1 I started reading over the SubmittingPatches document for Linux
2 kernel, primarily because I wanted to have a document similar to
3 it for the core GIT to make sure people understand what they are
4 doing when they write "Signed-off-by" line.
6 But the patch submission requirements are a lot more relaxed
7 here on the technical/contents front, because the core GIT is
8 thousand times smaller ;-). So here is only the relevant bits.
11 (1) Make separate commits for logically separate changes.
13 Unless your patch is really trivial, you should not be sending
14 out a patch that was generated between your working tree and
15 your commit head. Instead, always make a commit with complete
16 commit message and generate a series of patches from your
17 repository. It is a good discipline.
19 Describe the technical detail of the change(s).
21 If your description starts to get too long, that's a sign that you
22 probably need to split up your commit to finer grained pieces.
24 Oh, another thing. I am picky about whitespaces. Make sure your
25 changes do not trigger errors with the sample pre-commit hook shipped
26 in templates/hooks--pre-commit. To help ensure this does not happen,
27 run git diff --check on your changes before you commit.
30 (2) Generate your patch using git tools out of your commits.
32 git based diff tools (git, Cogito, and StGIT included) generate
33 unidiff which is the preferred format.
35 You do not have to be afraid to use -M option to "git diff" or
36 "git format-patch", if your patch involves file renames. The
37 receiving end can handle them just fine.
39 Please make sure your patch does not include any extra files
40 which do not belong in a patch submission. Make sure to review
41 your patch after generating it, to ensure accuracy. Before
42 sending out, please make sure it cleanly applies to the "master"
43 branch head. If you are preparing a work based on "next" branch,
44 that is fine, but please mark it as such.
47 (3) Sending your patches.
49 People on the git mailing list need to be able to read and
50 comment on the changes you are submitting. It is important for
51 a developer to be able to "quote" your changes, using standard
52 e-mail tools, so that they may comment on specific portions of
53 your code. For this reason, all patches should be submitted
54 "inline". WARNING: Be wary of your MUAs word-wrap
55 corrupting your patch. Do not cut-n-paste your patch; you can
56 lose tabs that way if you are not careful.
58 It is a common convention to prefix your subject line with
59 [PATCH]. This lets people easily distinguish patches from other
60 e-mail discussions.
62 "git format-patch" command follows the best current practice to
63 format the body of an e-mail message. At the beginning of the
64 patch should come your commit message, ending with the
65 Signed-off-by: lines, and a line that consists of three dashes,
66 followed by the diffstat information and the patch itself. If
67 you are forwarding a patch from somebody else, optionally, at
68 the beginning of the e-mail message just before the commit
69 message starts, you can put a "From: " line to name that person.
71 You often want to add additional explanation about the patch,
72 other than the commit message itself. Place such "cover letter"
73 material between the three dash lines and the diffstat.
75 Do not attach the patch as a MIME attachment, compressed or not.
76 Do not let your e-mail client send quoted-printable. Do not let
77 your e-mail client send format=flowed which would destroy
78 whitespaces in your patches. Many
79 popular e-mail applications will not always transmit a MIME
80 attachment as plain text, making it impossible to comment on
81 your code. A MIME attachment also takes a bit more time to
82 process. This does not decrease the likelihood of your
83 MIME-attached change being accepted, but it makes it more likely
84 that it will be postponed.
86 Exception: If your mailer is mangling patches then someone may ask
87 you to re-send them using MIME, that is OK.
89 Do not PGP sign your patch, at least for now. Most likely, your
90 maintainer or other people on the list would not have your PGP
91 key and would not bother obtaining it anyway. Your patch is not
92 judged by who you are; a good patch from an unknown origin has a
93 far better chance of being accepted than a patch from a known,
94 respected origin that is done poorly or does incorrect things.
96 If you really really really really want to do a PGP signed
97 patch, format it as "multipart/signed", not a text/plain message
98 that starts with '-----BEGIN PGP SIGNED MESSAGE-----'. That is
99 not a text/plain, it's something else.
