Documentation / SubmittingPatcheson commit Merge branch 'jc/everyday-markup' into maint (7974b3f)
   1Here are some guidelines for people who want to contribute their code
   2to this software.
   3
   4(0) Decide what to base your work on.
   5
   6In general, always base your work on the oldest branch that your
   7change is relevant to.
   8
   9 - A bugfix should be based on 'maint' in general. If the bug is not
  10   present in 'maint', base it on 'master'. For a bug that's not yet
  11   in 'master', find the topic that introduces the regression, and
  12   base your work on the tip of the topic.
  13
  14 - A new feature should be based on 'master' in general. If the new
  15   feature depends on a topic that is in 'pu', but not in 'master',
  16   base your work on the tip of that topic.
  17
  18 - Corrections and enhancements to a topic not yet in 'master' should
  19   be based on the tip of that topic. If the topic has not been merged
  20   to 'next', it's alright to add a note to squash minor corrections
  21   into the series.
  22
  23 - In the exceptional case that a new feature depends on several topics
  24   not in 'master', start working on 'next' or 'pu' privately and send
  25   out patches for discussion. Before the final merge, you may have to
  26   wait until some of the dependent topics graduate to 'master', and
  27   rebase your work.
  28
  29 - Some parts of the system have dedicated maintainers with their own
  30   repositories (see the section "Subsystems" below).  Changes to
  31   these parts should be based on their trees.
  32
  33To find the tip of a topic branch, run "git log --first-parent
  34master..pu" and look for the merge commit. The second parent of this
  35commit is the tip of the topic branch.
  36
  37(1) Make separate commits for logically separate changes.
  38
  39Unless your patch is really trivial, you should not be sending
  40out a patch that was generated between your working tree and
  41your commit head.  Instead, always make a commit with complete
  42commit message and generate a series of patches from your
  43repository.  It is a good discipline.
  44
  45Give an explanation for the change(s) that is detailed enough so
  46that people can judge if it is good thing to do, without reading
  47the actual patch text to determine how well the code does what
  48the explanation promises to do.
  49
  50If your description starts to get too long, that's a sign that you
  51probably need to split up your commit to finer grained pieces.
  52That being said, patches which plainly describe the things that
  53help reviewers check the patch, and future maintainers understand
  54the code, are the most beautiful patches.  Descriptions that summarise
  55the point in the subject well, and describe the motivation for the
  56change, the approach taken by the change, and if relevant how this
  57differs substantially from the prior version, are all good things
  58to have.
  59
  60Make sure that you have tests for the bug you are fixing.  See
  61t/README for guidance.
  62
  63When adding a new feature, make sure that you have new tests to show
  64the feature triggers the new behaviour when it should, and to show the
  65feature does not trigger when it shouldn't.  Also make sure that the
  66test suite passes after your commit.  Do not forget to update the
  67documentation to describe the updated behaviour.
  68
  69Speaking of the documentation, it is currently a liberal mixture of US
  70and UK English norms for spelling and grammar, which is somewhat
  71unfortunate.  A huge patch that touches the files all over the place
  72only to correct the inconsistency is not welcome, though.  Potential
  73clashes with other changes that can result from such a patch are not
  74worth it.  We prefer to gradually reconcile the inconsistencies in
  75favor of US English, with small and easily digestible patches, as a
  76side effect of doing some other real work in the vicinity (e.g.
  77rewriting a paragraph for clarity, while turning en_UK spelling to
  78en_US).  Obvious typographical fixes are much more welcomed ("teh ->
  79"the"), preferably submitted as independent patches separate from
  80other documentation changes.
  81
  82Oh, another thing.  We are picky about whitespaces.  Make sure your
  83changes do not trigger errors with the sample pre-commit hook shipped
  84in templates/hooks--pre-commit.  To help ensure this does not happen,
  85run git diff --check on your changes before you commit.
  86
  87
  88(2) Describe your changes well.
  89
  90The first line of the commit message should be a short description (50
  91characters is the soft limit, see DISCUSSION in git-commit(1)), and
  92should skip the full stop.  It is also conventional in most cases to
  93prefix the first line with "area: " where the area is a filename or
  94identifier for the general area of the code being modified, e.g.
  95
  96  . archive: ustar header checksum is computed unsigned
  97  . git-cherry-pick.txt: clarify the use of revision range notation
  98
  99If in doubt which identifier to use, run "git log --no-merges" on the
 100files you are modifying to see the current conventions.
 101
 102The body should provide a meaningful commit message, which:
 103
 104  . explains the problem the change tries to solve, iow, what is wrong
 105    with the current code without the change.
 106
 107  . justifies the way the change solves the problem, iow, why the
 108    result with the change is better.
 109
 110  . alternate solutions considered but discarded, if any.
