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

Re: [PATCH jh/notes-merge-in-git-dir-worktree] fixup! t3310 on Windows

From
Junio C Hamano <gitster@pobox.com>
Date
Mar 14, 2012, 17:44 UTC
Message-ID
<7vlin3qdpt.fsf@alter.siamese.dyndns.org>
In-Reply-To
<CALKQrgdWZM959OyrEp+WCCehczZmMA3K8_RAcf23aAczKBCfvA@mail.gmail.com>
Johan Herland <johan@herland.net> writes:
Show 14 quoted lines
> On Wed, Mar 14, 2012 at 12:59, Johannes Sixt <j.sixt@viscovery.net> wrote:
> ...
>> I doubt that the use-case that is tested here makes sense.
>
> As David wrote, the use case is likely to pop up among regular users.
> We can't simply ignore it.
>
>> Or .git/NOTES_MERGE_WORKTREE should not be removed. Would it be an option
>> to clear it out only when it is needed, right before it is filled again?
>
> Maybe, but then we wouldn't be able to warn or abort in the case where
> there is a previous unfinished notes merge, and the user tries to
> start a new notes merge. Instead, we'd silently overwrite the previous
> unfinished notes merge...
We cannot simply ignore it.
You _could_ do is perhaps to do something like:
 - when you notice that you need to ask the user to hand-merge a temporary
   file (i.e. when 'notes merge' is run), check if .git/N_M_W/done exists.
    - If you see that marker file, remove it everything in the
      directory, deposit the temporary file you want the user to edit,
      and expect the user to later tell you that he is done by "notes
      merge --commit";
    - If you don't, the user didn't give you "notes merge --commit"
      during the earlier run.  Tell him to first abandon the previous
      round with "notes merge --abandon" or something.
 - "notes merge --commit" will just mark .git/N_M_W/ with "done".
 - "notes merge --abandon" will remove .git/N_M_W/.

but it only supports the simplest "Run 'notes merge', cd to .git/N_M_W, edit all, run 'notes merge --commit'" workflow. If the user decides to abandon instead of commit, he will be inside the problematic directory, so trying to be more elaborate like the above does not really solve anything fundamental. The user has to come out of that temporary directory eventually anyway.

So I think the following is not just an improvement over the current code, but probably an acceptable solution, given its simplicity.

> Maybe it's better to simply detect if cwd is inside
> .git/NOTES_MERGE_WORKTREE, and then abort, telling the user to chdir
> out before trying again?
Previous: Johan HerlandNext: Johan Herland
Message 20 of 50 in “read_directory() rewrite to support struct pathspec”
  1. 00/11 read_directory() rewrite to support struct pathspecNguyễn Thái Ngọc Duy, Oct 24, 2011
  2. 01/11 Introduce "check-attr --excluded" as a replacement for "add --ignore-missing"Nguyễn Thái Ngọc Duy, Oct 24, 2011
  3. Junio C HamanoOct 27, 2011
  4. Nguyen Thai Ngoc DuyOct 28, 2011
  5. 02/11 notes-merge: use opendir/readdir instead of using read_directory()Nguyễn Thái Ngọc Duy, Oct 24, 2011
  6. Junio C HamanoOct 25, 2011
  7. Nguyen Thai Ngoc DuyOct 26, 2011
  8. Junio C HamanoOct 26, 2011
  9. Nguyen Thai Ngoc DuyOct 27, 2011
  10. Junio C HamanoOct 27, 2011
  11. Nguyen Thai Ngoc DuyOct 28, 2011
  12. 1/2 t3310: Add testcase demonstrating failure to --commit from within another dirJohan Herland, Mar 12, 2012
  13. 2/2 notes-merge: use opendir/readdir instead of using read_directory()Johan Herland, Mar 12, 2012
  14. Nguyen Thai Ngoc DuyMar 12, 2012
  15. fixup! t3310 on WindowsJohannes Sixt, Mar 14, 2012
  16. Johan HerlandMar 14, 2012
  17. Johannes SixtMar 14, 2012
  18. David BremnerMar 14, 2012
  19. Johan HerlandMar 14, 2012
  20. Junio C HamanoMar 14, 2012
  21. 3/2 notes-merge: Don't remove .git/NOTES_MERGE_WORKTREE; it may be the user's cwdJohan Herland, Mar 14, 2012
  22. Junio C HamanoMar 15, 2012
  23. Junio C HamanoMar 15, 2012
  24. Johan HerlandMar 15, 2012
  25. Re* [PATCH 3/2] notes-merge: Don't remove .git/NOTES_MERGE_WORKTREE; it may be the user's cwdJunio C Hamano, Mar 15, 2012
  26. Junio C HamanoMar 15, 2012
  27. Johannes SixtMar 15, 2012
  28. 03/11 t5403: avoid doing "git add foo/bar" where foo/.git existsNguyễn Thái Ngọc Duy, Oct 24, 2011
  29. Junio C HamanoOct 25, 2011
  30. Nguyen Thai Ngoc DuyOct 26, 2011
  31. Junio C HamanoOct 26, 2011
  32. Nguyen Thai Ngoc DuyOct 27, 2011
  33. Junio C HamanoOct 27, 2011
  34. Nguyen Thai Ngoc DuyOct 30, 2011
  35. Junio C HamanoOct 30, 2011
  36. Nguyen Thai Ngoc DuyOct 30, 2011
  37. Junio C HamanoOct 30, 2011
  38. 04/11 tree-walk.c: do not leak internal structure in tree_entry_len()Nguyễn Thái Ngọc Duy, Oct 24, 2011
  39. Junio C HamanoOct 25, 2011
  40. 05/11 symbolize return values of tree_entry_interesting()Nguyễn Thái Ngọc Duy, Oct 24, 2011
  41. Junio C HamanoOct 25, 2011
  42. Junio C HamanoOct 27, 2011
  43. Nguyen Thai Ngoc DuyOct 30, 2011
  44. 06/11 read_directory_recursive: reduce one indentation levelNguyễn Thái Ngọc Duy, Oct 24, 2011
  45. 07/11 tree_entry_interesting: make use of local pointer "item"Nguyễn Thái Ngọc Duy, Oct 24, 2011
  46. 08/11 tree-walk: mark useful pathspecsNguyễn Thái Ngọc Duy, Oct 24, 2011
  47. 09/11 tree_entry_interesting: differentiate partial vs full matchNguyễn Thái Ngọc Duy, Oct 24, 2011
  48. 10/11 read-dir: stop using path_simplify code in favor of tree_entry_interesting()Nguyễn Thái Ngọc Duy, Oct 24, 2011
  49. 11/11 dir.c: remove dead code after read_directory() rewriteNguyễn Thái Ngọc Duy, Oct 24, 2011
  50. Junio C HamanoOct 24, 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.