docs/master

annotate ch03.roff @ 87:7d5b180de542

find changesets by author, revision, files, or words in the commit message
All kinds of rework plus new refs.
author markus schnalke <meillo@marmaro.de>
date Tue, 12 Jun 2012 18:04:55 +0200
parents fb0d1b4c8fb1
children 30830e3b9e98
rev   line source
meillo@58 1 .H0 "Discussion
meillo@0 2 .P
meillo@58 3 This main chapter discusses the practical work done in the mmh project.
meillo@58 4 It is structured along the goals to achieve. The concrete work done
meillo@58 5 is described in the examples of how the general goals were achieved.
meillo@87 6 The discussion compares the current version of mmh with the state of
meillo@87 7 nmh just before the mmh project started, i.e. Fall 2011.
meillo@87 8 Current changes of nmh will be mentioned only as side notes.
meillo@87 9 .\" XXX where do I discuss the parallel development of nmh?
meillo@58 10
meillo@58 11
meillo@58 12
meillo@58 13 .H1 "Stream-lining
meillo@58 14
meillo@0 15 .P
meillo@58 16 MH had been considered an all-in-one system for mail handling.
meillo@58 17 The community around nmh has a similar understanding.
meillo@87 18 In fundamental difference, mmh shall be a MUA only.
meillo@87 19 I believe that the development of all-in-one mail systems is obsolete.
meillo@87 20 Today, email is too complex to be fully covered by single projects.
meillo@87 21 Such a project won't be able to excel in all aspects.
meillo@87 22 Instead, the aspects of email should be covered my multiple projects,
meillo@87 23 which then can be combined to form a complete system.
meillo@87 24 Excellent implementations for the various aspects of email exist already.
meillo@87 25 Just to name three examples: Postfix is a specialized MTA,
meillo@87 26 Procmail is a specialized MDA, and Fetchmail is a specialized MRA.
meillo@87 27 I believe that it is best to use such specilized tools instead of
meillo@87 28 providing the same function again as a side-component in the project.
meillo@58 29 .P
meillo@87 30 Doing something well, requires to focus on a small set of specific aspects.
meillo@87 31 Under the assumption that focused development produces better results
meillo@87 32 in the particular area, specialized projects will likely be superior
meillo@87 33 in their field of focus.
meillo@87 34 Hence, all-in-one mail system projects \(en no matter if monolithic
meillo@87 35 or modular \(en will never be the best choice in any of the fields.
meillo@87 36 Even in providing the best consistent all-in-one system they are likely
meillo@87 37 to be beaten by projects that focus only on integrating existing mail
meillo@87 38 components to a homogenious system.
meillo@87 39 .P
meillo@87 40 The limiting resource in Free Software community development
meillo@87 41 is usually man power.
meillo@87 42 If the development power is spread over a large development area,
meillo@87 43 it becomes even more difficult to compete with the specialists in the
meillo@87 44 various fields.
meillo@87 45 The concrete situation for MH-based mail systems is even tougher,
meillo@87 46 given the small and aged community, including both developers and users,
meillo@87 47 it has.
meillo@87 48 .P
meillo@87 49 In consequence, I believe that the available development resources
meillo@87 50 should be focused on the point where MH is most unique.
meillo@87 51 This is clearly the user interface \(en the MUA.
meillo@87 52 Peripheral parts should be removed to stream-line mmh for the MUA task.
meillo@60 53
meillo@60 54
meillo@87 55 .H2 "Removal of the Mail Transfer Facilities
meillo@60 56 .P
meillo@60 57 In contrast to nmh, which also provides mail submission and mail retrieval
meillo@87 58 agents, mmh is a MUA only.
meillo@66 59 This general difference in the view on the character of nmh
meillo@76 60 initiated the development of mmh.
meillo@66 61 Removing the mail transfer facilities had been the first work task
meillo@76 62 in the mmh project.
meillo@60 63 .P
meillo@66 64 The MSA is called \fIMessage Transfer Service\fP (MTS) in nmh.
meillo@76 65 The facility established network connections and spoke SMTP to submit
meillo@60 66 messages for relay to the outside world.
meillo@76 67 This part was implemented by the
meillo@60 68 .Pn post
meillo@60 69 command.
meillo@87 70 The changes in emailing in the last years
meillo@87 71 demanded changes in this part of nmh too.
meillo@76 72 Encryption and authetication for network connections
meillo@87 73 needed to be supported, hence TLS and SASL were introduced into nmh.
meillo@87 74 This added complexity to nmh without improving it in its core functions.
meillo@87 75 Also, keeping up with recent developments in the field of
meillo@87 76 mail transfer requires development power and specialists.
meillo@87 77 In mmh this whole facility was simply cut off.
meillo@76 78 .Ci f6aa95b724fd8c791164abe7ee5468bf5c34f226
meillo@76 79 .Ci fecd5d34f65597a4dfa16aeabea7d74b191532c3
meillo@76 80 .Ci 156d35f6425bea4c1ed3c4c79783dc613379c65b
meillo@87 81 Instead, mmh depends on an external MSA.
meillo@60 82 The only outgoing interface available to mmh is the
meillo@60 83 .Pn sendmail
meillo@87 84 command, which almost any MSA provides.
meillo@87 85 If not, a wrapper program can be written.
meillo@87 86 It must read the message from the standard input, extract the
meillo@87 87 recipient addresses from the message header, and hand the message
meillo@87 88 over to the MSA.
meillo@87 89 For example, a wrapper script for qmail would be:
meillo@87 90 .VS
meillo@87 91 #!/bin/sh
meillo@87 92 # ignore command line arguments
meillo@87 93 exec qmail-inject
meillo@87 94 VE
meillo@87 95 The requirement to parse the recipient addresses out of the message header
meillo@87 96 is likely to be removed in the future.
meillo@87 97 Then mmh would give the recipient addresses as command line arguments.
meillo@87 98 This is clearly the better interface, but mmh does not provide it yet.
meillo@87 99 .\" XXX implement it
meillo@60 100 .P
meillo@60 101 To retrieve mail, the
meillo@60 102 .Pn inc
meillo@76 103 command established network connections
meillo@76 104 and spoke POP3 to retrieve mail from remote servers.
meillo@76 105 As with mail submission, the network connections required encryption and
meillo@87 106 authentication, thus TLS and SASL were added.
meillo@87 107 Support for message retrieval through IMAP will become necessary
meillo@87 108 to be added soon, too, and so on for any changes in mail transfer.
meillo@76 109 Mmh has dropped the support for retrieving mail from remote locations.
meillo@76 110 .Ci ab7b48411962d26439f92f35ed084d3d6275459c
meillo@76 111 Instead, it depends on an external tool to cover this task.
meillo@87 112 In mmh there exist two paths for messages to enter mmh's mail storage:
meillo@87 113 (1) Mail can be incorporate with
meillo@60 114 .Pn inc
meillo@87 115 from the system maildrop, or (2) with
meillo@60 116 .Pn rcvstore
meillo@87 117 by reading them, one at a time, from the standard input.
meillo@60 118 .P
meillo@60 119 With the removal of the MSA and MRA, mmh converted from an all-in-one
meillo@87 120 mail system to being a MUA only.
meillo@87 121 Following the Unix philosophy, it now focuses on one job and
meillo@76 122 tries to do that one well.
meillo@76 123 Not only the programs follow that tenet but also the project itself does so.
meillo@60 124 Now, of course, mmh depends on third-party software.
meillo@87 125 An external MSA is required to transfer mail to the outside world;
meillo@60 126 an external MRA is required to retrieve mail from remote machines.
meillo@60 127 There exist excellent implementations of such software,
meillo@76 128 which do this specific task likely better than the internal
meillo@87 129 versions had done it.
meillo@87 130 Also, the best suiting programs can be freely chosen.
meillo@60 131 .P
meillo@60 132 As it had already been possible to use an external MSA or MRA,
meillo@60 133 why not keep the internal version for convenience?
meillo@76 134 The question whether there is sense in having a fall-back pager in all
meillo@76 135 the command line tools, for the cases when
meillo@60 136 .Pn more
meillo@60 137 or
meillo@60 138 .Pn less
meillo@76 139 aren't available, appears to be ridiculous.
meillo@87 140 Now, an MSA or MRA is more complex than a text pager
meillo@87 141 and not necessarily available but still the concept of orthogonal
meillo@87 142 design holds: ``Write programs that do one thing and do it well.''
meillo@87 143 .[
meillo@87 144 mcilroy unix phil
meillo@87 145 p. 53
meillo@87 146 .]
meillo@87 147 .[
meillo@87 148 mcilroy bstj foreword
meillo@87 149 .]
meillo@87 150 Here, this part of the Unix philosophy was applied not only
meillo@87 151 to the programs but to the project itself.
meillo@87 152 In other words:
meillo@87 153 ``Develop projects that focus on one thing and do it well.''
meillo@87 154 Projects grown complex should be split for the same reasons programs grown
meillo@87 155 complex should be split.
meillo@87 156 If it is conceptionally more elegant to have the MSA and MRA
meillo@87 157 separate projects then they should be separated.
meillo@87 158 This is the case here, in my opinion.
meillo@87 159 The RFCs propose this separation by clearly distinguishing the different
meillo@87 160 mail handling tasks.
meillo@87 161 .[
meillo@87 162 rfc 821
meillo@87 163 .]
meillo@87 164 The small interfaces between the mail agents support the separation.