101 Note that your maintainer does not necessarily read everything
102 on the git mailing list. If your patch is for discussion first,
103 send it "To:" the mailing list, and optionally "cc:" him. If it
104 is trivially correct or after the list reached a consensus, send
105 it "To:" the maintainer and optionally "cc:" the list.
107 Also note that your maintainer does not actively involve himself in
108 maintaining what are in contrib/ hierarchy. When you send fixes and
109 enhancements to them, do not forget to "cc: " the person who primarily
110 worked on that hierarchy in contrib/.
113 (4) Sign your work
115 To improve tracking of who did what, we've borrowed the
116 "sign-off" procedure from the Linux kernel project on patches
117 that are being emailed around. Although core GIT is a lot
118 smaller project it is a good discipline to follow it.
120 The sign-off is a simple line at the end of the explanation for
121 the patch, which certifies that you wrote it or otherwise have
122 the right to pass it on as a open-source patch. The rules are
123 pretty simple: if you can certify the below:
125 Developer's Certificate of Origin 1.1
127 By making a contribution to this project, I certify that:
129 (a) The contribution was created in whole or in part by me and I
130 have the right to submit it under the open source license
131 indicated in the file; or
133 (b) The contribution is based upon previous work that, to the best
134 of my knowledge, is covered under an appropriate open source
135 license and I have the right under that license to submit that
136 work with modifications, whether created in whole or in part
137 by me, under the same open source license (unless I am
138 permitted to submit under a different license), as indicated
139 in the file; or
141 (c) The contribution was provided directly to me by some other
142 person who certified (a), (b) or (c) and I have not modified
143 it.
145 (d) I understand and agree that this project and the contribution
146 are public and that a record of the contribution (including all
147 personal information I submit with it, including my sign-off) is
148 maintained indefinitely and may be redistributed consistent with
149 this project or the open source license(s) involved.
151 then you just add a line saying
153 Signed-off-by: Random J Developer <random@developer.example.org>
155 This line can be automatically added by git if you run the git-commit
156 command with the -s option.
158 Some people also put extra tags at the end. They'll just be ignored for
159 now, but you can do this to mark internal company procedures or just
160 point out some special detail about the sign-off.
163 ------------------------------------------------
164 MUA specific hints
166 Some of patches I receive or pick up from the list share common
167 patterns of breakage. Please make sure your MUA is set up
168 properly not to corrupt whitespaces. Here are two common ones
169 I have seen:
171 * Empty context lines that do not have _any_ whitespace.
173 * Non empty context lines that have one extra whitespace at the
174 beginning.
176 One test you could do yourself if your MUA is set up correctly is:
178 * Send the patch to yourself, exactly the way you would, except
179 To: and Cc: lines, which would not contain the list and
180 maintainer address.
182 * Save that patch to a file in UNIX mailbox format. Call it say
183 a.patch.
185 * Try to apply to the tip of the "master" branch from the
186 git.git public repository:
188 $ git fetch http://kernel.org/pub/scm/git/git.git master:test-apply
189 $ git checkout test-apply
190 $ git reset --hard
191 $ git applymbox a.patch
193 If it does not apply correctly, there can be various reasons.
195 * Your patch itself does not apply cleanly. That is _bad_ but
196 does not have much to do with your MUA. Please rebase the
197 patch appropriately.
199 * Your MUA corrupted your patch; applymbox would complain that
200 the patch does not apply. Look at .dotest/ subdirectory and
201 see what 'patch' file contains and check for the common
202 corruption patterns mentioned above.
204 * While you are at it, check what are in 'info' and
205 'final-commit' files as well. If what is in 'final-commit' is
206 not exactly what you would want to see in the commit log
207 message, it is very likely that your maintainer would end up
208 hand editing the log message when he applies your patch.
209 Things like "Hi, this is my first patch.\n", if you really
210 want to put in the patch e-mail, should come after the
211 three-dash line that signals the end of the commit message.
