unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
@ 2015-04-14  6:01 Glenn Morris
  2015-04-17  6:46 ` Paul Eggert
  0 siblings, 1 reply; 40+ messages in thread
From: Glenn Morris @ 2015-04-14  6:01 UTC (permalink / raw)
  To: 20325

Package: emacs
Version: 25.0.50

The syntax that Emacs adopted for bug references in ChangeLogs was
"(Bug#123)", or "(bug#123)".

So gitlog-to-changelog should convert "Fixes: *(de)?bugs:123" in commit
logs to "(Bug#123)" in the output ChangeLog.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-14  6:01 bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog Glenn Morris
@ 2015-04-17  6:46 ` Paul Eggert
  2015-04-17 11:25   ` Dmitry Gutov
  2015-04-20 21:54   ` Glenn Morris
  0 siblings, 2 replies; 40+ messages in thread
From: Paul Eggert @ 2015-04-17  6:46 UTC (permalink / raw)
  To: 20325

Wouldn't it be better to use the same format in commit messages and in 
ChangeLog.2?  That will be less confusing for everybody.  I don't care what 
format is used; "Fixes: bug#NNN" is fine, as is "(Bug#NNN)".  I'm more used to 
the latter format, and I don't know where the former format came from.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17  6:46 ` Paul Eggert
@ 2015-04-17 11:25   ` Dmitry Gutov
  2015-04-17 15:19     ` Paul Eggert
  2015-04-17 21:45     ` Michael Albinus
  2015-04-20 21:54   ` Glenn Morris
  1 sibling, 2 replies; 40+ messages in thread
From: Dmitry Gutov @ 2015-04-17 11:25 UTC (permalink / raw)
  To: Paul Eggert, 20325

On 04/17/2015 09:46 AM, Paul Eggert wrote:
> Wouldn't it be better to use the same format in commit messages and in
> ChangeLog.2?  That will be less confusing for everybody.  I don't care
> what format is used; "Fixes: bug#NNN" is fine, as is "(Bug#NNN)".  I'm
> more used to the latter format, and I don't know where the former format
> came from.

I think debbugs:NNN is a bit better than Bug#NNN. It's more specific.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 11:25   ` Dmitry Gutov
@ 2015-04-17 15:19     ` Paul Eggert
  2015-04-17 15:32       ` Dmitry Gutov
  2015-04-17 21:45     ` Michael Albinus
  1 sibling, 1 reply; 40+ messages in thread
From: Paul Eggert @ 2015-04-17 15:19 UTC (permalink / raw)
  To: Dmitry Gutov, 20325

Dmitry Gutov wrote:
> I think debbugs:NNN is a bit better than Bug#NNN. It's more specific.

Currently, if you put ' (Bug#1234)' in a commit message, Emacs VC (via 
.dir-locals.el's setting of log-edit-rewrite-fixes) alters this to 'Fixes: 
debbugs:1234'.  Looking at recent commits, it appears that this is commonly done 
in practice, though I don't know what keystrokes trigger this.

How do you commit changes involving bug numbers?  What Emacs commands do you 
use?  (I use vc-dwim, but evidently I'm in the minority.)  Perhaps the "obvious" 
information about what keystrokes to use should be put into CONTRIBUTE, and 
CONTRIBUTE should recommend "Fixes: debbugs:1234' instead of its current 
recommendation 'Fixes: bug#1234'.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 15:19     ` Paul Eggert
@ 2015-04-17 15:32       ` Dmitry Gutov
  2015-04-17 15:56         ` Paul Eggert
  2015-04-17 16:20         ` Jan D.
  0 siblings, 2 replies; 40+ messages in thread
From: Dmitry Gutov @ 2015-04-17 15:32 UTC (permalink / raw)
  To: Paul Eggert, 20325

On 04/17/2015 06:19 PM, Paul Eggert wrote:

> Currently, if you put ' (Bug#1234)' in a commit message, Emacs VC (via
> .dir-locals.el's setting of log-edit-rewrite-fixes) alters this to
> 'Fixes: debbugs:1234'.

You put (Bug#1234) into the top-level ChangeLog file. Later, 
`log-edit-insert-changelog' performs the conversion according to the 
value of `log-edit-rewrite-fixes'.

> Looking at recent commits, it appears that this
> is commonly done in practice

Indeed, that's why I mentioned it.

> How do you commit changes involving bug numbers?  What Emacs commands do
> you use?  (I use vc-dwim, but evidently I'm in the minority.)

I either use the ChangeLog, or type out "Fixes: debbugs:1234" manually.

 > Perhaps
> the "obvious" information about what keystrokes to use should be put
> into CONTRIBUTE,

Already covered, in the "Generating ChangeLog entries" section.

> and CONTRIBUTE should recommend "Fixes: debbugs:1234'
> instead of its current recommendation 'Fixes: bug#1234'.

I think so, yes.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 15:32       ` Dmitry Gutov
@ 2015-04-17 15:56         ` Paul Eggert
  2015-04-17 16:02           ` Dmitry Gutov
  2015-04-17 16:25           ` Jan D.
  2015-04-17 16:20         ` Jan D.
  1 sibling, 2 replies; 40+ messages in thread
From: Paul Eggert @ 2015-04-17 15:56 UTC (permalink / raw)
  To: Dmitry Gutov, 20325

Dmitry Gutov wrote:
>
>  > Perhaps
>> the "obvious" information about what keystrokes to use should be put
>> into CONTRIBUTE,
>
> Already covered, in the "Generating ChangeLog entries" section.

I tried using that procedure, and it didn't work for me.  So it appears that the 
procedure could be documented better for new contributors who are not used to 
Emacs VC (as I am not; I use vc-dwim).  Let me explain the problem I ran into, 
and then perhaps we can come up with an improvement to CONTRIBUTE.

I made a trivial change to Makefile.in, and then typed C-x 4 a.  This created a 
file ChangeLog, and I edited it to look like this:

2015-04-17  Paul Eggert  <eggert@cs.ucla.edu>

	Try to make a trivial change (Bug#1234)
	* Makefile.in: A trivial change.

I then typed C-x C-s to save it, and type C-c C-a as CONTRIBUTE suggested. 
Emacs responded "C-c C-a is undefined".  What did I do wrong?





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 15:56         ` Paul Eggert
@ 2015-04-17 16:02           ` Dmitry Gutov
  2015-04-17 16:25           ` Jan D.
  1 sibling, 0 replies; 40+ messages in thread
From: Dmitry Gutov @ 2015-04-17 16:02 UTC (permalink / raw)
  To: Paul Eggert, 20325

On 04/17/2015 06:56 PM, Paul Eggert wrote:

> I made a trivial change to Makefile.in, and then typed C-x 4 a.  This
> created a file ChangeLog, and I edited it to look like this:
>
> 2015-04-17  Paul Eggert  <eggert@cs.ucla.edu>
>
>      Try to make a trivial change (Bug#1234)
>      * Makefile.in: A trivial change.
>
> I then typed C-x C-s to save it, and type C-c C-a as CONTRIBUTE
> suggested. Emacs responded "C-c C-a is undefined".  What did I do wrong?

`C-c C-a' is only bound in the *vc-log* buffer. Before it's created, 
Emacs has no place to insert the ChangeLog entry anyway.

So, press `C-x v v' first (either in Makefile.in, or in vc-dir buffer 
after selecting the files to commit).