meillo@76 165 .P
meillo@87 166 In the beginning, email had been small and simple.
meillo@76 167 (\c
meillo@60 168 .Pn /bin/mail
meillo@76 169 had once covered anything there was to email and still had been small
meillo@76 170 and simple.)
meillo@76 171 Then the essential complexity of email increased.
meillo@87 172 (Essential complexity is the complexity defined by the problem itself.\0
meillo@87 173 .[[
meillo@87 174 brooks no silver bullet
meillo@87 175 .]])
meillo@87 176 Email systems reacted to this change: They grew.
meillo@87 177 RFCs started to introduce mail agents and separated the various tasks
meillo@87 178 because the existing tasks became more extensive and new tasks appeared.
meillo@87 179 Again, email systems grew, or they split parts off.
meillo@76 180 In nmh, for instance, the POP server, which the original MH had included,
meillo@76 181 was removed.
meillo@87 182 Now is the time to go one step further and remove the MSA and MRA, too.
meillo@87 183 Not only does this decrease the code size of the project,
meillo@87 184 but, more important, it unburdens mmh of the whole field of
meillo@87 185 message transfer with all its implications for the project.
meillo@87 186 There's no more need to concern with changes in network transfer.
meillo@76 187 This independence is received by depending on an external program
meillo@76 188 that covers the field.
meillo@76 189 Today, this is a reasonable exchange.
meillo@60 190 .P
meillo@87 191 Function can be added in three different ways:
meillo@87 192 .BU
meillo@87 193 Implementing the function originally in the project.
meillo@87 194 .BU
meillo@87 195 Depending on a library that provides the function.
meillo@87 196 .BU
meillo@87 197 Depending on a program that provides the function.
meillo@87 198 .P
meillo@87 199 Whereas adding the function originally to the project increases the
meillo@76 200 code size most and requires most maintenance and development work,
meillo@87 201 it makes the project most independent of other software.
meillo@87 202 Using libraries or external programs require less maintenance work
meillo@87 203 but introduces dependencies on external software.
meillo@87 204 Programs have the smallest interfaces and provide the best separation
meillo@87 205 but possibly limit the information exchange.
meillo@87 206 External libraries are stronger connected than external programs,
meillo@87 207 thus information can be exchanged more flexible.
meillo@87 208 Adding code to a project increases maintenance work.
meillo@87 209 .\" XXX ref
meillo@87 210 Implementing complex functions originally in the project will add
meillo@87 211 a lot of code.
meillo@87 212 This should be avoided if possible.
meillo@66 213 Hence, the dependencies only change in kind, not in their existence.
meillo@66 214 In mmh, library dependencies on
meillo@66 215 .Pn libsasl2
meillo@66 216 and
meillo@66 217 .Pn libcrypto /\c
meillo@66 218 .Pn libssl
meillo@66 219 were treated against program dependencies on an MSA and an MRA.
meillo@87 220 This also meant treating build-time dependencies against run-time
meillo@87 221 dependencies.
meillo@66 222 Besides program dependencies providing the stronger separation
meillo@66 223 and being more flexible, they also allowed
meillo@66 224 over 6\|000 lines of code to be removed from mmh.
meillo@66 225 This made mmh's code base about 12\|% smaller.
meillo@87 226 Reducing the project's code size by such an amount without actually
meillo@87 227 losing functionality is a convincing argument.
meillo@87 228 Actually, as external MSAs and MRAs are likely superior to the
meillo@87 229 project's internal versions, the common user even gains functionality.
meillo@66 230 .P
meillo@76 231 Users of MH should not have problems to set up an external MSA and MRA.
meillo@60 232 Also, the popular MSAs and MRAs have large communities and a lot
meillo@60 233 of documentation available.
meillo@87 234 Choices for MSAs range from full-featured MTAs like
meillo@60 235 .I Postfix
meillo@87 236 over mid-size MTAs like
meillo@60 237 .I masqmail
meillo@60 238 and
meillo@60 239 .I dma
meillo@60 240 to small forwarders like
meillo@60 241 .I ssmtp
meillo@60 242 and
meillo@60 243 .I nullmailer .
meillo@60 244 Choices for MRAs include
meillo@60 245 .I fetchmail ,
meillo@60 246 .I getmail ,
meillo@60 247 .I mpop
meillo@60 248 and
meillo@60 249 .I fdm .
meillo@60 250
meillo@60 251
meillo@60 252 .H2 "Removal of non-MUA Tools
meillo@60 253 .P
meillo@87 254 One goal of mmh is to remove the tools that are not part of the MUA's task.
meillo@87 255 Further more, any tools that don't improve the MUA's job significently
meillo@87 256 should be removed.
meillo@87 257 Loosely related and rarely used tools distract from the lean appearance.
meillo@87 258 They require maintenance work without adding much to the core task.
meillo@87 259 On removing these tools, the project shall become more stream-lined
meillo@87 260 and focused.
meillo@76 261 In mmh the following tools are not available anymore:
meillo@62 262 .BU
meillo@58 263 .Pn conflict
meillo@87 264 was removed
meillo@76 265 .Ci 8b235097cbd11d728c07b966cf131aa7133ce5a9
meillo@87 266 because it is a mail system maintenance tool that is not MUA-related.
meillo@87 267 It even checked
meillo@58 268 .Fn /etc/passwd
meillo@58 269 and
meillo@58 270 .Fn /etc/group
meillo@87 271 for consistency, which is completely unrelated to email.
meillo@87 272 A tool like
meillo@87 273 .Pn conflict
meillo@87 274 is surely useful, but it should not be shipped with mmh.
meillo@76 275 .\" XXX historic reasons?
meillo@62 276 .BU
meillo@58 277 .Pn rcvtty
meillo@87 278 was removed
meillo@87 279 .Ci 14767c94b3827be7c867196467ed7aea5f6f49b0
meillo@87 280 because its usecase of writing to the user's terminal
meillo@76 281 on receiving of mail is obsolete.
meillo@87 282 If users like to be informed of new mail, the shell's
meillo@58 283 .Ev MAILPATH
meillo@87 284 variable or graphical notifications are technically more appealing.
meillo@62 285 Writing directly to a terminals is hardly ever wanted today.
meillo@62 286 If though one wants to have it this way, the standard tool
meillo@58 287 .Pn write
meillo@58 288 can be used in a way similar to:
meillo@82 289 .VS
meillo@58 290 scan -file - | write `id -un`
meillo@82 291 VE
meillo@62 292 .BU
meillo@58 293 .Pn viamail
meillo@87 294 was removed
meillo@87 295 .Ci eda72d6a7a7c20ff123043fb7f19c509ea01f932
meillo@87 296 when the new attachment system was activated, because
meillo@58 297 .Pn forw
meillo@76 298 could then cover the task itself.
meillo@62 299 The program
meillo@58 300 .Pn sendfiles
meillo@62 301 was rewritten as a shell script wrapper around
meillo@58 302 .Pn forw .
meillo@76 303 .Ci 0e82199cf3c991a173e0ac8aa776efdb3ded61e6
meillo@62 304 .BU
meillo@58 305 .Pn msgchk
meillo@87 306 was removed
meillo@87 307 .Ci bb9360ead7eb7a3fedcce2eeedfc660014e41dbe ,
meillo@87 308 because it lost its use case when POP support was removed.
meillo@76 309 A call to
meillo@58 310 .Pn msgchk
meillo@87 311 provided hardly more information than:
meillo@82 312 .VS
meillo@58 313 ls -l /var/mail/meillo
meillo@82 314 VE
meillo@87 315 It did distinguished between old and new mail, but
meillo@87 316 this detail information and can be retrieved with
meillo@76 317 .Pn stat (1),
meillo@62 318 too.
meillo@62 319 A very small shell script could be written to output the information
meillo@76 320 in a similar way, if truly necessary.
meillo@76 321 As mmh's
meillo@76 322 .Pn inc
meillo@87 323 only incorporates mail from the user's local maildrop,
meillo@62 324 and thus no data transfers over slow networks are involved,
meillo@76 325 there's hardly any need to check for new mail before incorporating it.
meillo@62 326 .BU
meillo@58 327 .Pn msh
meillo@87 328 was removed
meillo@76 329 .Ci 916690191222433a6923a4be54b0d8f6ac01bd02
meillo@87 330 because the tool was in conflict with the philosophy of MH.
meillo@76 331 It provided an interactive shell to access the features of MH,
meillo@76 332 but it wasn't just a shell, tailored to the needs of mail handling.
meillo@76 333 Instead it was one large program that had several MH tools built in.
meillo@76 334 This conflicts with the major feature of MH of being a tool chest.
meillo@76 335 .Pn msh 's
meillo@76 336 main use case had been accessing Bulletin Boards, which have seized to
meillo@62 337 be popular.
meillo@62 338 .P
meillo@62 339 Removing
meillo@58 340 .Pn msh ,
meillo@76 341 together with the truly archaic code relicts
meillo@58 342 .Pn vmh
meillo@58 343 and
meillo@58 344 .Pn wmh ,
meillo@62 345 saved more than 7\|000 lines of C code \(en
meillo@66 346 about 15\|% of the project's original source code amount.
meillo@76 347 .P
meillo@76 348 Having less code (with equal readability, of course)
meillo@76 349 for the same functionality is an advantage.
meillo@63 350 Less code means less bugs and less maintenance work.
