation where the profile is the only
meillo@133 3727 configuration file.
meillo@133 3728 Why else should the mail storage and the configuration files be intermixed?
meillo@220 3729 They are of different kind:
meillo@220 3730 One kind is the data to be operated on and the other kind is
meillo@220 3731 the configuration to change how tools operate.
meillo@133 3732 Splitting the configuration between the profile and the MH directory
meillo@220 3733 is inappropriate, as well.
meillo@220 3734 I improved the situation by breaking compatibility.
meillo@133 3735 .P
meillo@220 3736 In mmh, personal data is grouped by type.
meillo@220 3737 This results in two distinct parts:
meillo@171 3738 the mail storage and the configuration.
meillo@220 3739 The mail storage directory still contains all the messages,
meillo@133 3740 but, in exception of public sequences files, nothing else.
meillo@133 3741 In difference to nmh, the auxiliary configuration files are no longer
meillo@133 3742 located there.
meillo@133 3743 Therefore, the directory is no longer called the user's \fIMH directory\fP
meillo@220 3744 but the user's \fImail storage\fP.
meillo@133 3745 Its location is still user-chosen, with the default name
meillo@220 3746 .Fn Mail
meillo@133 3747 in the user's home directory.
meillo@220 3748 The configuration is grouped together in the hidden directory
meillo@133 3749 .Fn \&.mmh
meillo@133 3750 in the user's home directory.
meillo@133 3751 This \fImmh directory\fP contains the context file, personal forms,
meillo@133 3752 scan formats, and the like, but also the user's profile, now named
meillo@133 3753 .Fn profile .
meillo@220 3754 The path to the profile is no longer
meillo@133 3755 .Fn $HOME/.mh_profile
meillo@220 3756 but
meillo@133 3757 .Fn $HOME/.mmh/profile .
meillo@220 3758 (The alternative of having file
meillo@133 3759 .Fn $HOME/.mh_profile
meillo@220 3760 and a configuration directory
meillo@133 3761 .Fn $HOME/.mmh
meillo@220 3762 appeared to be inconsistent.)
meillo@220 3763 .P
meillo@133 3764 The approach chosen for mmh is consistent, simple, and familiar to
meillo@133 3765 Unix users.
meillo@220 3766 The main achievement of the change is the clear and sensible separation
meillo@220 3767 of the mail storage and the configuration.
meillo@168 3768 .Ci 7030d7edb099bff36ded7548bb5380f7acab4f9b
meillo@133 3769 .P
meillo@220 3770 As MH allows users to have multiple MH setups,
meillo@220 3771 it is necessary to switch the profile.
meillo@133 3772 The profile is the single entry point to access the rest of a
meillo@133 3773 personal MH setup.
meillo@133 3774 In nmh, the environment variable
meillo@133 3775 .Ev MH
meillo@220 3776 is used to specify a different profile.
meillo@220 3777 To operate in the same MH setup with a separate context, the
meillo@133 3778 .Ev MHCONTEXT
meillo@220 3779 environment variable is used.
meillo@220 3780 This allows having a separate current folder in each terminal at
meillo@220 3781 the same time, for instance.
meillo@220 3782 In mmh, three environment variables replace the two of nmh.
meillo@133 3783 .Ev MMH
meillo@133 3784 overrides the default location of the mmh directory (\c
meillo@133 3785 .Fn .mmh ).
meillo@133 3786 .Ev MMHP
meillo@133 3787 and
meillo@133 3788 .Ev MMHC
meillo@220 3789 override the paths to the profile and context file, respectively.
meillo@133 3790 This approach allows the set of personal configuration files to be chosen
meillo@220 3791 independently of the profile, context, and mail storage.
meillo@133 3792 The new approach has no functional disadvantages,
meillo@133 3793 as every setup I can imagine can be implemented with both approaches,
meillo@220 3794 possibly even easier with the new one.
meillo@220 3795 .Ci 7030d7edb099bff36ded7548bb5380f7acab4f9b
meillo@133 3796
meillo@133 3797
meillo@133 3798
meillo@133 3799
meillo@133 3800
meillo@118 3801 .H2 "Modularization
meillo@220 3802 .Id modularization
meillo@118 3803 .P
meillo@123 3804 The source code of the mmh tools is located in the
meillo@122 3805 .Fn uip
meillo@123 3806 (``user interface programs'') directory.
meillo@180 3807 Each tool has a source file with the name of the command.
meillo@122 3808 For example,
meillo@122 3809 .Pn rmm
meillo@122 3810 is built from
meillo@122 3811 .Fn uip/rmm.c .
meillo@123 3812 Some source files are used for multiple programs.