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 15:32       ` Dmitry Gutov
  2015-04-17 15:56         ` Paul Eggert
@ 2015-04-17 16:20         ` Jan D.
  2015-04-17 16:30           ` Dmitry Gutov
  1 sibling, 1 reply; 40+ messages in thread
From: Jan D. @ 2015-04-17 16:20 UTC (permalink / raw)
  To: Dmitry Gutov; +Cc: 20325, Paul Eggert


> 17 apr 2015 kl. 17:32 skrev Dmitry Gutov <dgutov@yandex.ru>:
> 
> On 04/17/2015 06:19 PM, Paul Eggert wrote:
> 
>> and CONTRIBUTE should recommend "Fixes: debbugs:1234'
>> instead of its current recommendation 'Fixes: bug#1234'.
> 
> I think so, yes.
> 

What about a few years from now when we switch to something other than debbugs?  Why hardcode a specific bug reporting tool as the recommendation?  That does not make sense to me.

	Jan D.






^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 15:56         ` Paul Eggert
  2015-04-17 16:02           ` Dmitry Gutov
@ 2015-04-17 16:25           ` Jan D.
  1 sibling, 0 replies; 40+ messages in thread
From: Jan D. @ 2015-04-17 16:25 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325, Dmitry Gutov


> 17 apr 2015 kl. 17:56 skrev Paul Eggert <eggert@cs.ucla.edu>:
> 
> Dmitry Gutov wrote:
>> 
>> > Perhaps
>>> the "obvious" information about what keystrokes to use should be put
>>> into CONTRIBUTE,
>> 
>> Already covered, in the "Generating ChangeLog entries" section.
> 
> I tried using that procedure, and it didn't work for me.  So it appears that the procedure could be documented better for new contributors who are not used to Emacs VC (as I am not; I use vc-dwim).  Let me explain the problem I ran into, and then perhaps we can come up with an improvement to CONTRIBUTE.
> 
> I made a trivial change to Makefile.in, and then typed C-x 4 a.  This created a file ChangeLog, and I edited it to look like this:
> 
> 2015-04-17  Paul Eggert  <eggert@cs.ucla.edu>
> 
> 	Try to make a trivial change (Bug#1234)
> 	* Makefile.in: A trivial change.
> 
> I then typed C-x C-s to save it, and type C-c C-a as CONTRIBUTE suggested. Emacs responded "C-c C-a is undefined".  What did I do wrong?

If CONTRIBUTE suggests that it is omitting a few steps.  After saving it you should do a VC-dir (C-x v d), and mark files you want to commit in the *vc-dir* buffer.  Then press v, and in that log buffer do C-c C-a.

	Jan D.







^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 16:20         ` Jan D.
@ 2015-04-17 16:30           ` Dmitry Gutov
  2015-04-17 17:06             ` Jan D.
  0 siblings, 1 reply; 40+ messages in thread
From: Dmitry Gutov @ 2015-04-17 16:30 UTC (permalink / raw)
  To: Jan D.; +Cc: 20325, Paul Eggert

On 04/17/2015 07:20 PM, Jan D. wrote:

> What about a few years from now when we switch to something other than debbugs?

That's exactly the situation where having used debbugs:1234 until that 
point will show its benefits.

 > Why hardcode a specific bug reporting tool as the recommendation?

Will all bugs be migrated to the new bug reporting tool? Will they keep 
their identifiers?

Unless you're 100% sure about that, explicitly documenting which bug 
tracker the bugs are reported in is a useful information.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 16:30           ` Dmitry Gutov
@ 2015-04-17 17:06             ` Jan D.
  2015-04-17 17:15               ` Dmitry Gutov
  0 siblings, 1 reply; 40+ messages in thread
From: Jan D. @ 2015-04-17 17:06 UTC (permalink / raw)
  To: Dmitry Gutov; +Cc: 20325, Paul Eggert

Hi.

> 17 apr 2015 kl. 18:30 skrev Dmitry Gutov <dgutov@yandex.ru>:
> 
> On 04/17/2015 07:20 PM, Jan D. wrote:
> 
>> What about a few years from now when we switch to something other than debbugs?
> 
> That's exactly the situation where having used debbugs:1234 until that point will show its benefits.
> 
> > Why hardcode a specific bug reporting tool as the recommendation?
> 
> Will all bugs be migrated to the new bug reporting tool? Will they keep their identifiers?

Probably not.  Why does that matter?

> 
> Unless you're 100% sure about that, explicitly documenting which bug tracker the bugs are reported in is a useful information.

We probably can remember the date we change bug tracker, so its not rocket sience to keep track of what bug tracker bugs are in.

We are talking about what people should write in vc-log.  If you tell them to write Bug#XXXX that recommendation is generic.  What actually goes into the git log is another matter.

	Jan D.






^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 17:06             ` Jan D.
@ 2015-04-17 17:15               ` Dmitry Gutov
  2015-04-18  7:46                 ` Jan D.
  0 siblings, 1 reply; 40+ messages in thread
From: Dmitry Gutov @ 2015-04-17 17:15 UTC (permalink / raw)
  To: Jan D.; +Cc: 20325, Paul Eggert

On 04/17/2015 08:06 PM, Jan D. wrote:

 > Probably not.  Why does that matter?

The old references will point to non-existent bugs.

> We probably can remember the date we change bug tracker, so its not rocket sience to keep track of what bug tracker bugs are in.

Will you implement this kind of date-tracking in bug-reference-mode?

> We are talking about what people should write in vc-log.  If you tell them to write Bug#XXXX that recommendation is generic.  What actually goes into the git log is another matter.

As soon as we switch to a new bug tracker, we'll update CONTRIBUTE.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 11:25   ` Dmitry Gutov
  2015-04-17 15:19     ` Paul Eggert
@ 2015-04-17 21:45     ` Michael Albinus
  2015-04-17 21:48       ` Dmitry Gutov
  1 sibling, 1 reply; 40+ messages in thread
From: Michael Albinus @ 2015-04-17 21:45 UTC (permalink / raw)
  To: Dmitry Gutov; +Cc: 20325, Paul Eggert

Dmitry Gutov <dgutov@yandex.ru> writes:

> On 04/17/2015 09:46 AM, Paul Eggert wrote:
>> Wouldn't it be better to use the same format in commit messages and in
>> ChangeLog.2?  That will be less confusing for everybody.  I don't care
>> what format is used; "Fixes: bug#NNN" is fine, as is "(Bug#NNN)".  I'm
>> more used to the latter format, and I don't know where the former format
>> came from.
>
> I think debbugs:NNN is a bit better than Bug#NNN. It's more specific.

I would prefer the (Bug#NNN) format. Otherwise, my commit messages /
ChangeLog entries must differ between the Emacs and Tramp repositories.

Best regards, Michael.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 21:45     ` Michael Albinus
@ 2015-04-17 21:48       ` Dmitry Gutov
  2015-04-17 22:02         ` Michael Albinus
  0 siblings, 1 reply; 40+ messages in thread
From: Dmitry Gutov @ 2015-04-17 21:48 UTC (permalink / raw)
  To: Michael Albinus; +Cc: 20325, Paul Eggert

On 04/18/2015 12:45 AM, Michael Albinus wrote:

> Otherwise, my commit messages /
> ChangeLog entries must differ between the Emacs and Tramp repositories.

How come? Don't you set the standards for the Tramp repository yourself?





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 21:48       ` Dmitry Gutov
@ 2015-04-17 22:02         ` Michael Albinus
  2015-04-17 22:44           ` Dmitry Gutov
  0 siblings, 1 reply; 40+ messages in thread
From: Michael Albinus @ 2015-04-17 22:02 UTC (permalink / raw)
  To: Dmitry Gutov; +Cc: 20325, Paul Eggert

Dmitry Gutov <dgutov@yandex.ru> writes:

> How come? Don't you set the standards for the Tramp repository yourself?