meillo@76 351 As
meillo@63 352 .Pn rcvtty
meillo@63 353 and
meillo@63 354 .Pn msgchk
meillo@87 355 are assumed to be rarely used and can be implemented in different ways,
meillo@87 356 why should one keep them?
meillo@76 357 Removing them stream-lines mmh.
meillo@63 358 .Pn viamail 's
meillo@63 359 use case is now partly obsolete and partly covered by
meillo@63 360 .Pn forw ,
meillo@76 361 hence there's no reason to still maintain it.
meillo@63 362 .Pn conflict
meillo@76 363 is not related to the mail client, and
meillo@63 364 .Pn msh
meillo@63 365 conflicts with the basic concept of MH.
meillo@76 366 Theses two tools might still be useful, but they should not be part of mmh.
meillo@63 367 .P
meillo@76 368 Finally, there's
meillo@76 369 .Pn slocal .
meillo@76 370 .Pn slocal
meillo@76 371 is an MDA and thus not directly MUA-related.
meillo@78 372 It should be removed, because including it is a violation
meillo@78 373 of the idea that mmh is a MUA only.
meillo@78 374 It should become a separate project.
meillo@87 375 However,
meillo@76 376 .Pn slocal
meillo@76 377 provides rule-based processing of messages, like filing them into
meillo@76 378 different folders, which is otherwise not available in mmh.
meillo@87 379 Although
meillo@76 380 .Pn slocal
meillo@87 381 does neither pull in dependencies nor does it include a separate
meillo@87 382 technical area (cf. Sec. XXX),
meillo@87 383 still it accounts for about 1\|000 lines of code that need to be maintained.
meillo@76 384 As
meillo@76 385 .Pn slocal
meillo@76 386 is almost self-standing, it should be split off into a separate project.
meillo@76 387 This would cut the strong connection between the MUA mmh and the MDA
meillo@76 388 .Pn slocal .
meillo@87 389 For anyone not using MH,
meillo@87 390 .Pn slocal
meillo@87 391 would become yet another independent MDA, like
meillo@87 392 .I procmail .
meillo@87 393 The need to install a complete MH system to have
meillo@87 394 .Pn slocal
meillo@87 395 would be gone.
meillo@76 396 Likewise, mmh users could decide to use
meillo@76 397 .I procmail
meillo@87 398 without having a second, unused MDA,
meillo@87 399 .Pn slocal ,
meillo@76 400 installed.
meillo@76 401 That's conceptionally the best solution.
meillo@76 402 Yet,
meillo@76 403 .Pn slocal
meillo@87 404 is not split off.
meillo@87 405 I feel unsure with removing it from mmh.
meillo@87 406 Hence, I defer the decision over
meillo@87 407 .Pn slocal .
meillo@87 408 In the meanwhile
meillo@78 409 .Pn slocal
meillo@87 410 does not hurt because it is unrelated to the rest of mmh.
meillo@0 411
meillo@58 412
meillo@76 413 .H2 "\fLshow\fP and \fPmhshow\fP
meillo@58 414 .P
meillo@69 415 Since the very beginning \(en already in the first concept paper \(en
meillo@58 416 .Pn show
meillo@62 417 had been MH's message display program.
meillo@58 418 .Pn show
meillo@76 419 mapped message numbers and sequences to files and invoked
meillo@58 420 .Pn mhl
meillo@76 421 to have the files formated.
meillo@76 422 For MIME, this approach wasn't sufficient anymore.
meillo@58 423 MIME messages can consist of multiple parts, some of which aren't
meillo@62 424 directly displayable, and text content might be encoded in
meillo@58 425 foreign charsets.
meillo@58 426 .Pn show 's
meillo@76 427 understanding of messages and
meillo@58 428 .Pn mhl 's
meillo@58 429 limited display facilities couldn't cope with the task any longer.
meillo@62 430 .P
meillo@76 431 Instead of extending these tools, additional tools were written from scratch
meillo@58 432 and then added to the MH tool chest. Doing so is encouraged by the
meillo@58 433 tool chest approach. The new tools could be added without interfering
meillo@78 434 with the existing ones.
meillo@78 435 Modular design is a great advantage for extending a system.
meillo@62 436 First, the new MIME features were added in form of the single program
meillo@58 437 .Pn mhn .
meillo@58 438 The command
meillo@82 439 .Cl "mhn -show 42
meillo@58 440 would show the MIME message numbered 42.
meillo@58 441 With the 1.0 release of nmh in February 1999, Richard Coleman finished
meillo@58 442 the split of
meillo@58 443 .Pn mhn
meillo@58 444 into a set of specialized programs, which together covered the
meillo@62 445 multiple aspects of MIME. One of these resulting tools was
meillo@69 446 .Pn mhshow ,
meillo@69 447 which replaced the
meillo@82 448 .Cl "mhn -show
meillo@62 449 call.
meillo@76 450 It was capable to display a MIME message appropriately.
meillo@62 451 .P
meillo@76 452 From then on, two message display tools were part of nmh:
meillo@76 453 .Pn show
meillo@76 454 and
meillo@76 455 .Pn mhshow .
meillo@76 456 Because the user should not need to invoke the right tool
meillo@69 457 whether the message uses MIME or not,
meillo@69 458 .Pn show
meillo@69 459 was extended to automatically hand the job over to
meillo@69 460 .Pn mhshow
meillo@69 461 if displaying the message would be beyond
meillo@69 462 .Pn show 's
meillo@69 463 abilities.
meillo@69 464 In consequence, the user would invoke
meillo@69 465 .Pn show
meillo@69 466 (possibly through
meillo@69 467 .Pn next
meillo@69 468 or
meillo@69 469 .Pn prev )
meillo@69 470 and get the message printed with either
meillo@69 471 .Pn show
meillo@69 472 or
meillo@69 473 .Pn mhshow ,
meillo@69 474 whatever was more appropriate.
meillo@69 475 (There was also a switch for
meillo@69 476 .Pn show
meillo@69 477 to never invoke
meillo@76 478 .Pn mhshow .
meillo@76 479 .Pn show
meillo@76 480 was able to display MIME messages if they contained only a single text
meillo@76 481 part.)
meillo@69 482 .P
meillo@69 483 Having two similar tools for essentially the same task is redundant.
meillo@76 484 The development of both programs needed to be in sync,
meillo@76 485 to ensure that the programs behaved in a similar way,
meillo@76 486 because they were used like a single tool.
meillo@76 487 Different behavior would have surprised the user.
meillo@69 488 .P
meillo@69 489 Today, non-MIME messages are rather seen to be a special case of
meillo@69 490 MIME messages, than MIME messages are seen to be an extension to
meillo@76 491 original email.
meillo@69 492 As
meillo@69 493 .Pn mhshow
meillo@69 494 had already be able to display non-MIME messages, it was natural
meillo@69 495 to drop
meillo@69 496 .Pn show
meillo@69 497 in favor of using
meillo@69 498 .Pn mhshow
meillo@69 499 exclusively.
meillo@76 500 This decision followed the idea of orthogonal design.
meillo@76 501 For convenience,
meillo@76 502 .Pn mhshow
meillo@76 503 was then renamed to
meillo@76 504 .Pn show .
meillo@72 505 .Ci 4c1efddfd499300c7e74263e57d8aa137e84c853
meillo@69 506 .P
meillo@76 507 To prepare for this transition,
meillo@69 508 .Pn mhshow
meillo@69 509 was reworked to behave more like
meillo@69 510 .Pn show
meillo@69 511 first.
meillo@76 512 (Section XXX describes this rework from a different perspective.)
meillo@69 513 Once the tools behaved similar, the replacing became a natural decision.
meillo@69 514 In mmh,
meillo@69 515 .Pn show
meillo@69 516 is the one single message display program again, but it handles
meillo@69 517 MIME messages as well as non-MIME messages.
meillo@76 518 Now, there's only one program to maintain, and users don't need to deal
meillo@69 519 with the existance of two display programs.
meillo@69 520 .P
meillo@76 521 There's one reason why removing the old
meillo@69 522 .Pn show
meillo@69 523 hurts: It had been such a simple program.
meillo@69 524 Its lean elegance is missing to
meillo@69 525 .Pn mhshow ,
meillo@69 526 i.e. the new
meillo@69 527 .Pn show .
meillo@69 528 But there is no chance, because supporting MIME causes essentially
meillo@69 529 higher complexity.
meillo@58 530
meillo@58 531
meillo@58 532 .H2 "Removal of Configure Options
meillo@58 533 .P
meillo@76 534 Customization is a double-edged sword.
meillo@76 535 It allows better suiting setups, but not for free.
meillo@76 536 There is the cost of code complexity to be able to customize.
meillo@76 537 There is the cost of less tested setups, because there are
meillo@72 538 more possible setups and especially corner-cases.
meillo@76 539 And, there is the cost of choice itself.
meillo@76 540 The code complexity directly affects the developers.
meillo@72 541 Less tested code affects both, users and developers.
meillo@76 542 The problem of choice affects the users, for once by having to
meillo@76 543 choose, but also by complexer interfaces that require more documentation.
meillo@72 544 Whenever options add little advantages, they should be considered for
meillo@72 545 removal.
meillo@72 546 I have reduced the number of project-specific configure options from
meillo@72 547 fifteen to three.