meillo@122 3813 For example
meillo@122 3814 .Fn uip/scansbr.c
meillo@173 3815 is used for both
meillo@122 3816 .Pn scan
meillo@122 3817 and
meillo@122 3818 .Pn inc .
meillo@122 3819 In nmh, 49 tools were built from 76 source files.
meillo@123 3820 This is a ratio of 1.6 source files per program.
meillo@123 3821 32 programs depended on multiple source files;
meillo@123 3822 17 programs depended on one source file only.
meillo@122 3823 In mmh, 39 tools are built from 51 source files.
meillo@123 3824 This is a ratio of 1.3 source files per program.
meillo@123 3825 18 programs depend on multiple source files;
meillo@123 3826 21 programs depend on one source file only.
meillo@123 3827 (These numbers and the ones in the following text ignore the MH library
meillo@123 3828 as well as shell scripts and multiple names for the same program.)
meillo@122 3829 .P
meillo@123 3830 Splitting the source code of a large program into multiple files can
meillo@220 3831 increase the readability of its source code,
meillo@220 3832 but most of the mmh tools are small and straight-forward programs.
meillo@220 3833 In exception of the MIME handling tools (i.e.
meillo@122 3834 .Pn mhbuild ,
meillo@122 3835 .Pn mhstore ,
meillo@122 3836 .Pn show ,
meillo@220 3837 etc.),
meillo@220 3838 .Pn pick
meillo@220 3839 is the only tool with more than one thousand lines of source code.
meillo@220 3840 Splitting programs with less than one thousand lines of code into
meillo@220 3841 multiple source files leads seldom to better readability.
meillo@220 3842 For such tools, splitting still makes sense
meillo@220 3843 when parts of the code are reused in other programs
meillo@179 3844 and the reused code fragment is (1) not general enough
meillo@179 3845 for including it in the MH library
meillo@179 3846 or (2) has dependencies on a library that only few programs need.
meillo@122 3847 .Fn uip/packsbr.c ,
meillo@122 3848 for instance, provides the core program logic for the
meillo@122 3849 .Pn packf
meillo@122 3850 and
meillo@122 3851 .Pn rcvpack
meillo@122 3852 programs.
meillo@122 3853 .Fn uip/packf.c
meillo@122 3854 and
meillo@122 3855 .Fn uip/rcvpack.c
meillo@122 3856 mainly wrap the core function appropriately.
meillo@122 3857 No other tools use the folder packing functions.
meillo@123 3858 As another example,
meillo@123 3859 .Fn uip/termsbr.c
meillo@220 3860 accesses terminal properties, which requires linking with the
meillo@220 3861 \fItermcap\fP or a \fIcurses\fP library.
meillo@220 3862 If
meillo@123 3863 .Fn uip/termsbr.c
meillo@220 3864 is included in the MH library, then every program needs to be linked
meillo@220 3865 with termcap or curses, although only few of the programs use
meillo@220 3866 the library.
meillo@122 3867 .P
meillo@122 3868 The task of MIME handling is complex enough that splitting its code
meillo@122 3869 into multiple source files improves the readability.
meillo@122 3870 The program
meillo@122 3871 .Pn mhstore ,
meillo@122 3872 for instance, is compiled out of seven source files with 2\|500
meillo@122 3873 lines of code in summary.
meillo@122 3874 The main code file
meillo@122 3875 .Fn uip/mhstore.c
meillo@220 3876 consists of 800 lines; the other 1\|700 lines are code reused in
meillo@123 3877 other MIME handling tools.
meillo@123 3878 It seems to be worthwhile to bundle the generic MIME handling code into
meillo@123 3879 a MH-MIME library, as a companion to the MH standard library.
meillo@220 3880 This is left to be done.
meillo@122 3881 .P
meillo@169 3882 The work already accomplished focussed on the non-MIME tools.
meillo@122 3883 The amount of code compiled into each program was reduced.
meillo@123 3884 This eases the understanding of the code base.
meillo@122 3885 In nmh,
meillo@122 3886 .Pn comp
meillo@122 3887 was built from six source files:
meillo@122 3888 .Fn comp.c ,
meillo@122 3889 .Fn whatnowproc.c ,
meillo@122 3890 .Fn whatnowsbr.c ,
meillo@122 3891 .Fn sendsbr.c ,
meillo@122 3892 .Fn annosbr.c ,
meillo@122 3893 and
meillo@122 3894 .Fn distsbr.c .
meillo@122 3895 In mmh, it builds from only two:
meillo@122 3896 .Fn comp.c
meillo@122 3897 and
meillo@122 3898 .Fn whatnowproc.c .
meillo@123 3899 In nmh's
meillo@123 3900 .Pn comp ,
meillo@123 3901 the core function of
meillo@122 3902 .Pn whatnow ,
meillo@122 3903 .Pn send ,
meillo@122 3904 and
meillo@122 3905 .Pn anno
meillo@220 3906 were all compiled into
meillo@122 3907 .Pn comp .