Sure. But the ChangeLog and some Tramp files contain already references
to (Bug#nnn). I do not intend to mix different formats, and I do not
intend to change such references to a new format.

That's possible of course, but I don't see why I shall. I've spoken
about my preferred format of bug references (the existing one); and I
haven't seen a convincing argument why it must be changed.

Bug references are not used only in Emacs ChangeLog files. There might
be other Lisp packages using them, and there might be even other software
but Emacs which uses them. See the list of packages hosted on
debbugs.gnu.org <http://debbugs.gnu.org/Packages.html>. Some of the
maintainers might use Emacs as editor.

Best regards, Michael.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 22:02         ` Michael Albinus
@ 2015-04-17 22:44           ` Dmitry Gutov
  2015-04-18  1:02             ` Paul Eggert
  2015-04-18  7:14             ` Eli Zaretskii
  0 siblings, 2 replies; 40+ messages in thread
From: Dmitry Gutov @ 2015-04-17 22:44 UTC (permalink / raw)
  To: Michael Albinus; +Cc: 20325, Paul Eggert

On 04/18/2015 01:02 AM, Michael Albinus wrote:

> That's possible of course, but I don't see why I shall. I've spoken
> about my preferred format of bug references (the existing one); and I
> haven't seen a convincing argument why it must be changed.

I don't know about "changed". We've used the debbugs:1234 format in 
commit messages for a while now.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 22:44           ` Dmitry Gutov
@ 2015-04-18  1:02             ` Paul Eggert
  2015-04-18  6:28               ` Michael Albinus
  2015-04-18  7:25               ` Eli Zaretskii
  2015-04-18  7:14             ` Eli Zaretskii
  1 sibling, 2 replies; 40+ messages in thread
From: Paul Eggert @ 2015-04-18  1:02 UTC (permalink / raw)
  To: Dmitry Gutov, Michael Albinus; +Cc: 20325

Dmitry Gutov wrote:
> We've used the debbugs:1234 format in commit messages for a while now.

We've used several formats.  "Fixes: debbugs:NNNN" is designed for Bzr I think, 
"(Bug#NNNN)" is the more-traditional form, and CONTRIBUTE currently recommends a 
third format "Fixes: bug#NNNN".

In practice, about as many commits these days use the Bzr-ish format as use the 
more-traditional form (see (*) below).  I don't know why we're still using a 
Bzr-ish format (is it just inertia?), and I don't know why CONTRIBUTE disagrees 
with both the commonly-used formats.

Since we're not using Bzr any more, how about if we change CONTRIBUTE to suggest 
the more-traditional "(Bug#NNNN)" form?  Any new technology that replaces 
debbugs would have to handle the same bug numbers as before, I expect.

(*) I surveyed the 2213 commits since November 1 (this is a bit before the Git 
transition), i.e., all commits from d87931d249a4c7f15d4c1178c6ce396ab18ce1f6 
through 35be45bbc9ac525bb2bc3e20b0fa27c1ceee0c15, and found the following number 
of instances of each format:

  count	example of the style
  262	Fixes: debbugs:20357
  172	(Bug#20011)
   49    (bug#20276)
   24	Fixes: bug#20338
    8	Fix bug#19891
    6	Fixes: Bug#20338
    5	Fixes: 18757
    5	Fixes Bug#15880.
    2	Fixes: debbugs:19387 debbugs:19405
    1	Fixes: debbugs:19134.
    1	Fixes: big#19528
    1	Fixes (bug#17284).
    1	Fix Bug#20249






^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-18  1:02             ` Paul Eggert
@ 2015-04-18  6:28               ` Michael Albinus
  2015-04-18  7:25               ` Eli Zaretskii
  1 sibling, 0 replies; 40+ messages in thread
From: Michael Albinus @ 2015-04-18  6:28 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325, Dmitry Gutov

Paul Eggert <eggert@cs.ucla.edu> writes:

> Dmitry Gutov wrote:
>> We've used the debbugs:1234 format in commit messages for a while now.
>
> We've used several formats.  "Fixes: debbugs:NNNN" is designed for Bzr
> I think, "(Bug#NNNN)" is the more-traditional form, and CONTRIBUTE
> currently recommends a third format "Fixes: bug#NNNN".

My major concern are ChangeLogs and other files, containing bug
references. Here we shall keep the established format (Bug#NNN), even in
generated files. That's what this bug report is about.

For consistency reasons, it might be preferred to use it also in commit
messages. 

Best regards, Michael.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 22:44           ` Dmitry Gutov
  2015-04-18  1:02             ` Paul Eggert
@ 2015-04-18  7:14             ` Eli Zaretskii
  2015-04-19  2:59               ` Paul Eggert
  1 sibling, 1 reply; 40+ messages in thread
From: Eli Zaretskii @ 2015-04-18  7:14 UTC (permalink / raw)
  To: Dmitry Gutov; +Cc: 20325, eggert, michael.albinus

> Date: Sat, 18 Apr 2015 01:44:59 +0300
> From: Dmitry Gutov <dgutov@yandex.ru>
> Cc: 20325@debbugs.gnu.org, Paul Eggert <eggert@cs.ucla.edu>
> 
> We've used the debbugs:1234 format in commit messages for a while
> now.

For some value of "we".  CONTRIBUTE still requires to use
"Fixes: bug#NNNN", and that's what I've switched to once those
requirements were written.  Which was already a change from the
previous procedure, which preferred simply "Bug#NNNN".

I wish you guys settle on something once and for all, never mind what,
and stop bikeshedding about it and changing the procedures every 2
weeks or so.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-18  1:02             ` Paul Eggert
  2015-04-18  6:28               ` Michael Albinus
@ 2015-04-18  7:25               ` Eli Zaretskii
  2015-04-18  9:25                 ` Ivan Shmakov
  1 sibling, 1 reply; 40+ messages in thread
From: Eli Zaretskii @ 2015-04-18  7:25 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325, michael.albinus, dgutov

> Date: Fri, 17 Apr 2015 18:02:51 -0700
> From: Paul Eggert <eggert@cs.ucla.edu>
> Cc: 20325@debbugs.gnu.org
> 
> "Fixes: debbugs:NNNN" is designed for Bzr I think,

Bzr didn't have this in the log messages, it had this as meta-data.
Conversion to Git moved this into the commit log, since AFAIK Git
doesn't have such an attribute in its meta-data.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17 17:15               ` Dmitry Gutov
@ 2015-04-18  7:46                 ` Jan D.
  2015-04-18  8:08                   ` Eli Zaretskii
  0 siblings, 1 reply; 40+ messages in thread
From: Jan D. @ 2015-04-18  7:46 UTC (permalink / raw)
  To: Dmitry Gutov; +Cc: 20325, Paul Eggert

Hi.

> 17 apr 2015 kl. 19:15 skrev Dmitry Gutov <dgutov@yandex.ru>:
> 
> On 04/17/2015 08:06 PM, Jan D. wrote:
> 
> > Probably not.  Why does that matter?
> 
> The old references will point to non-existent bugs.
> 
>> We probably can remember the date we change bug tracker, so its not rocket sience to keep track of what bug tracker bugs are in.
> 
> Will you implement this kind of date-tracking in bug-reference-mode?

A file with a date in it?  Sure.

> 
>> We are talking about what people should write in vc-log.  If you tell them to write Bug#XXXX that recommendation is generic.  What actually goes into the git log is another matter.
> 
> As soon as we switch to a new bug tracker, we'll update CONTRIBUTE.