meillo@74 548
meillo@76 549 .U3 "Mail Transfer Facilities
meillo@74 550 .P
meillo@85 551 With the removal of the mail transfer facilities five configure
meillo@85 552 options vanished:
meillo@85 553 .P
meillo@85 554 The switches
meillo@85 555 .Sw --with-tls
meillo@85 556 and
meillo@85 557 .Sw --with-cyrus-sasl
meillo@85 558 had activated the support for transfer encryption and authetication.
meillo@85 559 This is not needed anymore.
meillo@85 560 .Ci fecd5d34f65597a4dfa16aeabea7d74b191532c3
meillo@85 561 .Ci 156d35f6425bea4c1ed3c4c79783dc613379c65b
meillo@85 562 .P
meillo@85 563 The configure switch
meillo@85 564 .Sw --enable-pop
meillo@85 565 activated the message retrieval facility.
meillo@85 566 The code area that would be conditionally compiled in for TLS and SASL
meillo@85 567 support had been small.
meillo@85 568 The conditionally compiled code area for POP support had been much larger.
meillo@85 569 Whereas the code base changes would only slightly change on toggling
meillo@85 570 TLS or SASL support, it changed much on toggling POP support.
meillo@85 571 The changes in the code base could hardly be overviewed.
meillo@85 572 By having POP support togglable a second code base had been created,
meillo@85 573 one that needed to be tested.
meillo@85 574 This situation is basically similar for the conditional TLS and SASL
meillo@85 575 code, but there the changes are minor and can yet be overviewed.
meillo@85 576 Still, conditional compilation of a code base creates variations
meillo@85 577 of the original program.
meillo@85 578 More variations require more testing and maintenance work.
meillo@85 579 .P
meillo@85 580 Two other options only specified default configuration values:
meillo@85 581 .Sw --with-mts=[smtp|sendmail]
meillo@85 582 defined the default transport service.
meillo@85 583 In mmh this fixed to
meillo@85 584 .Ar sendmail .
meillo@85 585 .Ci f6aa95b724fd8c791164abe7ee5468bf5c34f226
meillo@85 586 With
meillo@85 587 .Sw --with-smtpservers=[server1...]
meillo@85 588 default SMTP servers for the
meillo@85 589 .Ar smtp
meillo@85 590 transport service could be specified.
meillo@72 591 .Ci 128545e06224233b7e91fc4c83f8830252fe16c9
meillo@85 592 Both of them became irrelevant.
meillo@72 593
meillo@74 594 .U3 "Backup Prefix
meillo@74 595 .P
meillo@76 596 The backup prefix is the string that was prepended to message
meillo@76 597 filenames to tag them as deleted.
meillo@76 598 By default it had been the comma character `\f(CW,\fP'.
meillo@78 599 In July 2000, Kimmo Suominen introduced
meillo@78 600 the configure option
meillo@78 601 .Sw --with-hash-backup
meillo@78 602 to change the default to the hash symbol `\f(CW#\fP'.
meillo@78 603 The choice was probably personal preference, because first, the
meillo@78 604 option was named
meillo@78 605 .Sw --with-backup-prefix.
meillo@78 606 and had the prefix symbol as argument.
meillo@78 607 Because giving the hash symbol as argument caused to many problems
meillo@78 608 for configure,
meillo@78 609 the option was limited to use the hash symbol as the default prefix.
meillo@78 610 This makes me believe, that the choice for the hash was personal preference.
meillo@78 611 Being it related or not, words that start with the hash symbol
meillo@78 612 introduce a comment in the Unix shell.
meillo@72 613 Thus, the command line
meillo@72 614 .Cl "rm #13 #15
meillo@72 615 calls
meillo@72 616 .Pn rm
meillo@72 617 without arguments because the first hash symbol starts the comment
meillo@72 618 that reaches until the end of the line.
meillo@72 619 To delete the backup files,
meillo@72 620 .Cl "rm ./#13 ./#15"
meillo@72 621 needs to be used.
meillo@78 622 Using the hash as backup prefix can be seen as a precaution agains
meillo@78 623 data loss.
meillo@78 624 .P
meillo@72 625 I removed the configure option but added the profile entry
meillo@72 626 .Pe backup-prefix ,
meillo@72 627 which allows to specify an arbitrary string as backup prefix.
meillo@72 628 .Ci 6c40d481d661d532dd527eaf34cebb6d3f8ed086
meillo@76 629 Profile entries are the common method to change mmh's behavior.
meillo@76 630 This change did not remove the choice but moved it to a location where
meillo@72 631 it suited better.
meillo@76 632 .P
meillo@78 633 Eventually, however, the new trash folder concept
meillo@78 634 .Cf "Sec. XXX
meillo@78 635 obsoleted the concept of the backup prefix completely.
meillo@78 636 .Ci 8edc5aaf86f9f77124664f6801bc6c6cdf258173
meillo@72 637 (Well, there still are corner-cases to remove until the backup
meillo@72 638 prefix can be layed to rest, eventually.)
meillo@72 639 .\" FIXME: Do this work in the code!
meillo@76 640
meillo@76 641 .U3 "Editor and Pager
meillo@74 642 .P
meillo@74 643 The two configure options
meillo@74 644 .CW --with-editor=EDITOR
meillo@74 645 .CW --with-pager=PAGER
meillo@74 646 were used to specify the default editor and pager at configure time.
meillo@74 647 Doing so at configure time made sense in the Eighties,
meillo@76 648 when the set of available editors and pagers varied much across
meillo@76 649 different systems.
meillo@76 650 Today, the situation is more homegeneic.
meillo@74 651 The programs
meillo@74 652 .Pn vi
meillo@74 653 and
meillo@74 654 .Pn more
meillo@76 655 can be expected to be available on every Unix system,
meillo@74 656 as they are specified by POSIX since two decades.
meillo@74 657 (The specifications for
meillo@74 658 .Pn vi
meillo@74 659 and
meillo@74 660 .Pn more
meillo@74 661 appeared in
meillo@74 662 .[
meillo@74 663 posix 1987
meillo@74 664 .]
meillo@74 665 and,
meillo@74 666 .[
meillo@74 667 posix 1992
meillo@74 668 .]
meillo@74 669 respectively.)
meillo@74 670 As a first step, these two tools were hard-coded as defaults.
meillo@74 671 .Ci 5d43a99db70c12a673028c7758c20cbe3e13ef5f
meillo@74 672 Not changed were the
meillo@74 673 .Pe editor
meillo@74 674 and
meillo@74 675 .Pe moreproc
meillo@76 676 profile entries, which allowed the user to override the system defaults.
meillo@74 677 Later, the concept was reworked to respect the standard environment
meillo@74 678 variables
meillo@74 679 .Ev VISUAL
meillo@74 680 and
meillo@74 681 .Ev PAGER
meillo@76 682 if they are set.
meillo@74 683 Today, mmh determines the editor to use in the following order,
meillo@74 684 taking the first available and non-empty item:
meillo@74 685 .IP (1)
meillo@74 686 Environment variable
meillo@74 687 .Ev MMHEDITOR
meillo@74 688 .IP (2)
meillo@74 689 Profile entry
meillo@74 690 .Pe Editor
meillo@74 691 .IP (3)
meillo@74 692 Environment variable
meillo@74 693 .Ev VISUAL
meillo@74 694 .IP (4)
meillo@74 695 Environment variable
meillo@74 696 .Ev EDITOR
meillo@74 697 .IP (5)
meillo@74 698 Command
meillo@74 699 .Pn vi .
meillo@74 700 .P
meillo@76 701 .Ci f85f4b7ae62e3d05a945dcd46ead51f0a2a89a9b
meillo@76 702 .P
meillo@76 703 The pager to use is deteminded in a similar order,
meillo@74 704 also taking the first available and non-empty item:
meillo@74 705 .IP (1)
meillo@74 706 Environment variable
meillo@74 707 .Ev MMHPAGER
meillo@74 708 .IP (2)
meillo@74 709 Profile entry
meillo@74 710 .Pe Pager
meillo@74 711 (replaces
meillo@74 712 .Pe moreproc )
meillo@74 713 .IP (3)
meillo@74 714 Environment variable
meillo@74 715 .Ev PAGER
meillo@74 716 .IP (4)
meillo@74 717 Command
meillo@74 718 .Pn more .
meillo@74 719 .P
meillo@74 720 .Ci 0c4214ea2aec6497d0d67b436bbee9bc1d225f1e
meillo@74 721 .P
meillo@76 722 By respecting the
meillo@74 723 .Ev VISUAL /\c
meillo@74 724 .Ev EDITOR
meillo@74 725 and
meillo@74 726 .Ev PAGER
meillo@76 727 environment variables,
meillo@76 728 the new behavior confirms better to the common style on Unix systems.
meillo@76 729 Additionally, the new approach is more uniform and clearer to users.
meillo@72 730
meillo@74 731 .U3 "Locale
meillo@74 732 .P
meillo@74 733 The configure option
meillo@74 734 .Sw --disable-locale
meillo@85 735 was removed because POSIX provides locale support and there's
meillo@85 736 hardly any need to disable locale support.
meillo@74 737 .Ci ccf4f175ef4c4e7522f9510a4a1149c15d810dd9
meillo@72 738
meillo@76 739 .U3 "ndbm
meillo@72 740 .P
meillo@74 741 .Pn slocal
meillo@78 742 used to depend on
meillo@74 743 .I ndbm ,
meillo@74 744 a database library.
meillo@76 745 The database is used to store the `\fLMessage-ID\fP's of all
meillo@76 746 messages delivered.
