docs/master: discussion.roff annotate

docs/master

annotate discussion.roff @ 157:0cce17978f0a

Wrote about signing and encrypting. Not finished yet.

author	markus schnalke <meillo@marmaro.de>
date	Sat, 07 Jul 2012 18:41:47 +0200
parents	75f20c925999
children	8b411125645d

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meillo@87	21 I believe that the development of all-in-one mail systems is obsolete.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