 111
 112Describe your changes in imperative mood, e.g. "make xyzzy do frotz"
 113instead of "[This patch] makes xyzzy do frotz" or "[I] changed xyzzy
 114to do frotz", as if you are giving orders to the codebase to change
 115its behaviour.  Try to make sure your explanation can be understood
 116without external resources. Instead of giving a URL to a mailing list
 117archive, summarize the relevant points of the discussion.
 118
 119
 120(3) Generate your patch using Git tools out of your commits.
 121
 122Git based diff tools generate unidiff which is the preferred format.
 123
 124You do not have to be afraid to use -M option to "git diff" or
 125"git format-patch", if your patch involves file renames.  The
 126receiving end can handle them just fine.
 127
 128Please make sure your patch does not add commented out debugging code,
 129or include any extra files which do not relate to what your patch
 130is trying to achieve. Make sure to review
 131your patch after generating it, to ensure accuracy.  Before
 132sending out, please make sure it cleanly applies to the "master"
 133branch head.  If you are preparing a work based on "next" branch,
 134that is fine, but please mark it as such.
 135
 136
 137(4) Sending your patches.
 138
 139Learn to use format-patch and send-email if possible.  These commands
 140are optimized for the workflow of sending patches, avoiding many ways
 141your existing e-mail client that is optimized for "multipart/*" mime
 142type e-mails to corrupt and render your patches unusable.
 143
 144People on the Git mailing list need to be able to read and
 145comment on the changes you are submitting.  It is important for
 146a developer to be able to "quote" your changes, using standard
 147e-mail tools, so that they may comment on specific portions of
 148your code.  For this reason, each patch should be submitted
 149"inline" in a separate message.
 150
 151Multiple related patches should be grouped into their own e-mail
 152thread to help readers find all parts of the series.  To that end,
 153send them as replies to either an additional "cover letter" message
 154(see below), the first patch, or the respective preceding patch.
 155
 156If your log message (including your name on the
 157Signed-off-by line) is not writable in ASCII, make sure that
 158you send off a message in the correct encoding.
 159
 160WARNING: Be wary of your MUAs word-wrap
 161corrupting your patch.  Do not cut-n-paste your patch; you can
 162lose tabs that way if you are not careful.
 163
 164It is a common convention to prefix your subject line with
 165[PATCH].  This lets people easily distinguish patches from other
 166e-mail discussions.  Use of additional markers after PATCH and
 167the closing bracket to mark the nature of the patch is also
 168encouraged.  E.g. [PATCH/RFC] is often used when the patch is
 169not ready to be applied but it is for discussion, [PATCH v2],
 170[PATCH v3] etc. are often seen when you are sending an update to
 171what you have previously sent.
 172
 173"git format-patch" command follows the best current practice to
 174format the body of an e-mail message.  At the beginning of the
 175patch should come your commit message, ending with the
 176Signed-off-by: lines, and a line that consists of three dashes,
 177followed by the diffstat information and the patch itself.  If
 178you are forwarding a patch from somebody else, optionally, at
 179the beginning of the e-mail message just before the commit
 180message starts, you can put a "From: " line to name that person.
 181
 182You often want to add additional explanation about the patch,
 183other than the commit message itself.  Place such "cover letter"
 184material between the three-dash line and the diffstat.  For
 185patches requiring multiple iterations of review and discussion,
 186an explanation of changes between each iteration can be kept in
 187Git-notes and inserted automatically following the three-dash
 188line via `git format-patch --notes`.
 189
 190Do not attach the patch as a MIME attachment, compressed or not.
 191Do not let your e-mail client send quoted-printable.  Do not let
 192your e-mail client send format=flowed which would destroy
 193whitespaces in your patches. Many
 194popular e-mail applications will not always transmit a MIME
 195attachment as plain text, making it impossible to comment on
 196your code.  A MIME attachment also takes a bit more time to
 197process.  This does not decrease the likelihood of your
 198MIME-attached change being accepted, but it makes it more likely
 199that it will be postponed.
 200
 201Exception:  If your mailer is mangling patches then someone may ask
 202you to re-send them using MIME, that is OK.
 203
 204Do not PGP sign your patch, at least for now.  Most likely, your
 205maintainer or other people on the list would not have your PGP
 206key and would not bother obtaining it anyway.  Your patch is not
 207judged by who you are; a good patch from an unknown origin has a
 208far better chance of being accepted than a patch from a known,
 209respected origin that is done poorly or does incorrect things.
 210
 211If you really really really really want to do a PGP signed
 212patch, format it as "multipart/signed", not a text/plain message
 213that starts with '-----BEGIN PGP SIGNED MESSAGE-----'.  That is
 214not a text/plain, it's something else.