meillo@123 3908 This saved the need to execute these programs with
meillo@220 3909 the expensive system calls
meillo@122 3910 .Fu fork()
meillo@122 3911 and
meillo@220 3912 .Fu exec() .
meillo@171 3913 Whereas this approach improved the time performance,
meillo@171 3914 it interwove the source code.
meillo@122 3915 Core functionalities were not encapsulated into programs but into
meillo@122 3916 function, which were then wrapped by programs.
meillo@122 3917 For example,
meillo@122 3918 .Fn uip/annosbr.c
meillo@122 3919 included the function
meillo@122 3920 .Fu annotate() .
meillo@122 3921 Each program that wanted to annotate messages, included the source file
meillo@123 3922 .Fn uip/annosbr.c
meillo@123 3923 and called
meillo@123 3924 .Fu annotate() .
meillo@123 3925 Because the function
meillo@123 3926 .Fu annotate()
meillo@123 3927 was used like the tool
meillo@123 3928 .Pn anno ,
meillo@123 3929 it had seven parameters, reflecting the command line switches of the tool.
meillo@122 3930 When another pair of command line switches was added to
meillo@122 3931 .Pn anno ,
meillo@122 3932 a rather ugly hack was implemented to avoid adding another parameter
meillo@220 3933 to the function
meillo@220 3934 .Ci d9b1d57351d104d7ec1a5621f090657dcce8cb7f .
meillo@122 3935 .P
meillo@122 3936 In mmh, the relevant code of
meillo@122 3937 .Pn comp
meillo@122 3938 comprises the two files
meillo@122 3939 .Fn uip/comp.c
meillo@122 3940 and
meillo@122 3941 .Fn uip/whatnowproc.c ,
meillo@220 3942 together 210 lines of code,
meillo@220 3943 whereas in nmh,
meillo@122 3944 .Pn comp
meillo@122 3945 comprises six files with 2\|450 lines.
meillo@220 3946 Not all of the code in these six files is actually used by
meillo@122 3947 .Pn comp ,
meillo@220 3948 but the reader needed to read it all to know which parts are relevant.
meillo@220 3949 Understanding nmh's
meillo@122 3950 .Pn comp ,
meillo@123 3951 required understanding the inner workings of
meillo@122 3952 .Fn uip/annosbr.c
meillo@122 3953 first.
meillo@123 3954 To be sure to fully understand a program, its whole source code needs
meillo@122 3955 to be examined.
meillo@123 3956 Not doing so is a leap of faith, assuming that the developers
meillo@122 3957 have avoided obscure programming techniques.
meillo@220 3958 Here, it should be recalled that information passed in obscure ways
meillo@220 3959 through the program's source base, due to the aforementioned hack
meillo@220 3960 to save an additional parameter in nmh's
meillo@220 3961 .Pn anno .
meillo@220 3962 .P
meillo@220 3963 In mmh, understanding
meillo@220 3964 .Pn comp
meillo@220 3965 requires to read only 210 lines of code to read, whereas the amount
meillo@220 3966 is ten times more for nmh's
meillo@220 3967 .Pn comp .
meillo@220 3968 .P
meillo@220 3969 By separating the tools on the program-level,
meillo@220 3970 the boundaries are clearly visible, as the interfaces are calls to
meillo@122 3971 .Fu exec()
meillo@122 3972 rather than arbitrary function calls.
meillo@220 3973 Additionally, this kind of separation is more strict because
meillo@220 3974 it is technically enforced by the operating system;
meillo@220 3975 it can not be simply bypassed with global variables.
meillo@220 3976 Good separation simplifies the understanding of program code
meillo@220 3977 because the area influenced by any particular statement is small.
meillo@220 3978 As I have read a lot in nmh's code base during the last two years,
meillo@220 3979 I have learned about the easy and the difficult parts.
meillo@220 3980 In my observation, the understanding of code is enormously eased
meillo@220 3981 if the influenced area is small and clearly bounded.
meillo@123 3982 .P
meillo@220 3983 Yet, the real problem is another:
meillo@123 3984 Nmh violates the golden ``one tool, one job'' rule of the Unix philosophy.
meillo@123 3985 Understanding
meillo@122 3986 .Pn comp
meillo@123 3987 requires understanding
meillo@123 3988 .Fn uip/annosbr.c
meillo@123 3989 and
meillo@123 3990 .Fn uip/sendsbr.c
meillo@123 3991 because
meillo@123 3992 .Pn comp
meillo@220 3993 annotates and sends messages.
