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

Re: [RFC/PATCH] commit notes workflow

From
Drew Northup <drew.northup@maine.edu>
Date
Mar 2, 2011, 12:45 UTC
Message-ID
<1299069921.17973.26.camel@drew-northup.unet.maine.edu>
In-Reply-To
<4D6DEB64.1080003@gmail.com>
On Wed, 2011-03-02 at 20:01 +1300, Chris Packham wrote:
Show 29 quoted lines
> On 02/03/11 10:59, Jeff King wrote:
> > On Fri, Feb 25, 2011 at 04:58:22PM +0100, Johan Herland wrote:
> >> Maybe we should use a slightly more verbose separator (i.e. more 
> >> unlikely to trigger false positives). As you say, we already have to 
> >> watch out for "---" because of "am", but that only applies to projects 
> >> that _use_ "am" (i.e. mailing-list-centric projects like git.git and 
> >> the Linux kernel). Other projects (e.g. github-centric projects or most 
> >> centralized "$dayjob-style" projects) seldom or never use "am" at all, 
> >> so I wouldn't expect those developers think of "---" as "special" in 
> >> any way.
> >>
> >> What about using something like "--- Notes ---" instead?
> > 
> > Yeah, it is true that many git users will never care about the
> > patch-through-mail workflow. And I think these days that is OK, because
> > rebase will take care to keep their commit message intact even if it
> > doesn't format well in a "format-patch | am" pipeline.
> > 
> > I really wanted to keep it short and natural, though. Because eventually
> > I'd like to have this on all the time via a config option, and I don't
> > want to see "--- Notes ---" in every commit that doesn't have notes. But
> > I _do_ want to be able to quickly say "oh, let me make a note on this"
> > and just add a quick separator.
> 
> <bikesheding>
> What about "#---"? Satisfies the quick to type and is a lot less likely
> to appear in commit messages. Not sure about the implications of finding
> that string before the commit message is stripped.
> </bikesheding>

True enough, but that would be seen as a comment and dropped outright the way things are currently standing. If you want short, definitely rare, and most likely intentional you'd need something harder to remember like "-!N" as the tag. I don't know how well that'd go over with people--it definitely isn't natural.

-- 
-Drew Northup
________________________________________________
"As opposed to vegetable or mineral error?"
-John Pescatore, SANS NewsBites Vol. 12 Num. 59
Previous: Chris PackhamNext: Piotr Krukowiecki
Message 17 of 27 in “commit notes workflow”
  1. commit notes workflowJeff King, Feb 25, 2011
  2. Johan HerlandFeb 25, 2011
  3. Jeff KingMar 1, 2011
  4. Johan HerlandMar 2, 2011
  5. Sverre RabbelierMar 3, 2011
  6. Junio C HamanoMar 3, 2011
  7. Sverre RabbelierMar 3, 2011
  8. commit, status: #comment diff output in verbose modeIan Ward Comfort, Mar 3, 2011
  9. Sverre RabbelierMar 3, 2011
  10. Jeff KingMar 7, 2011
  11. 1/2 notes: make expand_notes_ref globally accessibleJeff King, Mar 7, 2011
  12. Johan HerlandMar 8, 2011
  13. 2/2 commit: allow editing notes in commit message editorJeff King, Mar 7, 2011
  14. Johan HerlandMar 8, 2011
  15. Michel LespinasseMar 8, 2011
  16. Chris PackhamMar 2, 2011
  17. Drew NorthupMar 2, 2011
  18. Piotr KrukowieckiMar 2, 2011
  19. Junio C HamanoFeb 25, 2011
  20. Drew NorthupFeb 25, 2011
  21. Jeff KingMar 1, 2011
  22. Drew NorthupMar 1, 2011
  23. Jeff KingMar 1, 2011
  24. Drew NorthupMar 1, 2011
  25. Michael J GruberFeb 27, 2011
  26. Jeff KingMar 1, 2011
  27. Yann DirsonMar 9, 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.