 215
 216Send your patch with "To:" set to the mailing list, with "cc:" listing
 217people who are involved in the area you are touching (the output from
 218"git blame $path" and "git shortlog --no-merges $path" would help to
 219identify them), to solicit comments and reviews.
 220
 221After the list reached a consensus that it is a good idea to apply the
 222patch, re-send it with "To:" set to the maintainer [*1*] and "cc:" the
 223list [*2*] for inclusion.
 224
 225Do not forget to add trailers such as "Acked-by:", "Reviewed-by:" and
 226"Tested-by:" lines as necessary to credit people who helped your
 227patch.
 228
 229    [Addresses]
 230     *1* The current maintainer: gitster@pobox.com
 231     *2* The mailing list: git@vger.kernel.org
 232
 233
 234(5) Sign your work
 235
 236To improve tracking of who did what, we've borrowed the
 237"sign-off" procedure from the Linux kernel project on patches
 238that are being emailed around.  Although core Git is a lot
 239smaller project it is a good discipline to follow it.
 240
 241The sign-off is a simple line at the end of the explanation for
 242the patch, which certifies that you wrote it or otherwise have
 243the right to pass it on as a open-source patch.  The rules are
 244pretty simple: if you can certify the below:
 245
 246        Developer's Certificate of Origin 1.1
 247
 248        By making a contribution to this project, I certify that:
 249
 250        (a) The contribution was created in whole or in part by me and I
 251            have the right to submit it under the open source license
 252            indicated in the file; or
 253
 254        (b) The contribution is based upon previous work that, to the best
 255            of my knowledge, is covered under an appropriate open source
 256            license and I have the right under that license to submit that
 257            work with modifications, whether created in whole or in part
 258            by me, under the same open source license (unless I am
 259            permitted to submit under a different license), as indicated
 260            in the file; or
 261
 262        (c) The contribution was provided directly to me by some other
 263            person who certified (a), (b) or (c) and I have not modified
 264            it.
 265
 266        (d) I understand and agree that this project and the contribution
 267            are public and that a record of the contribution (including all
 268            personal information I submit with it, including my sign-off) is
 269            maintained indefinitely and may be redistributed consistent with
 270            this project or the open source license(s) involved.
 271
 272then you just add a line saying
 273
 274        Signed-off-by: Random J Developer <random@developer.example.org>
 275
 276This line can be automatically added by Git if you run the git-commit
 277command with the -s option.
 278
 279Notice that you can place your own Signed-off-by: line when
 280forwarding somebody else's patch with the above rules for
 281D-C-O.  Indeed you are encouraged to do so.  Do not forget to
 282place an in-body "From: " line at the beginning to properly attribute
 283the change to its true author (see (2) above).
 284
 285Also notice that a real name is used in the Signed-off-by: line. Please
 286don't hide your real name.
 287
 288If you like, you can put extra tags at the end:
 289
 2901. "Reported-by:" is used to credit someone who found the bug that
 291   the patch attempts to fix.
 2922. "Acked-by:" says that the person who is more familiar with the area
 293   the patch attempts to modify liked the patch.
 2943. "Reviewed-by:", unlike the other tags, can only be offered by the
 295   reviewer and means that she is completely satisfied that the patch
 296   is ready for application.  It is usually offered only after a
 297   detailed review.
 2984. "Tested-by:" is used to indicate that the person applied the patch
 299   and found it to have the desired effect.
 300
 301You can also create your own tag or use one that's in common usage
 302such as "Thanks-to:", "Based-on-patch-by:", or "Mentored-by:".
 303
 304------------------------------------------------
 305Subsystems with dedicated maintainers
 306
 307Some parts of the system have dedicated maintainers with their own
 308repositories.
 309
 310 - git-gui/ comes from git-gui project, maintained by Pat Thoyts:
 311
 312        git://repo.or.cz/git-gui.git
 313
 314 - gitk-git/ comes from Paul Mackerras's gitk project:
 315
 316        git://ozlabs.org/~paulus/gitk
 317
 318 - po/ comes from the localization coordinator, Jiang Xin:
 319
 320        https://github.com/git-l10n/git-po/
 321
 322Patches to these parts should be based on their trees.
 323
 324------------------------------------------------
 325An ideal patch flow
 326
 327Here is an ideal patch flow for this project the current maintainer
 328suggests to the contributors:
 329
 330 (0) You come up with an itch.  You code it up.
 331
 332 (1) Send it to the list and cc people who may need to know about
 333     the change.
 334
 335     The people who may need to know are the ones whose code you
 336     are butchering.  These people happen to be the ones who are
 337     most likely to be knowledgeable enough to help you, but
 338     they have no obligation to help you (i.e. you ask for help,
 339     don't demand).  "git log -p -- $area_you_are_modifying" would
 340     help you find out who they are.