The hard part is to force all contributors to change their habits.

	Jan D.






^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-18  7:46                 ` Jan D.
@ 2015-04-18  8:08                   ` Eli Zaretskii
  0 siblings, 0 replies; 40+ messages in thread
From: Eli Zaretskii @ 2015-04-18  8:08 UTC (permalink / raw)
  To: Jan D.; +Cc: 20325, eggert, dgutov

> From: "Jan D." <jan.h.d@swipnet.se>
> Date: Sat, 18 Apr 2015 09:46:37 +0200
> Cc: 20325@debbugs.gnu.org, Paul Eggert <eggert@cs.ucla.edu>
> 
> The hard part is to force all contributors to change their habits.

Which in this particular case I had to do already 3 times since last
November, with the 4th time looming.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-18  7:25               ` Eli Zaretskii
@ 2015-04-18  9:25                 ` Ivan Shmakov
  2015-04-18  9:46                   ` Eli Zaretskii
  0 siblings, 1 reply; 40+ messages in thread
From: Ivan Shmakov @ 2015-04-18  9:25 UTC (permalink / raw)
  To: 20325, dgutov

>>>>> Eli Zaretskii <eliz@gnu.org> writes:
>>>>> From: Paul Eggert <eggert@cs.ucla.edu>

 >> "Fixes: debbugs:NNNN" is designed for Bzr I think,

 > Bzr didn't have this in the log messages, it had this as meta-data.
 > Conversion to Git moved this into the commit log, since AFAIK Git
 > doesn't have such an attribute in its meta-data.

	I guess the above means there’s no valid reason to stick to this
	format any longer?

 >> count  example of the style
 >> 262    Fixes: debbugs:20357
 >> 172    (Bug#20011)
 >>  49    (bug#20276)

	I believe these two are essentially the same format – subject to
	the usual capitalization rules.  Consider, e. g.:

	* lisp/foo.el (foo-bar): Do not call qux.  (Bug#54321)
	* lisp/baz.el (baz-mode): Let-bind foo-1 to nil (bug#53210) and
	call bar.

	FWIW, I’d prefer to just stick with the (bug#) markers for
	commit messages, and let them propagate as is to the ChangeLog
	when one’s generated.

-- 
FSF associate member #7257  http://boycottsystemd.org/  … 3013 B6A0 230E 334A





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-18  9:25                 ` Ivan Shmakov
@ 2015-04-18  9:46                   ` Eli Zaretskii
  0 siblings, 0 replies; 40+ messages in thread
From: Eli Zaretskii @ 2015-04-18  9:46 UTC (permalink / raw)
  To: Ivan Shmakov; +Cc: 20325, dgutov

> From: Ivan Shmakov <ivan@siamics.net>
> Date: Sat, 18 Apr 2015 09:25:30 +0000
> 
> >>>>> Eli Zaretskii <eliz@gnu.org> writes:
> >>>>> From: Paul Eggert <eggert@cs.ucla.edu>
> 
>  >> "Fixes: debbugs:NNNN" is designed for Bzr I think,
> 
>  > Bzr didn't have this in the log messages, it had this as meta-data.
>  > Conversion to Git moved this into the commit log, since AFAIK Git
>  > doesn't have such an attribute in its meta-data.
> 
> 	I guess the above means there’s no valid reason to stick to this
> 	format any longer?

Probably not.

> 	FWIW, I’d prefer to just stick with the (bug#) markers for
> 	commit messages, and let them propagate as is to the ChangeLog
> 	when one’s generated.

Agreed, as that scales better to all the possible ways of making a
commit (some use VC, some use magit, some use command line plus
emacsclient, etc.).





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-18  7:14             ` Eli Zaretskii
@ 2015-04-19  2:59               ` Paul Eggert
  2015-04-19 14:23                 ` Eli Zaretskii
  0 siblings, 1 reply; 40+ messages in thread
From: Paul Eggert @ 2015-04-19  2:59 UTC (permalink / raw)
  To: Eli Zaretskii, Dmitry Gutov; +Cc: 20325, michael.albinus

[-- Attachment #1: Type: text/plain, Size: 1059 bytes --]

Eli Zaretskii wrote:
> For some value of "we".  CONTRIBUTE still requires to use
> "Fixes: bug#NNNN", and that's what I've switched to once those
> requirements were written.  Which was already a change from the
> previous procedure, which preferred simply "Bug#NNNN".
>
> I wish you guys settle on something once and for all, never mind what,

"you guys"?  I thought you were one of us!  :-)

I did not choose the "Fixes: bug#NNNN" syntax.  I vaguely recall it came from an 
offhand remark in one of Stefan's emails a few months ago and that it migrated 
into CONTRIBUTE when Stephen Leake attempted to summarize that discussion, but 
possibly it was a typo in Stefan's message and anyway I expect he doesn't much 
care what the format is.

 From the earlier discussion in this thread it appears that the "Fixes: 
debbugs:NNNN" syntax is an artifact of the Bzr-to-Git conversion that now causes 
more confusion than it's worth.  To try to lessen this I installed the attached 
patches (the third one is mechanical; it's the result of 'make change-history').

[-- Attachment #2: 0001-Prefer-Bug-1234-in-commit-messages.patch --]
[-- Type: text/x-patch, Size: 2142 bytes --]

From f2da220b5fde8575812784f6a9e5f7df9560e838 Mon Sep 17 00:00:00 2001
From: Paul Eggert <eggert@cs.ucla.edu>
Date: Sat, 18 Apr 2015 19:08:53 -0700
Subject: [PATCH 1/4] Prefer "Bug#1234" in commit messages

* .dir-locals.el (log-edit-mode): Don't rewrite Bug#,
as this isn't useful for Git.
* CONTRIBUTE: Suggest "Bug#1234" instead of "Fixes: debbugs:1234".
---
 .dir-locals.el |  4 +---
 CONTRIBUTE     | 11 ++++-------
 2 files changed, 5 insertions(+), 10 deletions(-)

diff --git a/.dir-locals.el b/.dir-locals.el
index f899b51..9853d7b 100644
--- a/.dir-locals.el
+++ b/.dir-locals.el
@@ -3,9 +3,7 @@
          (fill-column . 70)))
  (c-mode . ((c-file-style . "GNU")))
  (objc-mode . ((c-file-style . "GNU")))
- (log-edit-mode . ((log-edit-rewrite-fixes
-                    "[ \n](bug#\\([0-9]+\\))" . "debbugs:\\1")
-                   (log-edit-font-lock-gnu-style . t)
+ (log-edit-mode . ((log-edit-font-lock-gnu-style . t)
                    (log-edit-setup-add-author . t)))
  (change-log-mode . ((add-log-time-zone-rule . t)
 		     (fill-column . 74)
diff --git a/CONTRIBUTE b/CONTRIBUTE
index 476b361..4205bc3 100644
--- a/CONTRIBUTE
+++ b/CONTRIBUTE
@@ -42,12 +42,11 @@ in its message.  Here is an example commit message (indented):
 	Deactivate shifted region
 
 	Do not silently extend a region that is not highlighted;
-	this can happen after a shift.
+	this can happen after a shift (Bug#19003).
 	* doc/emacs/mark.texi (Shift Selection): Document the change.
 	* lisp/window.el (handle-select-window):
 	* src/frame.c (Fhandle_switch_frame, Fselected_frame):
 	Deactivate the mark.
-	Fixes: bug#19003
 
 The general format is as follows.
 
@@ -113,11 +112,9 @@ The general format is as follows.
 