meillo@74 747 This enables
meillo@74 748 .Pn slocal
meillo@74 749 to suppress delivering the same message to the same user twice.
meillo@74 750 (This features was enabled by the
meillo@74 751 .Sw -suppressdup
meillo@74 752 switch.)
meillo@74 753 .P
meillo@74 754 A variety of version of the database library exist.
meillo@78 755 .[
meillo@78 756 wolter unix incompat notes dbm
meillo@78 757 .]
meillo@74 758 Complicated autoconf code was needed to detect them correctly.
meillo@74 759 Further more, the configure switches
meillo@74 760 .Sw --with-ndbm=ARG
meillo@74 761 and
meillo@74 762 .Sw --with-ndbmheader=ARG
meillo@74 763 were added to help with difficult setups that would
meillo@78 764 not be detected automatically or correctly.
meillo@74 765 .P
meillo@74 766 By removing the suppress duplicates feature of
meillo@74 767 .Pn slocal ,
meillo@74 768 the dependency on
meillo@74 769 .I ndbm
meillo@78 770 vanished and 120 lines of complex autoconf code could be saved.
meillo@74 771 .Ci ecd6d6a20cb7a1507e3a20d6c4cb3a1cf14c6bbf
meillo@78 772 The change removed funtionality too, but that is minor to the
meillo@78 773 improvement by dropping the dependency and the complex autoconf code.
meillo@72 774
meillo@74 775 .U3 "mh-e Support
meillo@72 776 .P
meillo@74 777 The configure option
meillo@74 778 .Sw --disable-mhe
meillo@74 779 was removed when the mh-e support was reworked.
meillo@74 780 Mh-e is the Emacs front-end to MH.
meillo@76 781 It requires MH to provide minor additional functions.
meillo@76 782 The
meillo@76 783 .Sw --disable-mhe
meillo@76 784 configure option could switch these extensions off.
meillo@76 785 After removing the support for old versions of mh-e,
meillo@74 786 only the
meillo@74 787 .Sw -build
meillo@76 788 switches of
meillo@74 789 .Pn forw
meillo@74 790 and
meillo@74 791 .Pn repl
meillo@76 792 are left to be mh-e extensions.
meillo@76 793 They are now always built in because they add little code and complexity.
meillo@76 794 In consequence, the
meillo@74 795 .Sw --disable-mhe
meillo@76 796 configure option was removed
meillo@72 797 .Ci a7ce7b4a580d77b6c2c4d980812beb589aa4c643
meillo@74 798 Removing the option removed a second code setup that would have
meillo@74 799 needed to be tested.
meillo@76 800 This change was first done in nmh and thereafter merged into mmh.
meillo@76 801 .P
meillo@76 802 The interface changes in mmh require mh-e to be adjusted in order
meillo@76 803 to be able to use mmh as back-end.
meillo@76 804 This will require minor changes to mh-e, but removing the
meillo@76 805 .Sw -build
meillo@76 806 switches would require more rework.
meillo@72 807
meillo@74 808 .U3 "Masquerading
meillo@72 809 .P
meillo@74 810 The configure option
meillo@74 811 .Sw --enable-masquerade
meillo@76 812 could take up to three arguments:
meillo@76 813 `draft_from', `mmailid', and `username_extension'.
meillo@74 814 They activated different types of address masquerading.
meillo@74 815 All of them were implemented in the SMTP-speaking
meillo@74 816 .Pn post
meillo@76 817 command, which provided an MSA.
meillo@76 818 Address masquerading is an MTA's task and mmh does not cover
meillo@76 819 this field anymore.
meillo@76 820 Hence, true masquerading needs to be implemented in the external MTA.
meillo@74 821 .P
meillo@74 822 The
meillo@74 823 .I mmailid
meillo@74 824 masquerading type is the oldest one of the three and the only one
meillo@74 825 available in the original MH.
meillo@74 826 It provided a
meillo@74 827 .I username
meillo@74 828 to
meillo@74 829 .I fakeusername
meillo@76 830 mapping, based on the password file's GECOS field.
meillo@74 831 The man page
meillo@74 832 .Mp mh-tailor(5)
meillo@74 833 described the use case as being the following:
meillo@74 834 .QP
meillo@74 835 This is useful if you want the messages you send to always
meillo@74 836 appear to come from the name of an MTA alias rather than your
meillo@74 837 actual account name. For instance, many organizations set up
meillo@74 838 `First.Last' sendmail aliases for all users. If this is
meillo@74 839 the case, the GECOS field for each user should look like:
meillo@74 840 ``First [Middle] Last <First.Last>''
meillo@74 841 .P
meillo@74 842 As mmh sends outgoing mail via the local MTA only,
meillo@76 843 the best location to do such global rewrites is there.
meillo@74 844 Besides, the MTA is conceptionally the right location because it
meillo@74 845 does the reverse mapping for incoming mail (aliasing), too.
meillo@76 846 Further more, masquerading set up there is readily available for all
meillo@74 847 mail software on the system.
meillo@76 848 Hence, mmailid masquerading was removed.
meillo@74 849 .Ci 0836c8000ccb34b59410ef1c15b1b7feac70ce5f
meillo@74 850 .P
meillo@74 851 The
meillo@74 852 .I username_extension
meillo@76 853 masquerading type did not replace the username but would append a suffix,
meillo@76 854 specified by the
meillo@74 855 .Ev USERNAME_EXTENSION
meillo@76 856 environment variable, to it.
meillo@76 857 This provided support for the
meillo@74 858 .I user-extension
meillo@74 859 feature of qmail and the similar
meillo@74 860 .I "plussed user
meillo@74 861 processing of sendmail.
meillo@74 862 The decision to remove this username_extension masquerading was
meillo@74 863 motivated by the fact that
meillo@74 864 .Pn spost
meillo@76 865 hadn't supported it already.
meillo@76 866 .Ci 2abae0bfd0ad5bf898461e50aa4b466d641f23d9
meillo@76 867 Username extensions are possible in mmh, but less convenient to use.
meillo@76 868 .\" XXX format file %(getenv USERNAME_EXTENSION)
meillo@74 869 .P
meillo@74 870 The
meillo@74 871 .I draft_from
meillo@74 872 masquerading type instructed
meillo@74 873 .Pn post
meillo@84 874 to use the value of the
meillo@84 875 .Hd From
meillo@84 876 header field as SMTP envelope sender.
meillo@76 877 Sender addresses could be replaced completely.
meillo@74 878 .Ci b14ea6073f77b4359aaf3fddd0e105989db9
meillo@76 879 Mmh offers a kind of masquerading similar in effect, but
meillo@74 880 with technical differences.
meillo@76 881 As mmh does not transfer messages itself, the local MTA has final control
meillo@76 882 over the sender's address. Any masquerading mmh introduces may be reverted
meillo@76 883 by the MTA.
meillo@76 884 In times of pedantic spam checking, an MTA will take care to use
meillo@76 885 sensible envelope sender addresses to keep its own reputation up.
meillo@84 886 Nonetheless, the MUA can set the
meillo@84 887 .Hd From
meillo@84 888 header field and thereby propose
meillo@76 889 a sender address to the MTA.
meillo@74 890 The MTA may then decide to take that one or generate the canonical sender
meillo@74 891 address for use as envelope sender address.
meillo@74 892 .P
meillo@74 893 In mmh, the MTA will always extract the recipient and sender from the
meillo@84 894 message header (\c
meillo@74 895 .Pn sendmail 's
meillo@74 896 .Sw -t
meillo@74 897 switch).
meillo@84 898 The
meillo@84 899 .Hd From
meillo@84 900 header field of the draft may be set arbitrary by the user.
meillo@74 901 If it is missing, the canonical sender address will be generated by the MTA.
meillo@74 902
meillo@74 903 .U3 "Remaining Options
meillo@74 904 .P
meillo@74 905 Two configure options remain in mmh.
meillo@74 906 One is the locking method to use:
meillo@74 907 .Sw --with-locking=[dot|fcntl|flock|lockf] .
meillo@76 908 The idea of removing all methods except the portable dot locking
meillo@76 909 and having that one as the default is appealing, but this change
meillo@76 910 requires deeper technical investigation into the topic.
meillo@76 911 The other option,
meillo@74 912 .Sw --enable-debug ,
meillo@74 913 compiles the programs with debugging symbols and does not strip them.
meillo@74 914 This option is likely to stay.
meillo@72 915
meillo@72 916
meillo@58 917
meillo@63 918
meillo@58 919 .H2 "Removal of switches
meillo@58 920 .P
meillo@58 921
meillo@58 922
meillo@58 923
meillo@58 924
meillo@74 925 .H1 "Modernizing
meillo@58 926
meillo@58 927
meillo@58 928 .H2 "Removal of Code Relicts
meillo@0 929 .P
meillo@51 930 The code base of mmh originates from the late Seventies,
meillo@51 931 had been extensively
meillo@51 932 worked on in the mid Eighties, and had been partly reorganized and extended
meillo@51 933 in the Nineties. Relicts of all those times had gathered in the code base.
meillo@12 934 My goal was to remove any ancient code parts. One part of the task was
meillo@12 935 converting obsolete code constructs to standard constructs, the other part
meillo@12 936 was dropping obsolete functions.
meillo@12 937 .P
meillo@12 938 As I'm not even thirty years old and have no more than seven years of
meillo@51 939 Unix experience, I needed to learn about the history in retrospective.