 341
 342 (2) You get comments and suggestions for improvements.  You may
 343     even get them in a "on top of your change" patch form.
 344
 345 (3) Polish, refine, and re-send to the list and the people who
 346     spend their time to improve your patch.  Go back to step (2).
 347
 348 (4) The list forms consensus that the last round of your patch is
 349     good.  Send it to the maintainer and cc the list.
 350
 351 (5) A topic branch is created with the patch and is merged to 'next',
 352     and cooked further and eventually graduates to 'master'.
 353
 354In any time between the (2)-(3) cycle, the maintainer may pick it up
 355from the list and queue it to 'pu', in order to make it easier for
 356people play with it without having to pick up and apply the patch to
 357their trees themselves.
 358
 359------------------------------------------------
 360Know the status of your patch after submission
 361
 362* You can use Git itself to find out when your patch is merged in
 363  master. 'git pull --rebase' will automatically skip already-applied
 364  patches, and will let you know. This works only if you rebase on top
 365  of the branch in which your patch has been merged (i.e. it will not
 366  tell you if your patch is merged in pu if you rebase on top of
 367  master).
 368
 369* Read the Git mailing list, the maintainer regularly posts messages
 370  entitled "What's cooking in git.git" and "What's in git.git" giving
 371  the status of various proposed changes.
 372
 373------------------------------------------------
 374MUA specific hints
 375
 376Some of patches I receive or pick up from the list share common
 377patterns of breakage.  Please make sure your MUA is set up
 378properly not to corrupt whitespaces.
 379
 380See the DISCUSSION section of git-format-patch(1) for hints on
 381checking your patch by mailing it to yourself and applying with
 382git-am(1).
 383
 384While you are at it, check the resulting commit log message from
 385a trial run of applying the patch.  If what is in the resulting
 386commit is not exactly what you would want to see, it is very
 387likely that your maintainer would end up hand editing the log
 388message when he applies your patch.  Things like "Hi, this is my
 389first patch.\n", if you really want to put in the patch e-mail,
 390should come after the three-dash line that signals the end of the
 391commit message.
 392
 393
 394Pine
 395----
 396
 397(Johannes Schindelin)
 398
 399I don't know how many people still use pine, but for those poor
 400souls it may be good to mention that the quell-flowed-text is
 401needed for recent versions.
 402
 403... the "no-strip-whitespace-before-send" option, too. AFAIK it
 404was introduced in 4.60.
 405
 406(Linus Torvalds)
 407
 408And 4.58 needs at least this.
 409
 410---
 411diff-tree 8326dd8350be64ac7fc805f6563a1d61ad10d32c (from e886a61f76edf5410573e92e38ce22974f9c40f1)
 412Author: Linus Torvalds <torvalds@g5.osdl.org>
 413Date:   Mon Aug 15 17:23:51 2005 -0700
 414
 415    Fix pine whitespace-corruption bug
 416
 417    There's no excuse for unconditionally removing whitespace from
 418    the pico buffers on close.
 419
 420diff --git a/pico/pico.c b/pico/pico.c
 421--- a/pico/pico.c
 422+++ b/pico/pico.c
 423@@ -219,7 +219,9 @@ PICO *pm;
 424            switch(pico_all_done){      /* prepare for/handle final events */
 425              case COMP_EXIT :          /* already confirmed */
 426                packheader();
 427+#if 0
 428                stripwhitespace();
 429+#endif
 430                c |= COMP_EXIT;
 431                break;
 432
 433
 434(Daniel Barkalow)
 435
 436> A patch to SubmittingPatches, MUA specific help section for
 437> users of Pine 4.63 would be very much appreciated.
 438
 439Ah, it looks like a recent version changed the default behavior to do the
 440right thing, and inverted the sense of the configuration option. (Either
 441that or Gentoo did it.) So you need to set the
 442"no-strip-whitespace-before-send" option, unless the option you have is
 443"strip-whitespace-before-send", in which case you should avoid checking
 444it.
 445
 446
 447Thunderbird, KMail, GMail
 448-------------------------
 449
 450See the MUA-SPECIFIC HINTS section of git-format-patch(1).
 451
 452Gnus
 453----
 454
 455'|' in the *Summary* buffer can be used to pipe the current
 456message to an external program, and this is a handy way to drive
 457"git am".  However, if the message is MIME encoded, what is
 458piped into the program is the representation you see in your
 459*Article* buffer after unwrapping MIME.  This is often not what
 460you would want for two reasons.  It tends to screw up non ASCII
 461characters (most notably in people's names), and also
 462whitespaces (fatal in patches).  Running 'C-u g' to display the
 463message in raw form before using '|' to run the pipe can work
 464this problem around.