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

Re: [PATCH 2/5] refs: split log_ref_write logic into log_ref_setup

From
Erick Mattos <erick.mattos@gmail.com>
Date
Jun 2, 2010, 23:16 UTC
Message-ID
<AANLkTikcQny9Es_cMJTg94qTJZD2s7T37h_Hur9Lt-Lv@mail.gmail.com>
In-Reply-To
<7vzkzdcneh.fsf@alter.siamese.dyndns.org>
Hi,
2010/6/2 Junio C Hamano <gitster@pobox.com>:
Show 12 quoted lines
> Erick Mattos <erick.mattos@gmail.com> writes:
>
>> Now just a question, Junio:
>>
>> I forgot to sign-off those patches, should I have to send them again?
>
> I would have appreciated a whole re-send while I was too distracted by
> non-git stuff during the past few weeks, but now I am more or less settled
> in, it's Ok to just send a separate "Signed-off-by:" like this one:
>
>    http://article.gmane.org/gmane.comp.version-control.git/148230/raw
>
All right, following then:
2010/5/21 Erick Mattos <erick.mattos@gmail.com>:
Show 41 quoted lines
> These series of patches are improvements to 'git checkout --orphan'.
>
> The main reason for them is a corner case which is not being solved by
> actual implementation.  As it is a quite improbable situation and as it was
> necessary to do more extensive changes to support it then its development
> was held to be presented in a new developing cycle.
>
> When someone set core.logAllRefUpdates to false reflogs are not created
> automatically.  This behavior is superseeded by -l option.  Actually this is
> not allowed with --orphan by current implementation.  Those new patches are
> made to fix that.
>
> There are also two other patches for configuring completion in bash and to
> enhance documentation.
>
> To be completely honest I don't see a point of not having the reflogs
> created and deleted automatically so I see no reason for -l and
> core.logAllRefUpdates at all.  But I do not like to do anything partially
> thus these new patches.  If someone could show me a case please do it.  ;-)
>
> [PATCH 1/5] Documentation: alter checkout --orphan description
>
> This one improves documentation text by late corrections from previous
> threads.
>
> [PATCH 2/5] refs: split log_ref_write logic into log_ref_setup
>
> Prepare the field by separating the logic to set up the reflog from the
> reflog writing action.
>
> [PATCH 3/5] checkout --orphan: respect -l option always
>
> This is the actual actor.
>
> [PATCH 4/5] t3200: test -l with core.logAllRefUpdates options
>
> Adjusting scripts to test everything extensively.
>
> [PATCH 5/5] bash completion: add --orphan to 'git checkout'
>
> Just do that git change.
Signed-off-by: Erick Mattos <erick.mattos@gmail.com>
Previous: Junio C HamanoNext: Erick Mattos
Message 7 of 18 in “checkout --orphan improvements”
  1. 0/5 checkout --orphan improvementsErick Mattos, May 22, 2010
  2. 1/5 Documentation: alter checkout --orphan descriptionErick Mattos, May 22, 2010
  3. 2/5 refs: split log_ref_write logic into log_ref_setupErick Mattos, May 22, 2010
  4. Junio C HamanoMay 26, 2010
  5. Erick MattosMay 26, 2010
  6. Junio C HamanoJun 2, 2010
  7. Erick MattosJun 2, 2010
  8. 3/5 checkout --orphan: respect -l option alwaysErick Mattos, May 22, 2010
  9. Junio C HamanoMay 26, 2010
  10. Erick MattosMay 26, 2010
  11. Erik Faye-LundMay 26, 2010
  12. Erick MattosMay 26, 2010
  13. Erick MattosJun 3, 2010
  14. Michael J GruberMay 26, 2010
  15. Erick MattosMay 26, 2010
  16. Michael J GruberMay 27, 2010
  17. 4/5 t3200: test -l with core.logAllRefUpdates optionsErick Mattos, May 22, 2010
  18. 5/5 bash completion: add --orphan to 'git checkout'Erick Mattos, May 22, 2010

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.