]> diplodocus.org Git - nmh/blob - docs/README.developers
Added const to arg of m_backup().
[nmh] / docs / README.developers
1 #
2 # README.developers
3 #
4
5 This file is intended to provide a few tips for anyone doing development on nmh.
6 Developers who learn things "the hard way" about the nmh codebase (as opposed to
7 local info best encoded in a comment) are encouraged to share their wisdom here.
8
9 Following a commit checklist, the topics are organized alphabetically.
10
11 ----------------
12 commit checklist
13 ----------------
14
15 1. code updated?
16 2. test added?
17 3. make distcheck passed?
18 4. man page and other documentation updated?
19 5. docs/pending-release-notes updated?
20 6. should commit message reference bug report?
21 7. update/close bug report (with commit id)?
22 8. notify nmh-users?
23
24
25 ---------------------------------
26 C library/system call usage notes
27 ---------------------------------
28 * Use m_mktemp2() or m_mktemp() instead of mkstemp(3) (see section on
29 nmh temporary files below).
30 * Use m_unlink() instead of unlink(3).
31 * Use done() instead of _exit(3) except after a fork(3).
32
33
34 -------------------------
35 autoconf & automake files
36 -------------------------
37
38 If you wish to change the `configure' script, the generated Makefile
39 or other related files, you'll need to first install GNU m4, available
40 from <ftp://ftp.gnu.org/pub/gnu/m4/>, then GNU autoconf
41 (<ftp://ftp.gnu.org/pub/gnu/autoconf/>) and GNU automake
42 (<ftp://ftp.gnu.org/pub/gnu/automake/>). Nmh is currently using a
43 minimum of autoconf 2.68 and automake 1.12.
44
45 Most of the configure-related files are automatically generated.
46 The only files you should need to manually edit are configure.ac
47 and any autoconf macros in the m4 directory. Don't, for instance,
48 edit config.h.in. Though it is an input file from the point of
49 view of the users (and the configure script) it is an output file
50 from the point of view of the developers (and the autoconf script).
51
52 If you wish to add a new autoconf macro, it should be placed in it's
53 own file and put in the m4 directory; aclocal will automatically pick
54 it up and automake will add it to the distribution target automatically.
55
56 If you wish to make changes to the Makefile, you will need to edit
57 Makefile.am. See the automake documentation if you need further help.
58 You should always check changes to Makefile.am by using "make distcheck".
59
60 Note that the automatically generated autotools files (such as config.h.in,
61 Makefile.in, and configure), are NOT kept in git. Thus, when you check out
62 a git tree, you need to run the autogen.sh script before you can build
63 anything:
64
65 % ./autogen.sh
66
67
68 -------------------
69 directory structure
70 -------------------
71
72 Following is a list of nmh's directories along with a brief description of the
73 purpose of each one. Meanings are given for the abbreviations, but note that
74 these meanings are just informed guesses as to what the MH developers were
75 thinking.
76
77 ./
78 The top-level directory. Contains files like README and INSTALL.
79
80 config/
81 Contains utility files for the `configure' process. Ordinarily nothing in
82 here needs to be messed with.
83
84 docs/
85 Contains more specialized documentation, such as this file and
86 the FAQ.
87
88 etc/
89 Contains files, file templates, and scripts to generate files that will be
90 installed in the ${prefix}/etc directory. Stuff like replcomps.
91
92 h/
93 Most of nmh's header (.h) files are kept not in the individual source
94 directories, but in this central location.
95
96 man/
97 Contains all the input files that are processed to generate nmh's manual
98 pages.
99
100 mts/
101 "mts" stands for "Message Transfer Service". Source files specific to the
102 different MTSs go in the subdirectories.
103
104 mts/smtp/
105 When nmh is configured to just talk to an SMTP server over TCP/IP, the
106 source in this directory is compiled.
107
108 sbr/
109 "sbr" stands for "subroutine(s)". For the most part, each source file in
110 this directory contains a single function with the same name as the source
111 file. These functions are of general use and are called from throughout
112 nmh.
113
114 SPECS/
115 Contains files such as RPM specs.
116
117 test/
118 The num unit test suite.
119
120 tools/
121 "tools" contains tools, scripts, and supporting files used by the
122 developers while writing, debugging, and testing the code.
123
124 uip/
125 "uip" stands for "User Interface Programs". Most nmh commands have a file
126 in this directory named <command>.c containing the code for that command
127 (e.g. repl.c). In some cases there is also an auxiliary file called
128 <command>sbr.c which contains additional subroutines called from <command>.c
129 (which would contain not much else besides main()).
