git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH] SubmittingPatches: clean up commit message tips

From
Junio C Hamano <gitster@pobox.com>
Date
Mar 9, 2011, 22:20 UTC
Message-ID
<7voc5k7x76.fsf@alter.siamese.dyndns.org>
In-Reply-To
<1299706069-5463-1-git-send-email-haircut@gmail.com>
Adam Monsen <haircut@gmail.com> writes:
> Removed uncommon acronyms in the "Checklist" section. I had to look up
> "iow" online, but this documentation should stand alone (without online
> access) as well as commit messages.

We have only a handful of instances of that phrase in the codebase (run "git grep" for them), so I am not strongly opposed to spelling it out, but use of IOW is quite common on this list---I suspect that it is largely because Linus uses the phrase quite often. Cf.

    $ git log | grep -e IOW -e '^Author: ' | grep -B1 -e IOW
> Leave wiggle room for including URLs in commit messages.

I don't think the updated text is too bad, but I don't very much like the above "wiggle room".

The guideline is written to suggest what you absolutely should include; it is obviously Ok to add other things as necessary. If common sense tells the reader external references will help recollection, the guideline does not forbid to include them. IOW, there are enough wiggle rooms already.

> Modify the section about trivial changes slightly... it makes more sense
> that it is discouraging diffs pasted in emails as opposed to patches
> generated with "git am".
s/am/format-patch/;
> Remove recommendations on commit messages from the "Make separate
> commits for logically separate changes" section,...
> sentence about breaking apart big commits into separate patches.

While I do not particularly hate this part, I think people who did the "Checklist vs Long Version" meant to make each of them stand on its own. Lazy people (or people who think they are experienced enough) read the former, while the others who pride themselves being thorough will skip the "for-lazy-people" digest version and read only "the real thing".

So overall, I am not enthused by this version. Input from others may be appreciated.

Previous: Adam MonsenNext: Adam Monsen
Message 21 of 26 in “frustrated forensics: hard to find diff that undid a fix”
  1. Adam MonsenMar 5, 2011
  2. Jonathan del StrotherMar 5, 2011
  3. Jakub NarebskiMar 5, 2011
  4. Jonathan NiederMar 5, 2011
  5. Jeff KingMar 5, 2011
  6. Adam MonsenMar 5, 2011
  7. 0/2 improve combined diff documentationAdam Monsen, Mar 5, 2011
  8. 1/2 documentation fix: git log -p does not imply -c.Adam Monsen, Mar 5, 2011
  9. Junio C HamanoMar 7, 2011
  10. Jeff KingMar 7, 2011
  11. Junio C HamanoMar 7, 2011
  12. Jeff KingMar 7, 2011
  13. Documentation fix: git log -p does not imply -c.Adam Monsen, Mar 7, 2011
  14. Junio C HamanoMar 8, 2011
  15. Documentation fix: git log -p does not imply -c.Adam Monsen, Mar 8, 2011
  16. Junio C HamanoMar 8, 2011
  17. Adam MonsenMar 8, 2011
  18. Junio C HamanoMar 9, 2011
  19. Adam MonsenMar 9, 2011
  20. SubmittingPatches: clean up commit message tipsAdam Monsen, Mar 9, 2011
  21. Junio C HamanoMar 9, 2011
  22. diff format documentation: clarify --cc and -cAdam Monsen, Mar 8, 2011
  23. diff format documentation: clarify --cc and -cAdam Monsen, Mar 8, 2011
  24. Jeff KingMar 8, 2011
  25. 2/2 English grammar fixes for combined diff doc.Adam Monsen, Mar 5, 2011
  26. Martin von ZweigbergkMar 5, 2011

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.