214 Pine
215 ----
217 (Johannes Schindelin)
219 I don't know how many people still use pine, but for those poor
220 souls it may be good to mention that the quell-flowed-text is
221 needed for recent versions.
223 ... the "no-strip-whitespace-before-send" option, too. AFAIK it
224 was introduced in 4.60.
226 (Linus Torvalds)
228 And 4.58 needs at least this.
230 ---
231 diff-tree 8326dd8350be64ac7fc805f6563a1d61ad10d32c (from e886a61f76edf5410573e92e38ce22974f9c40f1)
232 Author: Linus Torvalds <torvalds@g5.osdl.org>
233 Date: Mon Aug 15 17:23:51 2005 -0700
235 Fix pine whitespace-corruption bug
237 There's no excuse for unconditionally removing whitespace from
238 the pico buffers on close.
240 diff --git a/pico/pico.c b/pico/pico.c
241 --- a/pico/pico.c
242 +++ b/pico/pico.c
243 @@ -219,7 +219,9 @@ PICO *pm;
244 switch(pico_all_done){ /* prepare for/handle final events */
245 case COMP_EXIT : /* already confirmed */
246 packheader();
247 +#if 0
248 stripwhitespace();
249 +#endif
250 c |= COMP_EXIT;
251 break;
254 (Daniel Barkalow)
256 > A patch to SubmittingPatches, MUA specific help section for
257 > users of Pine 4.63 would be very much appreciated.
259 Ah, it looks like a recent version changed the default behavior to do the
260 right thing, and inverted the sense of the configuration option. (Either
261 that or Gentoo did it.) So you need to set the
262 "no-strip-whitespace-before-send" option, unless the option you have is
263 "strip-whitespace-before-send", in which case you should avoid checking
264 it.
267 Thunderbird
268 -----------
270 (A Large Angry SCM)
272 Here are some hints on how to successfully submit patches inline using
273 Thunderbird.
275 This recipe appears to work with the current [*1*] Thunderbird from Suse.
277 The following Thunderbird extensions are needed:
278 AboutConfig 0.5
279 http://aboutconfig.mozdev.org/
280 External Editor 0.7.2
281 http://globs.org/articles.php?lng=en&pg=8
283 1) Prepare the patch as a text file using your method of choice.
285 2) Before opening a compose window, use Edit->Account Settings to
286 uncheck the "Compose messages in HTML format" setting in the
287 "Composition & Addressing" panel of the account to be used to send the
288 patch. [*2*]
290 3) In the main Thunderbird window, _before_ you open the compose window
291 for the patch, use Tools->about:config to set the following to the
292 indicated values:
293 mailnews.send_plaintext_flowed => false
294 mailnews.wraplength => 0
296 4) Open a compose window and click the external editor icon.
298 5) In the external editor window, read in the patch file and exit the
299 editor normally.
301 6) Back in the compose window: Add whatever other text you wish to the
302 message, complete the addressing and subject fields, and press send.
304 7) Optionally, undo the about:config/account settings changes made in
305 steps 2 & 3.
308 [Footnotes]
309 *1* Version 1.0 (20041207) from the MozillaThunderbird-1.0-5 rpm of Suse
310 9.3 professional updates.
312 *2* It may be possible to do this with about:config and the following
313 settings but I haven't tried, yet.
314 mail.html_compose => false
315 mail.identity.default.compose_html => false
316 mail.identity.id?.compose_html => false
320 Gnus
321 ----
323 '|' in the *Summary* buffer can be used to pipe the current
324 message to an external program, and this is a handy way to drive
325 "git am". However, if the message is MIME encoded, what is
326 piped into the program is the representation you see in your
327 *Article* buffer after unwrapping MIME. This is often not what
328 you would want for two reasons. It tends to screw up non ASCII
329 characters (most notably in people's names), and also
330 whitespaces (fatal in patches). Running 'C-u g' to display the
331 message in raw form before using '|' to run the pipe can work
332 this problem around.