Log messages in CVS

Log messages in CVS

[Eli Zaretskii]

This fragment from "cvs log":

  date: 2007-11-20 00:57:08 +0000;  author: dann;  state: Exp;  lines: +4 -0;  commitid: qqjdnHpVbIePxfGs;
  * progmodes/idlw-help.el: Require browse-url unconditionally, it
  is available by default.
  (idlwave-help-browse-url-available): Change default to t.

  * emulation/edt.el (defgroup, defcustom): Remove definition.
  (eval-when-compile): Remove.
  (c-mark-function):
  * textmodes/reftex-dcr.el (bibtex-beginning-of-entry):
  * textmodes/fill.el (comment-search-forward)
  (comment-string-strip):
  * progmodes/prolog.el (comint-mode, comint-send-string)
  (comint-send-region, comint-send-eof):
  * progmodes/dcl-mode.el (imenu-default-create-index-function):
  * emulation/viper-util.el (viper-forward-Word):
  * emulation/vi.el (c-mark-function):
  * emulation/edt-vt100.el (vt100-wide-mode):
  * emacs-lisp/timer.el (diary-entry-time): Declare as functions.

  * url-mailto.el (mail-send-and-exit):
  * url-http.el (url-dav-file-attributes):
  * url-file.el (ange-ftp-set-passwd, ange-ftp-copy-file-internal):
  Declare as functions.

  * url-privacy.el (url-device-type): Define unconditionally.

indicates that several files were committed with one "cvs ci" command,
which is (AFAIK) against the Emacs development policies.

Re: Log messages in CVS

[Tassilo Horn]

Eli Zaretskii <eliz@gnu.org> writes:

Hi Eli,

> This fragment from "cvs log":

[...]

> indicates that several files were committed with one "cvs ci" command,
> which is (AFAIK) against the Emacs development policies.

Oh, I think I did that a few days ago, too, unless PCL-CVS commits each
file separately if many are marked.  What are the pros of committing
each file separately?  I can only see the drawback that with that
convention inconsistent states are introduced in the repository, if the
changes of one file depend on changes in another file.

Anyway, where can I read up the emacs development policies?  I grepped
the whole source tree for "cvs c" or "commit" but couldn't find
something relevant.

Bye,
Tassilo

Re: Log messages in CVS

[Eli Zaretskii]

> From: Tassilo Horn <tassilo@member.fsf.org>
> Date: Sat, 29 Dec 2007 14:12:44 +0100
> 
> Anyway, where can I read up the emacs development policies?  I grepped
> the whole source tree for "cvs c" or "commit" but couldn't find
> something relevant.

I once posted a summary that I know about; see:

  http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00229.html
  http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00234.html
  http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00312.html

Richard commented here, basically approving my summary:

  http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00276.html

I suggested at the time to have these guidelines under admin/ but I
guess no one volunteered to do that yet.

Re: Log messages in CVS

[Tassilo Horn]

Eli Zaretskii <eliz@gnu.org> writes:

Hi Eli,

> I once posted a summary that I know about; see:
>
>   http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00229.html
>   http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00234.html
>   http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00312.html
>
> Richard commented here, basically approving my summary:
>
>   http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00276.html

Ok, I've read the thread and that Richard agreed with the rules.  But on
the other hand others like Miles disagreed.  Is the only purpose of
those single file commits really to keep the CVS logs clean?

Bye,
Tassilo

Re: Log messages in CVS

[Eli Zaretskii]

> From: Tassilo Horn <tassilo@member.fsf.org>
> Date: Sat, 29 Dec 2007 20:45:05 +0100
> Cc: emacs-devel@gnu.org
> 
> > I once posted a summary that I know about; see:
> >
> >   http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00229.html
> >   http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00234.html
> >   http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00312.html
> >
> > Richard commented here, basically approving my summary:
> >
> >   http://lists.gnu.org/archive/html/emacs-devel/2006-11/msg00276.html
> 
> Ok, I've read the thread and that Richard agreed with the rules.  But on
> the other hand others like Miles disagreed.  Is the only purpose of
> those single file commits really to keep the CVS logs clean?

I'll let Richard reply to this question, as well as to disagreements
by Miles and others.  Unlike some others, I don't care much about
whether files should or should not be committed one by one, as long as
whatever practice we adopt is adhered to by everybody.

Re: Log messages in CVS

[Miles Bader]

Eli Zaretskii <eliz@gnu.org> writes:
>> Ok, I've read the thread and that Richard agreed with the rules.  But on
>> the other hand others like Miles disagreed.  Is the only purpose of
>> those single file commits really to keep the CVS logs clean?
>
> I'll let Richard reply to this question, as well as to disagreements
> by Miles and others.  Unlike some others, I don't care much about
> whether files should or should not be committed one by one, as long as
> whatever practice we adopt is adhered to by everybody.

I will not make single file commits, in general.

-Miles

-- 
In New York, most people don't have cars, so if you want to kill a person, you
have to take the subway to their house.  And sometimes on the way, the train
is delayed and you get impatient, so you have to kill someone on the subway.
  [George Carlin]

Re: Log messages in CVS

[Stefan Monnier]

>>> Ok, I've read the thread and that Richard agreed with the rules.  But on
>>> the other hand others like Miles disagreed.  Is the only purpose of
>>> those single file commits really to keep the CVS logs clean?
>> 
>> I'll let Richard reply to this question, as well as to disagreements
>> by Miles and others.  Unlike some others, I don't care much about
>> whether files should or should not be committed one by one, as long as
>> whatever practice we adopt is adhered to by everybody.

> I will not make single file commits, in general.

I usually do single-file commits but I also will not in general refrain
from multi-file commits.  I see no benefit in ruling them out just for
the sake of "simplicity" or some other such principle.  Let's worry
about consistency where it matters, but only there.


        Stefan