130
131
132 ---
133 git
134 ---
135
136 As of December 2010, nmh has switched to using git for revision control
137 instead of CVS. While the topic of git is beyond the scope of this FAQ,
138 to get started with git & nmh, you can run the following command to checkout
139 the nmh repository (with read-only access to it):
140
141 % git clone git://git.savannah.nongnu.org/nmh.git
142
143 That will create a workspace called nmh. To update that workspace
144 with changes to the master, cd to it and run:
145
146 % git pull
147
148 If you are a project member and want write access to the repository,
149 you'll have to checkout with the following command instead of the one
150 above:
151
152 % git clone <username>@git.sv.nongnu.org:/srv/git/nmh.git
153
154 We suggest using git pull --rebase instead of the default merge for
155 git pull. If you don't want to add the --rebase option every time,
156 you can tell git pull to always rebase in your nmh workspace by
157 cd'ing to it and running the following command:
158
159 % git config --bool branch.master.rebase true
160
161 And you'll probably want the following, also, so that --rebase applies
162 to any new branches that you create:
163
164 % git config branch.autosetuprebase always
165
166
167 -------------------------------------------------------
168 nmh-local functions to use in preference to OS versions
169 -------------------------------------------------------
170
171 For some system functions whose availability or behavior varies from OS to OS,
172 nmh conditionally uses a local definition with the same name as the OS function
173 (e.g. snprintf()). For other functions, developers need to avoid the OS
174 versions and always use the nmh-supplied function. Here is a list of such
175 functions:
176
177 OS function nmh-local version to use instead
178 =========== ================================
179 getpass() nmh_getpass()
180
181
182 -------------------
183 nmh temporary files
184 -------------------
185
186 To create a temporary file, use m_mktemp2() or m_mktemp(). They use
187 mkstemp(3), but they also register the temporary file for removal on
188 program termination. So, do not use mkstemp() directly.
189
190 To further support this, nmh_init() must be called at the beginning of
191 main(). And, if a child process is not going to immediately call one
192 of the exec(3) functions or _exit(3) after a fork(3), it should call
193 unregister_for_removal(0). Finally, nmh_init() sets up signal handlers
194 for several signals: these signal handlers should not be disabled.
195
196
197 --------------
198 nmh test suite
199 --------------
200
201 The nmh test suite is run through the Makefile, with "make check"
202 or "make distcheck".
203
204 In the nmh test suite, nmh programs to be tested should be invoked
205 through the run_test or run_prog shell functions defined in
206 test/common.sh.
207
208 Instead of echoing test progress, use start_test()/finish_test()
209 from tests/common.sh. These will report the particular test name,
210 within the test, only if there is a failure.
211
212 To enable the use of valgrind, where available, set the environment
213 variable NMH_VALGRIND to a non-null value. However, a separate
214 environment variable, VALGRIND_ME, triggers the use of valgrind in
215 test/inc/test-eom-align because it greatly extends the duration of
216 that test.
217
218 If valgrind complains about "serious error when reading debuginfo"
219 from a library, either update or remove the debuginfo package for
220 the offending library.
221
222
223 -------------
224 releasing nmh
225 -------------
226
227 To make a public release of nmh (we'll use version 1.5 as the example
228 here; the convention for release candidates is to use something like
229 "1.5-RC1"):
230
231 1. Create a release branch. The convention is to name release branches
232 with the name "<version>-release".
233
234 % git branch 1.5-release
235
236 Note you are still on the master branch at this point. Mark the
237 current revision as the branchpoint for the new release branch:
238
239 % git tag -a -m "This tag marks the point where we started the branch for 1.5" 1.5-branchpoint
240
241 Now mark the master branch with a post-release version number (the
242 convention here is to use VERSION+dev as the version number).
243
244 % echo 1.5+dev > VERSION
245 % git commit VERSION
246 % git push
247 % git push --tags
248
249 Then do:
250
251 % git checkout 1.5-release
252
253 You are now on the 1.5 release branch.
254
255 2. % echo 1.5 > VERSION
256 % date +"%e %B %Y" > DATE
257 (DATE should contain something like "30 December 2000")
258
259 3. % git commit VERSION DATE; git push
260
261 4. % git tag -a 1.5 -m 'Releasing nmh-1.5.'
262 % git push --tags
263
264 Note that the new convention for tagging is to simply tag with the
265 version number (tag formats in the past have varied).
266
267 5. % make distcheck
268
269 If you want to check the distribution build with some particular
270 configure options, set the DISTCHECK_CONFIGURE_FLAGS variable.
271 E.g.:
272
273 % make distcheck DISTCHECK_CONFIGURE_FLAGS=--with-cyrus-sasl
274
275 6. Upload the distribution file to savannah. You can automate this process
276 by doing:
277
278 % make upload SAVANNAH_USERNAME=username
279
280 This will automatically call gpg to sign the release. You can bypass
281 this step by setting the SKIP_GPG_SIG variable.
282
283 7. Update the http://www.nongnu.org/nmh/ homepage. (It lives in the CVS
284 'webpages repository'; see https://savannah.nongnu.org/cvs/?group=nmh)
285
286 8. Add a news item to the savannah nmh page. You'll have to submit it first
287 and then separately approve it (under News->Manage).
288
289 9. Send the release announcement email to the following places:
290 nmh-workers@nongnu.org
291 nmh-announce@nongnu.org
292 exmh-users@redhat.com
293 exmh-workers@redhat.com
294 mh-e-users@lists.sourceforge.net
295
296 If the release fixes significant security holes, also send an announcement
297 to bugtraq@securityfocus.com. The exmh lists require you to be subscribed
298 in order to post. Note that you don't need to post separately to
299 comp.mail.mh, as the mh-users mailing list is apparently bidirectionally
300 gatewayed to it.
301
302 Preferably, the announcement should contain the MD5 hash generated above,
303 and should be PGP-signed. It should include the URL for the tarball as
304 well as the URL of the website. It should contain a brief summary of
305 visible changes, as well as the URL of the git diff page that would show
306 a detailed list of changes. The changes between 1.5 and 1.4 would be
307 shown by [this is just a guess, I don't know anything about cgit, and
308 it assumes that we tag with nmh-x_x-release from now on]:
309
310 http://git.savannah.gnu.org/cgit/nmh.git/diff/?h=nmh-1_5-release?h=nmh-1_4-release