meillo@51 940 Older people likely have used those ancient constructs themselves
meillo@51 941 and have suffered from their incompatibilities and have longed for
meillo@12 942 standardization. Unfortunately, I have only read that others had done so.
meillo@12 943 This put me in a much more difficult positions when working on the old
meillo@12 944 code. I needed to recherche what other would have known by heart from
meillo@12 945 experience. All my programming experience comes from a time past ANSI C
meillo@12 946 and past POSIX. Although I knew about the times before, I took the
meillo@51 947 current state implicitly for granted most of the time.
meillo@12 948 .P
meillo@12 949 Being aware of
meillo@12 950 these facts, I rather let people with more historic experience solve the
meillo@12 951 task of converting the ancient code constructs to standardized ones.
meillo@12 952 Luckily, Lyndon Nerenberg focused on this task at the nmh project.
meillo@12 953 He converted large parts of the code to POSIX constructs, removing
meillo@12 954 the conditionals compilation for now standardized features.
meillo@12 955 I'm thankful for this task being solved. I only pulled the changes into
meillo@12 956 mmh.
meillo@12 957 .P
meillo@20 958 The other task \(en dropping ancient functionality to remove old code \(en
meillo@12 959 I did myself, though. My position to strip mmh to the bare minimum of
meillo@12 960 frequently used features is much more revolutional than the nmh community
meillo@20 961 likes it. Without the need to justify my decisions, I was able to quickly
meillo@20 962 remove functionality I considered ancient.
meillo@20 963 The need to discuss my decisions with
meillo@20 964 peers likely would have slowed this process down. Of course, I researched
meillo@12 965 if a particular feature really should be dropped. Having not had any
meillo@12 966 contact to this feature within my computer life was a first indicator to
meillo@12 967 drop it, but I also asked others and searched the literature for modern
meillo@12 968 usage of the feature. If it appeared to be truly ancient, I dropped it.
meillo@12 969 The reason for dropping is always part of the commit message in the
meillo@12 970 version control system. Thus, it is easy for others to check their
meillo@12 971 view on the topic with mine and possibly to argue for reinclusion.
meillo@12 972
meillo@12 973 .U2 "MMDF maildrop support
meillo@12 974 .P
meillo@12 975 I did drop any support for the MMDF maildrop format. This type of format
meillo@12 976 is conceptionally similar to the mbox format, but uses four bytes with
meillo@12 977 value 1 (\fL^A^A^A^A\fP) as message delimiter,
meillo@18 978 instead of the string ``\fLFrom\ \fP''.
meillo@12 979 Due to the similarity and mbox being the de-facto standard maildrop
meillo@12 980 format on Unix, but also due to the larger influence of Sendmail than MMDF,
meillo@12 981 the MMDF maildrop format had vanished.
meillo@12 982 .P
meillo@12 983 The simplifications within the code were only moderate. Switches could
meillo@12 984 be removed from tools like
meillo@12 985 .L packf ,
meillo@12 986 which generate packed mailboxes. Only one packed mailbox format remained:
meillo@12 987 mbox.
meillo@12 988 The most important changes affect the equally named mail parsing routine in
meillo@12 989 .L sbr/m_getfld.c .
meillo@12 990 The direct MMDF code had been removed, but as now only one packed mailbox
meillo@12 991 format is left, code structure simplifications are likely possible.
meillo@12 992 The reason why they are still outstanding is the heavily optimized code
meillo@18 993 of
meillo@18 994 .Fu m_getfld() .
meillo@18 995 Changes beyond a small local scope \(en
meillo@12 996 which restructuring in its core is \(en cause a high risk of damaging
meillo@12 997 the intricate workings of the optimized code. This problem is know
meillo@12 998 to the developers of nmh, too. They also avoid touching this minefield
meillo@12 999 if possible.
meillo@12 1000
meillo@12 1001 .U2 "UUCP Bang Paths
meillo@12 1002 .P
meillo@12 1003 More questionably than the former topic is the removal of support for the
meillo@12 1004 UUCP bang path address style. However, the user may translate the bang
meillo@12 1005 paths on retrieval to Internet addresses and the other way on posting
meillo@12 1006 messages. The former can be done my an MDA like procmail; the latter
meillo@12 1007 by a sendmail wrapper. This would ensure that any address handling would
meillo@12 1008 work as expected. However, it might just work well without any
meillo@12 1009 such modifications, as mmh does not touch addresses much, in general.
meillo@12 1010 But I can't ensure as I have never used an environment with bang paths.
meillo@12 1011 Also, the behavior might break at any point in further development.
meillo@12 1012
meillo@12 1013 .U2 "Hardcopy terminal support
meillo@12 1014 .P
meillo@12 1015 More of a funny anecdote is the remaining of a check for printing to a
meillo@12 1016 hardcopy terminal until Spring 2012, when I finally removed it.
meillo@12 1017 I surely would be very happy to see such a terminal in action, maybe
meillo@12 1018 actually being able to work on it, but I fear my chances are null.
meillo@12 1019 .P
meillo@12 1020 The check only prevented a pager to be placed between the outputting
meillo@18 1021 program (\c
meillo@18 1022 .Pn mhl )
meillo@18 1023 and the terminal. This could have been ensured with
meillo@18 1024 the
meillo@82 1025 .Sw -nomoreproc
meillo@18 1026 at the command line statically, too.
meillo@12 1027
meillo@12 1028 .U2 "Removed support for header fields
meillo@12 1029 .P
meillo@84 1030 The
meillo@84 1031 .Hd Encrypted
meillo@84 1032 header field had been introduced by RFC\^822, but already
meillo@12 1033 marked legacy in RFC 2822. It was superseded by FIXME.
meillo@84 1034 Mmh does no more support this header field.
meillo@12 1035 .P
meillo@84 1036 Native support for
meillo@84 1037 .Hd Face
meillo@84 1038 header fields had been removed, as well.
meillo@84 1039 The feature is similar to the
meillo@84 1040 .Hd X-Face
meillo@84 1041 header field in its intent,
meillo@21 1042 but takes a different approach to store the image.
meillo@84 1043 Instead of encoding the image data directly into the header field,
meillo@84 1044 the it contains the hostname and UDP port where the image
meillo@21 1045 date could be retrieved.
meillo@84 1046 Neither
meillo@84 1047 .Hd X-Face
meillo@84 1048 nor the here described
meillo@84 1049 .Hd Face
meillo@84 1050 system
meillo@21 1051 \**
meillo@21 1052 .FS
meillo@21 1053 There is also a newer but different system, invented 2005,
meillo@84 1054 using
meillo@84 1055 .Hd Face
meillo@84 1056 headers.
meillo@84 1057 It is the successor of
meillo@84 1058 .Hd X-Face
meillo@84 1059 providing colored PNG images.
meillo@21 1060 .FE
meillo@21 1061 became well used in the large scale.
meillo@21 1062 It's still possible to use a Face systems,
meillo@21 1063 although mmh does not provide support for any of the different systems
meillo@21 1064 anymore. It's fairly easy to write a small shell script to
meillo@21 1065 extract the embedded or fetch the external Face data and display the image.
meillo@84 1066 Own
meillo@84 1067 .Hd Face
meillo@84 1068 header field can be added into the draft template files.
meillo@21 1069 .P
meillo@84 1070 .Hd Content-MD5
meillo@84 1071 header fields were introduced by RFC\^1864. They provide only
meillo@12 1072 a verification of data corruption during the transfer. By no means can
meillo@12 1073 they ensure verbatim end-to-end delivery of the contents. This is clearly
meillo@12 1074 stated in the RFC. The proper approach to provide verificationability
meillo@12 1075 of content in an end-to-end relationship is the use of digital cryptography
meillo@12 1076 (RFCs FIXME). On the other hand, transfer protocols should ensure the
meillo@12 1077 integrity of the transmission. In combinations these two approaches
meillo@84 1078 make the
meillo@84 1079 .Hd Content-MD5
meillo@84 1080 header field useless. In consequence, I removed
meillo@12 1081 the support for it. By this removal, MD5 computation is not needed
meillo@12 1082 anywhere in mmh. Hence, over 500 lines of code were removed by this one
meillo@84 1083 change. Even if the
meillo@84 1084 .Hd Content-MD5
meillo@84 1085 header field is useful sometimes,
meillo@12 1086 I value its usefulnes less than the improvement in maintainability, caused
meillo@12 1087 by the removal.
meillo@12 1088
meillo@20 1089 .U2 "Prompter's Control Keys
meillo@20 1090 .P
meillo@20 1091 The program
meillo@20 1092 .Pn prompter
meillo@20 1093 queries the user to fill in a message form. When used by
meillo@20 1094 .Pn comp
meillo@20 1095 as:
meillo@82 1096 .VS
meillo@82 1097 comp -editor prompter
meillo@82 1098 VE
meillo@20 1099 the resulting behavior is similar to
meillo@20 1100 .Pn mailx .
meillo@51 1101 Apparently,
meillo@20 1102 .Pn prompter
meillo@20 1103 hadn't been touched lately. Otherwise it's hardly explainable why it
meillo@20 1104 still offered the switches
meillo@84 1105 .Sw -erase
meillo@84 1106 .Ar chr
meillo@20 1107 and
meillo@84 1108 .Sw -kill
meillo@84 1109 .Ar chr
meillo@20 1110 to name the characters for command line editing.