meillo@220 3994 In nmh, there surely exist the tools
meillo@220 3995 .Pn anno
meillo@220 3996 and
meillo@122 3997 .Pn send ,
meillo@220 3998 which cover these jobs,
meillo@220 3999 but
meillo@122 4000 .Pn comp
meillo@123 4001 and
meillo@122 4002 .Pn repl
meillo@122 4003 and
meillo@122 4004 .Pn forw
meillo@122 4005 and
meillo@122 4006 .Pn dist
meillo@122 4007 and
meillo@122 4008 .Pn whatnow
meillo@122 4009 and
meillo@220 4010 .Pn viamail
meillo@220 4011 \(en they all (!) \(en
meillo@220 4012 have the same annotating and sending functions included, once more.
meillo@220 4013 As a result,
meillo@123 4014 .Pn comp
meillo@123 4015 sends messages without using
meillo@123 4016 .Pn send .
meillo@123 4017 The situation is the same as if
meillo@123 4018 .Pn grep
meillo@220 4019 would page its output without using
meillo@123 4020 .Pn more
meillo@123 4021 just because both programs are part of the same code base.
meillo@123 4022 .P
meillo@220 4023 The clear separation on the surface of nmh
meillo@220 4024 \(en the tool chest approach \(en
meillo@123 4025 is violated on the level below.
meillo@122 4026 This violation is for the sake of time performance.
meillo@220 4027 Decades ago, sacrificing readability and conceptional beauty
meillo@220 4028 for speed might have been necessary to prevent MH from being
meillo@220 4029 unusably slow, but today this is not the case anymore.
meillo@220 4030 No longer should speed improvements that became unnecessary be kept.
meillo@220 4031 No longer should readability or conceptional beauty be sacrificed.
meillo@220 4032 No longer should the Unix philosophy's ``one tool, one job''
meillo@220 4033 guideline be violated.
meillo@123 4034 Therefore, mmh's
meillo@123 4035 .Pn comp
meillo@220 4036 no longer sends messages.
meillo@220 4037 .P
meillo@123 4038 In mmh, different jobs are divided among separate programs that
meillo@122 4039 invoke each other as needed.
meillo@123 4040 In consequence,
meillo@123 4041 .Pn comp
meillo@123 4042 invokes
meillo@123 4043 .Pn whatnow
meillo@123 4044 which thereafter invokes
meillo@220 4045 .Pn send
meillo@168 4046 .Ci 3df5ab3c116e6d4a2fb4bb5cc9dfc5f781825815
meillo@220 4047 .Ci c73c00bfccd22ec77e9593f47462aeca4a8cd9c0 .
meillo@123 4048 The clear separation on the surface is maintained on the level below.
meillo@220 4049 Human users and other tools use the same interface \(en
meillo@123 4050 annotations, for example, are made by invoking
meillo@123 4051 .Pn anno ,
meillo@220 4052 no matter if requested by programs or by human beings
meillo@168 4053 .Ci 469a4163c2a1a43731d412eaa5d9cae7d670c48b
meillo@168 4054 .Ci aed384169af5204b8002d06e7a22f89197963d2d
meillo@220 4055 .Ci 3caf9e298a8861729ca8b8a84f57022b6f3ea742 .
meillo@123 4056 The decrease of tools built from multiple source files and thus
meillo@123 4057 the decrease of
meillo@123 4058 .Fn uip/*sbr.c
meillo@220 4059 files confirm the improvement
meillo@168 4060 .Ci 9e6d91313f01c96b4058d6bf419a8ca9a207bc33
meillo@168 4061 .ci 81744a46ac9f845d6c2b9908074d269275178d2e
meillo@168 4062 .Ci f0f858069d21111f0dbea510044593f89c9b0829
meillo@168 4063 .Ci 0503a6e9be34f24858b55b555a5c948182b9f24b
meillo@168 4064 .Ci 27826f9353e0f0b04590b7d0f8f83e60462b90f0
meillo@168 4065 .Ci d1da1f94ce62160aebb30df4063ccbc53768656b
meillo@220 4066 .Ci c42222869e318fff5dec395eca3e776db3075455 .
meillo@220 4067 This is also visible in the complexity of the build dependency graphs:
meillo@145 4068
meillo@220 4069 .sp
meillo@220 4070 Nmh:
meillo@220 4071 .BP input/deps-nmh.eps .5i
meillo@220 4072 .EP
meillo@220 4073 .sp
meillo@220 4074 Mmh:
meillo@220 4075 .BP input/deps-mmh.eps .8i
meillo@220 4076 .EP
meillo@145 4077
meillo@226 4078 The figures display all program to source file relationships where
meillo@226 4079 programs (ellipses) are built from multiple source files (rectangles).
meillo@220 4080 The primary source file of each program is omited from the graph.