 	Copyright-paperwork-exempt: yes
 
-- If the commit fixes a bug, append a separate line
-
-	Fixes: bug#NNNN
-
-  where NNNN is the bug number.
+- The commit message should contain "Bug#NNNNN" if it is related to
+  bug number NNNNN in the debbugs database.  This string is often
+  parenthesized, as in "(Bug#19003)".
 
 - In ChangeLog entries, there is no standard or recommended way to
   identify revisions.
-- 
2.1.0


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #3: 0002-Convert-ChangeLog-history-files-to-Bug-NNNN.patch --]
[-- Type: text/x-patch; name="0002-Convert-ChangeLog-history-files-to-Bug-NNNN.patch", Size: 11287 bytes --]

From 870287327b1fba7105599eeabac5a2aa5ebadf19 Mon Sep 17 00:00:00 2001
From: Paul Eggert <eggert@cs.ucla.edu>
Date: Sat, 18 Apr 2015 19:45:38 -0700
Subject: [PATCH 2/4] ; Convert ChangeLog history files to "Bug#NNNN"

---
 ChangeLog.1       |  8 ++++----
 ChangeLog.2       | 49 ++++++++++++++++++++++++-------------------------
 lisp/ChangeLog.17 |  4 ++--
 src/ChangeLog.13  | 12 ++++++------
 test/ChangeLog.1  |  2 +-
 5 files changed, 37 insertions(+), 38 deletions(-)

diff --git a/ChangeLog.1 b/ChangeLog.1
index 6e2b4fc..d1b8d0f 100644
--- a/ChangeLog.1
+++ b/ChangeLog.1
@@ -55,7 +55,7 @@
 	(HAVE_PTHREAD, LIB_PTHREAD): Port better to FreeBSD,
 	by also checking for pthread_create, pthread_self, pthread_sigmask.
 	Tighten the test for pthread_atfork while we're at it.
-	Fixes: bug#20136
+	(Bug#20136)
 
 	Merge from gnulib
 	This incorporates:
@@ -76,7 +76,7 @@
 	Don't require GNU putenv
 	* configure.ac: Use system putenv even if it lacks GNU features, as
 	we don't need them.  This works around a bug in FreeBSD 10.1 getenv.
-	Fixes: bug#19874
+	(Bug#19874)
 
 2015-02-25  Paul Eggert  <eggert@cs.ucla.edu>
 
@@ -170,7 +170,7 @@
 	If gnustep-config works, use 'gnustep-config --objc-flags' and
 	'gnustep-config --gui-libs' to compute GNUstep configuration
 	variables, instead of attempting to infer them individually.
-	Fixes: bug#19507
+	(Bug#19507)
 
 2015-01-21  Ulrich Müller  <ulm@gentoo.org>
 
@@ -198,7 +198,7 @@
 	* INSTALL: Document this.
 	* configure.ac: Add AM_SILENT_RULES([yes]).
 	(AM_DEFAULT_VERBOSITY): Remove now-unnecessary initialization.
-	Fixes: bug#19501
+	(Bug#19501)
 
 2015-01-06  Paul Eggert  <eggert@cs.ucla.edu>
 
diff --git a/ChangeLog.2 b/ChangeLog.2
index 2ecf2b5..044f888 100644
--- a/ChangeLog.2
+++ b/ChangeLog.2
@@ -16,7 +16,7 @@
 
 	Pre-4.6 GCC succeeds with unknown option
 	* configure.ac (emacs_cv_prog_cc_nopie): Port to pre-4.6 GCC.
-	Fixes: bug#20338
+	(Bug#20338)
 
 2015-04-15  Paul Eggert  <eggert@cs.ucla.edu>
 
@@ -46,7 +46,7 @@
 	both jpeglib.h and jerror.h features.  Remove special case for
 	mingw32, which should no longer be needed (and if it were needed,
 	should now be addressable by hotwiring emacs_cv_jpeglib).
-	Fixes: bug#20332
+	(Bug#20332)
 
 2015-04-15  Stefan Monnier  <monnier@iro.umontreal.ca>
 
@@ -100,7 +100,7 @@
 2015-04-15  Stefan Monnier  <monnier@iro.umontreal.ca>
 
 	* src/lread.c (intern_1): Make sure we'd find the symbol we add
-	Fixes: debbugs:20334
+	(Bug#20334)
 	* src/xfaces.c (resolve_face_name): Don't use `intern' with Lisp_Strings.
 
 2015-04-15  Glenn Morris  <rgm@gnu.org>
@@ -169,7 +169,7 @@
 	Describe problems with cursor caused by Windows Magnifier
 	* etc/PROBLEMS: Describe the problem with cursor shape on
 	MS-Windows due to Windows Magnifier.
-	Fixes: Bug#20271
+	(Bug#20271)
 
 	Make [:print:] support non-ASCII characters correctly
 	* src/regex.c (ISPRINT): Call 'printablep' for multibyte characters.
@@ -254,7 +254,7 @@
 2015-04-13  Dmitry Gutov  <dgutov@yandex.ru>
 
 	Change diff-switches default to `-u'
-	Fixes: debbugs:20290
+	(Bug#20290)
 	* doc/emacs/files.texi (Comparing Files): Document the new default
 	value of `diff-switches'.
 	* doc/emacs/trouble.texi (Sending Patches): Document the preference
@@ -276,7 +276,7 @@
 2015-04-13  Stefan Monnier  <monnier@iro.umontreal.ca>
 
 	* lisp/loadup.el ("cus-start"): Load it after loaddefs.el
-	Fixes: debbugs:20321
+	(Bug#20321)
 	* lisp/cus-start.el (read-buffer-function): Don't advertize
 	iswitchb-read-buffer any more.
 	(iswitchb): Don't tweak this obsolete group any more.
@@ -432,7 +432,7 @@
 	Use completion-separator-chars rather than local key binding.
 
 	* src/*.c: Set deactivate_mark buffer-locally
-	Fixes: debbugs:20260
+	(Bug#20260)
 	* src/insdel.c (prepare_to_modify_buffer_1):
 	* src/fileio.c (Finsert_file_contents): Set deactivate_mark
 	buffer-locally.
@@ -440,7 +440,7 @@
 2015-04-12  Fabián Ezequiel Gallina  <fgallina@gnu.org>
 
 	python.el: Keep symmetry on sexp navigation with parens
-	Fixes: debbugs:19954
+	(Bug#19954)
 	* lisp/progmodes/python.el
 	(python-nav--forward-sexp): Add argument skip-parens-p.
 	(python-nav-forward-sexp, python-nav-backward-sexp)
@@ -466,7 +466,7 @@
 	* configure.ac (MAIL_USE_MMDF): Remove.
 	* etc/NEWS: Document this.
 	* lib-src/movemail.c: Assume MAIL_USE_MMDF is not defined.
-	Fixes: bug#20308
+	(Bug#20308)
 
 	* doc/man/ChangeLog.01: Rename from doc/man/ChangeLog.1.
 	That way, 'make install' won't think it's a man page.
@@ -675,7 +675,7 @@
 	* src/gnutls.c (syms_of_gnutls) <libgnutls-version>: New DEFSYM.
 	* lisp/term/w32-win.el (dynamic-library-alist): Determine which
 	GnuTLS DLL to load according to value of libgnutls-version.
-	Fixes: bug#20294
+	(Bug#20294)
 
 2015-04-11  Paul Eggert  <eggert@cs.ucla.edu>
 
@@ -863,7 +863,7 @@
 2015-04-09  Simen Heggestøyl  <simenheg@gmail.com>
 
 	css-mode.el: Add "not" pseudo-class
-	Fixes: debbugs:20267
+	(Bug#20267)
 	* lisp/textmodes/css-mode.el (css-pseudo-class-ids): Add "not" to
 	list of CSS pseudo-classes.
 