meillo@21 1111 The times when this had been necessary are long time gone.
meillo@20 1112 Today these things work out-of-the-box, and if not, are configured
meillo@20 1113 with the standard tool
meillo@20 1114 .Pn stty .
meillo@20 1115
meillo@21 1116 .U2 "Vfork and Retry Loops
meillo@21 1117 .P
meillo@51 1118 MH creates many processes, which is a consequence of the tool chest approach.
meillo@21 1119 In earlier times
meillo@21 1120 .Fu fork()
meillo@21 1121 had been an expensive system call, as the process's whole image needed
meillo@21 1122 to be duplicated. One common case is replacing the image with
meillo@21 1123 .Fu exec()
meillo@21 1124 right after having forked the child process.
meillo@21 1125 To speed up this case, the
meillo@21 1126 .Fu vfork()
meillo@21 1127 system call was invented at Berkeley. It completely omits copying the
meillo@21 1128 image. If the image gets replaced right afterwards then unnecessary
meillo@21 1129 work is omited. On old systems this results in large speed ups.
meillo@21 1130 MH uses
meillo@21 1131 .Fu vfork()
meillo@21 1132 whenever possible.
meillo@21 1133 .P
meillo@21 1134 Memory management units that support copy-on-write semantics make
meillo@21 1135 .Fu fork()
meillo@21 1136 almost as fast as
meillo@21 1137 .Fu vfork()
meillo@21 1138 in the cases when they can be exchanged.
meillo@21 1139 With
meillo@21 1140 .Fu vfork()
meillo@51 1141 being more error-prone and hardly faster, it's preferable to simply
meillo@21 1142 use
meillo@21 1143 .Fu fork()
meillo@21 1144 instead.
meillo@21 1145 .P
meillo@21 1146 Related to the costs of
meillo@21 1147 .Fu fork()
meillo@21 1148 is the probability of its success.
meillo@21 1149 Today on modern systems, the system call will succeed almost always.
meillo@51 1150 In the Eighties on heavy loaded systems, as they were common at
meillo@21 1151 universities, this had been different. Thus, many of the
meillo@21 1152 .Fu fork()
meillo@21 1153 calls were wrapped into loops to retry to fork several times in
meillo@21 1154 short intervals, in case of previous failure.
meillo@21 1155 In mmh, the program aborts at once if the fork failed.
meillo@21 1156 The user can reexecute the command then. This is expected to be a
meillo@21 1157 very rare case on modern systems, especially personal ones, which are
meillo@21 1158 common today.
meillo@21 1159
meillo@12 1160
meillo@58 1161 .H2 "Attachments
meillo@22 1162 .P
meillo@58 1163 MIME
meillo@58 1164
meillo@58 1165
meillo@58 1166 .H2 "Digital Cryptography
meillo@22 1167 .P
meillo@58 1168 Signing and encryption.
meillo@58 1169
meillo@58 1170
meillo@58 1171 .H2 "Good Defaults
meillo@22 1172 .P
meillo@58 1173 foo
meillo@58 1174
meillo@58 1175
meillo@58 1176
meillo@58 1177
meillo@58 1178 .H1 "Code style
meillo@22 1179 .P
meillo@58 1180 foo
meillo@58 1181
meillo@58 1182
meillo@58 1183 .H2 "Standard Code
meillo@22 1184 .P
meillo@58 1185 POSIX
meillo@22 1186
meillo@22 1187
meillo@58 1188 .H2 "Separation
meillo@14 1189
meillo@58 1190 .U2 "MH Directory Split
meillo@0 1191 .P
meillo@19 1192 In MH and nmh, a personal setup had consisted of two parts:
meillo@19 1193 The MH profile, named
meillo@19 1194 .Fn \&.mh_profile
meillo@19 1195 and being located directly in the user's home directory.
meillo@19 1196 And the MH directory, where all his mail messages and also his personal
meillo@19 1197 forms, scan formats, other configuration files are stored. The location
meillo@19 1198 of this directory could be user-chosen. The default was to name it
meillo@19 1199 .Fn Mail
meillo@19 1200 and have it directly in the home directory.
meillo@19 1201 .P
meillo@19 1202 I've never liked the data storage and the configuration to be intermixed.
meillo@19 1203 They are different kinds of data. One part, are the messages,
meillo@19 1204 which are the data to operate on. The other part, are the personal
meillo@19 1205 configuration files, which are able to change the behavior of the operations.
meillo@19 1206 The actual operations are defined in the profile, however.
meillo@19 1207 .P
meillo@19 1208 When storing data, one should try to group data by its type.
meillo@19 1209 There's sense in the Unix file system hierarchy, where configuration
meillo@19 1210 file are stored separate (\c
meillo@19 1211 .Fn /etc )
meillo@19 1212 to the programs (\c
meillo@19 1213 .Fn /bin
meillo@19 1214 and
meillo@19 1215 .Fn /usr/bin )
meillo@19 1216 to their sources (\c
meillo@19 1217 .Fn /usr/src ).
meillo@19 1218 Such separation eases the backup management, for instance.
meillo@19 1219 .P
meillo@19 1220 In mmh, I've reorganized the file locations.
meillo@19 1221 Still there are two places:
meillo@19 1222 There's the mail storage directory, which, like in MH, contains all the
meillo@19 1223 messages, but, unlike in MH, nothing else.
meillo@19 1224 Its location still is user-chosen, with the default name
meillo@19 1225 .Fn Mail ,
meillo@19 1226 in the user's home directory. This is much similar to the case in nmh.
meillo@19 1227 The configuration files, however, are grouped together in the new directory
meillo@19 1228 .Fn \&.mmh
meillo@19 1229 in the user's home directory.
meillo@19 1230 The user's profile now is a file, named
meillo@19 1231 .Fn profile ,
meillo@19 1232 in this mmh directory.
meillo@19 1233 Consistently, the context file and all the personal forms, scan formats,
meillo@19 1234 and the like, are also there.
meillo@19 1235 .P
meillo@19 1236 The naming changed with the relocation.
meillo@19 1237 The directory where everything, except the profile, had been stored (\c
meillo@19 1238 .Fn $HOME/Mail ),
meillo@19 1239 used to be called \fIMH directory\fP. Now, this directory is called the
meillo@19 1240 user's \fImail storage\fP. The name \fImmh directory\fP is now given to
meillo@19 1241 the new directory
meillo@19 1242 (\c
meillo@19 1243 .Fn $HOME/.mmh ),
meillo@19 1244 containing all the personal configuration files.
meillo@19 1245 .P
meillo@19 1246 The separation of the files by type of content is logical and convenient.
meillo@19 1247 There are no functional differences as any possible setup known to me
meillo@19 1248 can be implemented with both approaches, although likely a bit easier
meillo@19 1249 with the new approach. The main goal of the change had been to provide
meillo@19 1250 sensible storage locations for any type of personal mmh file.
meillo@19 1251 .P
meillo@19 1252 In order for one user to have multiple MH setups, he can use the
meillo@19 1253 environment variable
meillo@19 1254 .Ev MH
meillo@19 1255 the point to a different profile file.
meillo@19 1256 The MH directory (mail storage plus personal configuration files) is
meillo@19 1257 defined by the
meillo@19 1258 .Pe Path
meillo@19 1259 profile entry.
meillo@19 1260 The context file could be defined by the
meillo@19 1261 .Pe context
meillo@19 1262 profile entry or by the
meillo@19 1263 .Ev MHCONTEXT
meillo@19 1264 environment variable.
meillo@19 1265 The latter is useful to have a distinct context (e.g. current folders)
meillo@19 1266 in each terminal window, for instance.
meillo@19 1267 In mmh, there are three environment variables now.
meillo@19 1268 .Ev MMH
meillo@19 1269 may be used to change the location of the mmh directory.
meillo@19 1270 .Ev MMHP
meillo@19 1271 and
meillo@19 1272 .Ev MMHC
meillo@19 1273 change the profile and context files, respectively.
meillo@19 1274 Besides providing a more consistent feel (which simply is the result
meillo@19 1275 of being designed anew), the set of personal configuration files can
meillo@19 1276 be chosen independently from the profile (including mail storage location)
meillo@19 1277 and context, now. Being it relevant for practical use or not, it
meillo@19 1278 de-facto is an improvement. However, the main achievement is the
meillo@19 1279 split between mail storage and personal configuration files.
meillo@17 1280
meillo@0 1281
meillo@58 1282 .H2 "Modularization
meillo@0 1283 .P
meillo@58 1284 whatnowproc
meillo@0 1285 .P
meillo@49 1286 The \fIMH library\fP
meillo@49 1287 .Fn libmh.a
meillo@49 1288 collects a bunch of standard functions that many of the MH tools need,
meillo@49 1289 like reading the profile or context files.
meillo@49 1290 This doesn't hurt the separation.
meillo@49 1291
meillo@58 1292
meillo@58 1293 .H2 "Style
meillo@58 1294 .P
meillo@58 1295 Code layout, goto, ...
meillo@58 1296
meillo@58 1297
meillo@58 1298
meillo@58 1299
meillo@58 1300 .H1 "Concept Exploitation/Homogeniety
meillo@58 1301
meillo@58 1302
meillo@58 1303 .H2 "Draft Folder
meillo@58 1304 .P
meillo@58 1305 Historically, MH provided exactly one draft message, named
meillo@58 1306 .Fn draft
meillo@58 1307 and
meillo@58 1308 being located in the MH directory. When starting to compose another message
meillo@58 1309 before the former one was sent, the user had been questioned whether to use,
meillo@58 1310 refile or replace the old draft. Working on multiple drafts at the same time
meillo@58 1311 was impossible. One could only work on them in alteration by refiling the
meillo@58 1312 previous one to some directory and fetching some other one for reediting.
meillo@58 1313 This manual draft management needed to be done each time the user wanted
meillo@58 1314 to switch between editing one draft to editing another.
meillo@58 1315 .P
meillo@58 1316 To allow true parallel editing of drafts, in a straight forward way, the
meillo@58 1317 draft folder facility exists. It had been introduced already in July 1984
meillo@58 1318 by Marshall T. Rose. The facility was deactivated by default.
meillo@58 1319 Even in nmh, the draft folder facility remained deactivated by default.
meillo@58 1320 At least, Richard Coleman added the man page
meillo@58 1321 .Mp mh-draft(5)
meillo@58 1322 to document
meillo@58 1323 the feature well.
meillo@58 1324 .P
meillo@58 1325 The only advantage of not using the draft folder facility is the static
meillo@58 1326 name of the draft file. This could be an issue for MH frontends like mh-e.
meillo@58 1327 But as they likely want to provide working on multiple drafts in parallel,
meillo@58 1328 the issue is only concerning compatibility. The aim of nmh to stay compatible
meillo@58 1329 prevented the default activation of the draft folder facility.
meillo@58 1330 .P
meillo@58 1331 On the other hand, a draft folder is the much more natural concept than
meillo@58 1332 a draft message. MH's mail storage consists of folders and messages,
meillo@58 1333 the messages named with ascending numbers. A draft message breaks with this
meillo@58 1334 concept by introducing a message in a file named
meillo@58 1335 .Fn draft .
meillo@58 1336 This draft
meillo@58 1337 message is special. It can not be simply listed with the available tools,
meillo@58 1338 but instead requires special switches. I.e. corner-cases were
meillo@58 1339 introduced. A draft folder, in contrast, does not introduce such
meillo@58 1340 corner-cases. The available tools can operate on the messages within that
meillo@58 1341 folder like on any messages within any mail folders. The only difference
meillo@58 1342 is the fact that the default folder for
meillo@58 1343 .Pn send
meillo@58 1344 is the draft folder,
meillo@58 1345 instead of the current folder, like for all other tools.
meillo@58 1346 .P
meillo@58 1347 The trivial part of the change was activating the draft folder facility
meillo@58 1348 by default and setting a default name for this folder. Obviously, I chose
meillo@58 1349 the name
meillo@58 1350 .Fn +drafts .
meillo@58 1351 This made the
meillo@82 1352 .Sw -draftfolder
meillo@58 1353 and
meillo@82 1354 .Sw -draftmessage
meillo@58 1355 switches useless, and I could remove them.
meillo@58 1356 The more difficult but also the part that showed the real improvement,
meillo@58 1357 was updating the tools to the new concept.
meillo@82 1358 .Sw -draft
meillo@58 1359 switches could
meillo@58 1360 be dropped, as operating on a draft message became indistinguishable to
meillo@58 1361 operating on any other message for the tools.
meillo@58 1362 .Pn comp
meillo@58 1363 still has its
meillo@82 1364 .Sw -use
meillo@58 1365 switch for switching between its two modes: (1) Compose a new
meillo@58 1366 draft, possibly by taking some existing message as a form. (2) Modify
meillo@58 1367 an existing draft. In either case, the behavior of
meillo@58 1368 .Pn comp is
meillo@58 1369 deterministic. There is no more need to query the user. I consider this
meillo@58 1370 a major improvement. By making
meillo@58 1371 .Pn send
meillo@58 1372 simply operate on the current
meillo@58 1373 message in the draft folder by default, with message and folder both
meillo@58 1374 overridable by specifying them on the command line, it is now possible
meillo@58 1375 to send a draft anywhere within the storage by simply specifying its folder
meillo@58 1376 and name.
meillo@58 1377 .P
meillo@58 1378 All theses changes converted special cases to regular cases, thus
meillo@58 1379 simplifying the tools and increasing the flexibility.
meillo@58 1380
meillo@58 1381
meillo@58 1382 .H2 "Trash Folder
meillo@58 1383 .P
meillo@58 1384 Similar to the situation for drafts is the situation for removed messages.
meillo@58 1385 Historically, a message was deleted by renaming. A specific
meillo@58 1386 \fIbackup prefix\fP, often comma (\c
meillo@58 1387 .Fn , )
meillo@58 1388 or hash (\c
meillo@58 1389 .Fn # ),
meillo@58 1390 being prepended to the file name. Thus, MH wouldn't recognize the file
meillo@58 1391 as a message anymore, as only files whose name consists of digits only
meillo@58 1392 are treated as messages. The removed messages remained as files in the
meillo@58 1393 same directory and needed some maintenance job to truly delete them after
meillo@58 1394 some grace time. Usually, by running a command similar to
meillo@82 1395 .VS
meillo@82 1396 find /home/user/Mail -ctime +7 -name ',*' | xargs rm
meillo@82 1397 VE
meillo@58 1398 in a cron job. Within the grace time interval
meillo@58 1399 the original message could be restored by stripping the
meillo@58 1400 the backup prefix from the file name. If however, the last message of
meillo@58 1401 a folder is been removed \(en say message
meillo@58 1402 .Fn 6
meillo@58 1403 becomes file
meillo@58 1404 .Fn ,6
meillo@58 1405 \(en and a new message enters the same folder, thus the same
meillo@58 1406 numbered being given again \(en in our case
meillo@58 1407 .Fn 6
meillo@58 1408 \(en, if that one
meillo@58 1409 is removed too, then the backup of the former message gets overwritten.
meillo@58 1410 Thus, the ability to restore removed messages does not only depend on
meillo@58 1411 the ``sweeping cron job'' but also on the removing of further messages.
meillo@58 1412 This is undesirable, because the real mechanism is hidden from the user
meillo@58 1413 and the consequences of further removals are not always obvious.
meillo@58 1414 Further more, the backup files are scattered within the whole mail
meillo@58 1415 storage, instead of being collected at one place.
meillo@58 1416 .P
meillo@58 1417 To improve the situation, the profile entry
meillo@58 1418 .Pe rmmproc
meillo@58 1419 (previously named
meillo@58 1420 .Pe Delete-Prog )
meillo@58 1421 was introduced, very early.
meillo@58 1422 It could be set to any command, which would care for the mail removal
meillo@58 1423 instead of taking the default action, described above.
meillo@58 1424 Refiling the to-be-removed files to some garbage folder was a common
meillo@58 1425 example. Nmh's man page
meillo@58 1426 .Mp rmm(1)
meillo@58 1427 proposes
meillo@58 1428 .Cl "refile +d
meillo@58 1429 to move messages to the garbage folder and
meillo@58 1430 .Cl "rm `mhpath +d all`
meillo@58 1431 the empty the garbage folder.
meillo@58 1432 Managing the message removal this way is a sane approach. It keeps
meillo@58 1433 the removed messages in one place, makes it easy to remove the backup
meillo@58 1434 files, and, most important, enables the user to use the tools of MH
meillo@58 1435 itself to operate on the removed messages. One can
meillo@58 1436 .Pn scan
meillo@58 1437 them,
meillo@58 1438 .Pn show
meillo@58 1439 them, and restore them with
meillo@58 1440 .Pn refile .
meillo@58 1441 There's no more
meillo@58 1442 need to use
meillo@58 1443 .Pn mhpath
meillo@58 1444 to switch over from MH tools to Unix tools \(en MH can do it all itself.
meillo@58 1445 .P
meillo@58 1446 This approach matches perfect with the concepts of MH, thus making
meillo@58 1447 it powerful. Hence, I made it the default. And even more, I also
meillo@58 1448 removed the old backup prefix approach, as it is clearly less powerful.
meillo@58 1449 Keeping unused alternative in the code is a bad choice as they likely
meillo@58 1450 gather bugs, by not being constantly tested. Also, the increased code
meillo@58 1451 size and more conditions crease the maintenance costs. By strictly
meillo@58 1452 converting to the trash folder approach, I simplified the code base.
meillo@58 1453 .Pn rmm
meillo@58 1454 calls
meillo@58 1455 .Pn refile
meillo@58 1456 internally to move the to-be-removed
meillo@58 1457 message to the trash folder (\c
meillo@58 1458 .Fn +trash
meillo@58 1459 by default). Messages
meillo@58 1460 there can be operated on like on any other message in the storage.
meillo@58 1461 The sweep clean, one can use
meillo@82 1462 .Cl "rmm -unlink +trash a" ,
meillo@58 1463 where the
meillo@82 1464 .Sw -unlink
meillo@58 1465 switch causes the files to be truly unliked instead
meillo@58 1466 of moved to the trash folder.
meillo@58 1467
meillo@58 1468
meillo@58 1469 .H2 "Path Notations
meillo@58 1470 .P
meillo@58 1471 foo
meillo@58 1472
meillo@58 1473
meillo@58 1474 .H2 "MIME Integration
meillo@58 1475 .P
meillo@58 1476 user-visible access to whole messages and MIME parts are inherently
meillo@58 1477 different
meillo@58 1478
meillo@58 1479
meillo@58 1480 .H2 "Of One Cast
meillo@58 1481 .P