@@ -904,12 +904,12 @@
 2015-04-09  Boruch Baum  <boruch_baum@gmx.com>  (tiny change)
 
 	* lisp/bookmark.el (bookmark-bmenu-goto-bookmark): Don't inf-loop.
-	Fixes: debbugs:20212
+	(Bug#20212)
 
 2015-04-09  Stefan Monnier  <monnier@iro.umontreal.ca>
 
 	Stop messing with the EMACS env var
-	Fixes: debbugs:20202
+	(Bug#20202)
 	* lisp/net/tramp-sh.el (tramp-remote-process-environment):
 	* lisp/comint.el (comint-exec-1):
 	* lisp/term.el (term-exec-1): Don't set EMACS envvar.
@@ -917,14 +917,14 @@
 	INSIDE_EMACS's format in line with other users.
 
 	css-mode.el (css-smie-rules): Fix indentation after complex selectors
-	Fixes: debbugs:20282
+	(Bug#20282)
 	* lisp/textmodes/css-mode.el (css-smie-rules): Don't get confused by
 	inner structure of selectors.
 
 2015-04-08  Fabián Ezequiel Gallina  <fgallina@gnu.org>
 
 	python.el: Indent docstring lines to base-indent
-	Fixes: debbugs:19595
+	(Bug#19595)
 	Thanks to immerrr <immerrr@gmail.com> for reporting and providing
 	an initial patch.
 	* lisp/progmodes/python.el
@@ -935,7 +935,7 @@
 	(python-indent-inside-string-2): Fix tests.
 
 	python.el: Increase native completion robustness
-	Fixes: debbugs:19755
+	(Bug#19755)
 	Thanks to Carlos Pita <carlosjosepita@gmail.com> for reporting
 	this and providing useful ideas.
 	* lisp/progmodes/python.el
@@ -952,7 +952,7 @@
 2015-04-08  Samer Masterson  <samer@samertm.com>
 
 	* lisp/eshell: Make backslash a no-op in front of normal chars
-	Fixes: debbugs:8531
+	(Bug#8531)
 	* lisp/eshell/esh-arg.el (eshell-parse-argument-hook): Update comment.
 	(eshell-parse-backslash): Return escaped character after backslash
 	if it is special.  Otherwise, if the backslash is not in a quoted
@@ -969,7 +969,7 @@
 
 	* lisp/vc/diff-mode.el (diff-hunk-file-names): Don't require a TAB
 	after the file name.
-	Fixes: debbugs:20276
+	(Bug#20276)
 
 2015-04-08  Paul Eggert  <eggert@cs.ucla.edu>
 
@@ -1006,7 +1006,7 @@
 2015-04-08  Stefan Monnier  <monnier@iro.umontreal.ca>
 
 	(eieio-copy-parents-into-subclass): Fix inheritance of initargs
-	Fixes: debbugs:20270
+	(Bug#20270)
 	* lisp/emacs-lisp/eieio-core.el (eieio-copy-parents-into-subclass):
 	Fix inheritance of initargs.
 
@@ -1023,7 +1023,7 @@
 2015-04-08  Eli Zaretskii  <eliz@gnu.org>
 
 	* src/eval.c (init_eval_once): Bump max_lisp_eval_depth to 800
-	Fixes: bug#17517
+	(Bug#17517)
 
 2015-04-08  Michael Albinus  <michael.albinus@gmx.de>
 
@@ -1043,7 +1043,7 @@
 2015-04-07  Dmitry Gutov  <dgutov@yandex.ru>
 
 	ffap: Support environment variable expansion in file names
-	Fixes: debbugs:19839
+	(Bug#19839)
 	* lisp/ffap.el (ffap-string-at-point-mode-alist): Support
 	environment variable expansion in file names.
 
@@ -1053,7 +1053,7 @@
 
 2015-04-07  Stefan Monnier  <monnier@iro.umontreal.ca>
 
-	Fixes: debbugs:20257
+	(Bug#20257)
 	* lisp/files.el (set-visited-file-name): Clear auto-save if nil.
 
 2015-04-07  Ivan Shmakov  <ivan@siamics.net>
@@ -1069,7 +1069,7 @@
 	Add doc strings for some Isearch state vars
 	* lisp/misearch.el (multi-isearch-buffer-list)
 	(multi-isearch-file-list): Add doc strings.
-	Fixes: bug#20232
+	(Bug#20232)
 
 2015-04-07  Alan Mackenzie  <acm@muc.de>
 
@@ -1110,7 +1110,7 @@
 	.git directory.  Distribute the new ChangeLog history files
 	instead of scattered ChangeLog files.  Distribute the new files
 	gitlog-to-changelog and gitlog-to-emacslog.
-	Fixes: bug#19113
+	(Bug#19113)
 
 	Rename ChangeLogs for gitlog-to-changelog
 	This patch was implemented via the following shell commands:
@@ -1146,4 +1146,3 @@ See ChangeLog.1 for earlier changes.
 ;; Local Variables:
 ;; coding: utf-8
 ;; End:
-
diff --git a/lisp/ChangeLog.17 b/lisp/ChangeLog.17
index 8984f48..a40f8f3 100644
--- a/lisp/ChangeLog.17
+++ b/lisp/ChangeLog.17
@@ -502,7 +502,7 @@
 
 2015-03-20  Alan Mackenzie  <acm@muc.de>
 
-	Fix debbugs#20146
+	Fix Bug#20146
 
 	* font-lock.el (font-lock-extend-jit-lock-region-after-change):
 	Return the calculated values, as per spec.
@@ -693,7 +693,7 @@
 
 2015-03-16  Alan Mackenzie  <acm@muc.de>
 
-	Edebug: Allow "S" to work during trace mode.  Fixes debbugs #20074.
+	Edebug: Allow "S" to work during trace mode.  Fixes Bug#20074.
 	Also display the overlay arrow in go and go-nonstop modes.
 
 	* emacs-lisp/edebug.el (edebug--display-1): Move the
diff --git a/src/ChangeLog.13 b/src/ChangeLog.13
index 33d0b9b..328f798 100644
--- a/src/ChangeLog.13
+++ b/src/ChangeLog.13
@@ -124,7 +124,7 @@
 	Fix minor ldexp issues
 	* floatfns.c (Fldexp): Require 2 args.  Avoid undefined behavior
 	if the exponent is out of 'int' range.  Improve documentation.
-	Fixes: bug#20185
+	(Bug#20185)
 
 2015-03-24  Daniel Colascione  <dancol@dancol.org>
 
@@ -215,7 +215,7 @@
 
 	Work around x86 glibc backtrace bug
 	* sysdep.c (emacs_backtrace): Don't dump core on x86.
-	Fixes: bug#19959
+	(Bug#19959)
 
 2015-03-05  Eli Zaretskii  <eliz@gnu.org>
 
@@ -1079,7 +1079,7 @@
 	to unused args.
 	* lisp.h (CALLMANY, CALLN): New macros.
 	* sound.c (Fplay_sound_internal): Coalesce duplicate code.
-	Fixes: bug#19634
+	(Bug#19634)
 
 	Use gnustep-config if available
 	* emacs.c [NS_IMPL_GNUSTEP]: Don't include <GNUstepBase/GSConfig.h>.
@@ -1103,7 +1103,7 @@
 
 	Don't downcase system diagnostics' first letters
 	* fileio.c (report_file_errno): Don't downcase, and simplify.
-	Fixes: bug#19642
+	(Bug#19642)
 
 	Isolate NIL_IS_ZERO-assuming code better
 	Suggested by Stefan Monnier in:
@@ -1148,7 +1148,7 @@
 
 	Correct an old fix for GTK font selection
 	* gtkutil.c (xg_get_font): Fix off-by-2 typo.
-	Fixes: bug#3228
+	(Bug#3228)
 
 	Fix minor bugs with printing null bytes
 	* minibuf.c (read_minibuf_noninteractive):
@@ -2137,7 +2137,7 @@
 	* fileio.c (Finsert_file_contents): Compute same_at_end character
 	position using the old buffer size, not the new one, since
 	restore_window_points wants the old size.
-	Fixes: debbugs:19161
+	(Bug#19161)
 
 2014-12-14  Alan Mackenzie  <acm@muc.de>
 
diff --git a/test/ChangeLog.1 b/test/ChangeLog.1
index 813f5dd..7f8eecf 100644
--- a/test/ChangeLog.1
+++ b/test/ChangeLog.1
@@ -307,7 +307,7 @@
 
 	* indent/shell.sh (bar): Use '[ $# -eq 0 ]', not '[ $# == 0 ]'.
 	This is more portable in shell scripts.
-	Fixes: bug#19658
+	(Bug#19658)
 
 2015-01-23  Dmitry Gutov  <dgutov@yandex.ru>
 
-- 
2.1.0


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #4: 0003-make-change-history-commit.patch --]
[-- Type: text/x-patch; name="0003-make-change-history-commit.patch", Size: 4458 bytes --]

From f48a28e34398551c719807c59819a5dab5f34aa8 Mon Sep 17 00:00:00 2001
From: Paul Eggert <eggert@cs.ucla.edu>
Date: Sat, 18 Apr 2015 19:51:18 -0700
Subject: [PATCH 3/4] ; make change-history-commit

---
 ChangeLog.2 | 96 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 Makefile.in |  2 +-
 2 files changed, 97 insertions(+), 1 deletion(-)

diff --git a/ChangeLog.2 b/ChangeLog.2
index 044f888..89fe77f 100644
--- a/ChangeLog.2
+++ b/ChangeLog.2
@@ -1,3 +1,99 @@
+2015-04-18  Paul Eggert  <eggert@cs.ucla.edu>
+
+	Prefer "Bug#1234" in commit messages
+	* .dir-locals.el (log-edit-mode): Don't rewrite Bug#,
+	as this isn't useful for Git.
+	* CONTRIBUTE: Suggest "Bug#1234" instead of "Fixes: debbugs:1234".
+
+2015-04-18  Glenn Morris  <rgm@gnu.org>
+
+	* lisp/files.el (auto-mode-alist): Use conf mode for gitconfig, hgrc.
+	Fixes: debbugs:19506
+
+2015-04-18  Tom Willemse  <tom@ryuslash.org>  (tiny change)
+
+	* lisp/elec-pair.el (electric-pair-post-self-insert-function): Do not use `chomp' as a function.
+	Fixes: debbugs:19505
+
+2015-04-18  Glenn Morris  <rgm@gnu.org>
+
+	* lisp/net/browse-url.el (browse-url, browse-url-at-point): Doc fixes.
+
+	* doc/emacs/misc.texi (Sorting): Small edit.
+	Fixes: debbugs:19896
+
+	* admin/admin.el (make-manuals): Add emacs-xtra in pdf and ps.
+
+2015-04-18  Simen Heggestøyl  <simenheg@gmail.com>
+
+	css-mode.el: Support multi-line comment filling
+	Fixes: debbugs:20256
+	* lisp/textmodes/css-mode.el (css-fill-paragraph): Support multi-line
+	comment filling.
+	(css-adaptive-fill): New function.
+	(css-mode): Set `adaptive-fill-function'.
+	(scss-mode): Set `comment-continue'.
+
+2015-04-18  Nicolas Petton  <nicolas@petton.fr>
+
+	* lisp/emacs-lisp/seq.el (seq-concatenate, seq-into): Better error messages.
+
+2015-04-18  Ivan Radanov Ivanov  <ivanradanov@yahoo.co.uk>  (tiny change)
+
+	Minor improvements in Bulgarian input methods
+	* lisp/leim/quail/cyrillic.el (bulgarian-phonetic, bulgarian-bds):
+	Replace U+042C with U+045D, as the former character is not used in
+	the modern Bulgarian language.
+	Fixes: Bug#20350
+
+2015-04-17  Thomas Fitzsimmons  <fitzsim@fitzsim.org>
+
+	Improve EUDC manual
+	* eudc.texi (LDAP Configuration): Mention simple and SASL
+	authentication schemes.  Add index items.  Shorten example server
+	name.
+
+2015-04-17  Dmitry Gutov  <dgutov@yandex.ru>
+
+	Don't show both feature and function with the same name
+	* lisp/progmodes/elisp-mode.el (elisp--xref-identifier-location):
+	Don't show both feature and function with the same name.
+
+	(elisp--xref-identifier-location): Skip variable, if it's also a functiong
+	* lisp/progmodes/elisp-mode.el (elisp--xref-identifier-location):
+	Avoid returning both the variable and the function for the same
+	minor mode.
+
+2015-04-17  Wolfgang Jenkner  <wjenkner@inode.at>
+
+	Fix fontification of keywords clobbered by the prompt.
+	* lisp/comint.el (comint-output-filter): Remove the uses of
+	with-silent-modifications I introduced as part of the last change.
+	This fixes, e.g., erratically missing highlighting when running
+	./configure --help; ./configure in a shell-mode buffer with
+	compilation-shell-minor-mode turned on.
+
+2015-04-17  Glenn Morris  <rgm@gnu.org>
+
+	* admin/authors.el (authors-valid-file-names, authors-renamed-files-alist): Additions.
+
+2015-04-17  Stefan Monnier  <monnier@iro.umontreal.ca>
+
+	* lisp/indent.el (indent-region): Don't deactivate the mark
+	Fixes: debbugs:20357
+
+2015-04-17  Sam Steingold  <sds@gnu.org>
+
+	lisp/net/rcirc.el (defun-rcirc-command): mark `target' as ignorable
+
+2015-04-16  Leo Liu  <sdl.web@gmail.com>
+
+	* lisp/progmodes/xref.el (xref-push-marker-stack): Add optional arg.
+
+2015-04-16  Stefan Monnier  <monnier@iro.umontreal.ca>
+
+	* lisp/erc/erc-pcomplete.el (erc-pcomplete): Don't use `pcomplete' any more.
+
 2015-04-16  Glenn Morris  <rgm@gnu.org>
 
 	* admin/authors.el (authors-lax-changelogs): Update for erc changes.
diff --git a/Makefile.in b/Makefile.in
index 5152a17..f065cd3 100644
--- a/Makefile.in
+++ b/Makefile.in
@@ -1092,7 +1092,7 @@ bootstrap: bootstrap-clean
 .PHONY: master-branch-is-current no-ChangeLog unchanged-history-files
 
 # The newest revision that should not appear in the generated ChangeLog.
-gen_origin = 455f13b8c91a51cd08e19fda737a4358a9a927d6
+gen_origin = 870287327b1fba7105599eeabac5a2aa5ebadf19
 
 # Convert git commit log to ChangeLog file.  make-dist uses this.
 ChangeLog:
-- 
2.1.0


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #5: 0004-Fix-more-recent-commit-messages-too.patch --]
[-- Type: text/x-patch; name="0004-Fix-more-recent-commit-messages-too.patch", Size: 2139 bytes --]

From 65ac8bc6a9e256b60c8ddfa3c99a1b28145a0763 Mon Sep 17 00:00:00 2001
From: Paul Eggert <eggert@cs.ucla.edu>
Date: Sat, 18 Apr 2015 19:54:58 -0700
Subject: [PATCH 4/4] ; Fix more-recent commit messages too

---
 ChangeLog.2 | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/ChangeLog.2 b/ChangeLog.2
index 89fe77f..90ef58c 100644
--- a/ChangeLog.2
+++ b/ChangeLog.2
@@ -1,6 +1,6 @@
 2015-04-18  Paul Eggert  <eggert@cs.ucla.edu>
 
-	Prefer "Bug#1234" in commit messages
+	Prefer "Bug#1234" in commit messages (Bug#20325)
 	* .dir-locals.el (log-edit-mode): Don't rewrite Bug#,
 	as this isn't useful for Git.
 	* CONTRIBUTE: Suggest "Bug#1234" instead of "Fixes: debbugs:1234".
@@ -8,26 +8,26 @@
 2015-04-18  Glenn Morris  <rgm@gnu.org>
 
 	* lisp/files.el (auto-mode-alist): Use conf mode for gitconfig, hgrc.
-	Fixes: debbugs:19506
+	(Bug#19506)
 
 2015-04-18  Tom Willemse  <tom@ryuslash.org>  (tiny change)
 
 	* lisp/elec-pair.el (electric-pair-post-self-insert-function): Do not use `chomp' as a function.
-	Fixes: debbugs:19505
+	(Bug#19505)
 
 2015-04-18  Glenn Morris  <rgm@gnu.org>
 
 	* lisp/net/browse-url.el (browse-url, browse-url-at-point): Doc fixes.
 
 	* doc/emacs/misc.texi (Sorting): Small edit.
-	Fixes: debbugs:19896
+	(Bug#19896)
 
 	* admin/admin.el (make-manuals): Add emacs-xtra in pdf and ps.
 
 2015-04-18  Simen Heggestøyl  <simenheg@gmail.com>
 
 	css-mode.el: Support multi-line comment filling
-	Fixes: debbugs:20256
+	(Bug#20256)
 	* lisp/textmodes/css-mode.el (css-fill-paragraph): Support multi-line
 	comment filling.
 	(css-adaptive-fill): New function.
@@ -44,7 +44,7 @@
 	* lisp/leim/quail/cyrillic.el (bulgarian-phonetic, bulgarian-bds):
 	Replace U+042C with U+045D, as the former character is not used in
 	the modern Bulgarian language.
-	Fixes: Bug#20350
+	(Bug#20350)
 
 2015-04-17  Thomas Fitzsimmons  <fitzsim@fitzsim.org>
 
@@ -80,7 +80,7 @@
 2015-04-17  Stefan Monnier  <monnier@iro.umontreal.ca>
 
 	* lisp/indent.el (indent-region): Don't deactivate the mark
-	Fixes: debbugs:20357
+	(Bug#20357)
 
 2015-04-17  Sam Steingold  <sds@gnu.org>
 
-- 
2.1.0


^ permalink raw reply related	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-19  2:59               ` Paul Eggert
@ 2015-04-19 14:23                 ` Eli Zaretskii
  0 siblings, 0 replies; 40+ messages in thread
From: Eli Zaretskii @ 2015-04-19 14:23 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325, michael.albinus, dgutov

> Date: Sat, 18 Apr 2015 19:59:02 -0700
> From: Paul Eggert <eggert@cs.ucla.edu>
> CC: michael.albinus@gmx.de, 20325@debbugs.gnu.org
> 
> Eli Zaretskii wrote:
> > For some value of "we".  CONTRIBUTE still requires to use
> > "Fixes: bug#NNNN", and that's what I've switched to once those
> > requirements were written.  Which was already a change from the
> > previous procedure, which preferred simply "Bug#NNNN".
> >
> > I wish you guys settle on something once and for all, never mind what,
> 
> "you guys"?  I thought you were one of us!  :-)

Not in this bikeshedding, sorry.

> I did not choose the "Fixes: bug#NNNN" syntax.  I vaguely recall it came from an 
> offhand remark in one of Stefan's emails a few months ago and that it migrated 
> into CONTRIBUTE when Stephen Leake attempted to summarize that discussion, but 
> possibly it was a typo in Stefan's message and anyway I expect he doesn't much 
> care what the format is.
> 
>  From the earlier discussion in this thread it appears that the "Fixes: 
> debbugs:NNNN" syntax is an artifact of the Bzr-to-Git conversion that now causes 
> more confusion than it's worth.  To try to lessen this I installed the attached 
> patches (the third one is mechanical; it's the result of 'make change-history').

Thanks.  Let's hope this one holds one a bit longer than its
predecessors.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-17  6:46 ` Paul Eggert
  2015-04-17 11:25   ` Dmitry Gutov
@ 2015-04-20 21:54   ` Glenn Morris
  2015-04-21  1:17     ` Stefan Monnier
  1 sibling, 1 reply; 40+ messages in thread
From: Glenn Morris @ 2015-04-20 21:54 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325

Paul Eggert wrote:

> Wouldn't it be better to use the same format in commit messages and in
> ChangeLog.2?  That will be less confusing for everybody.

I didn't have a problem with it (different formats), and judging from
the logs not many other people seemed to either. For a long time, Emacs
has provided the tools to seamlessly create a commit log in the right
format from a ChangeLog. I'd have preferred to see gitlog-to-changelog
get the small update needed to go in the other direction, rather than
having the Emacs convention changed (soon after we just had a brand-new
repo created using the "fixes: debbugs:123" convention throughout). (But
since it's already a fait accompli this is just me bitching rather than
asking you to change it back.)

Presumably this means the solution to http://debbugs.gnu.org/20324
will be to modify gitlog-to-changelog to recognize "(tiny change)".
Emacs will still need to be modified to pass (tiny change) through to
the VC log though.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-20 21:54   ` Glenn Morris
@ 2015-04-21  1:17     ` Stefan Monnier
  2015-04-21  3:41       ` Paul Eggert
  0 siblings, 1 reply; 40+ messages in thread
From: Stefan Monnier @ 2015-04-21  1:17 UTC (permalink / raw)
  To: Glenn Morris; +Cc: 20325, Paul Eggert

> repo created using the "fixes: debbugs:123" convention throughout). (But

I like "fixes: debbugs:123" and see no reason to change it.


        Stefan





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-21  1:17     ` Stefan Monnier
@ 2015-04-21  3:41       ` Paul Eggert
  2015-04-21 14:48         ` Stefan Monnier
  2015-04-21 14:49         ` Eli Zaretskii
  0 siblings, 2 replies; 40+ messages in thread
From: Paul Eggert @ 2015-04-21  3:41 UTC (permalink / raw)
  To: Stefan Monnier, Glenn Morris; +Cc: 20325

Stefan Monnier wrote:
> I like "fixes: debbugs:123" and see no reason to change it.

Hmm, that's news to me.  Until recently CONTRIBUTE said "Fixes: bug#NNNNN" 
because Eli suggested it 
<http://lists.gnu.org/archive/html/emacs-devel/2014-12/msg00892.html>, and you 
said that it sounded good and that you were also OK with "(bug#NNNNN)" 
<http://lists.gnu.org/archive/html/emacs-devel/2014-12/msg00917.html>.

I don't care what format is used, and it's fine with me to revert my attempt to 
bring order to this little corner of chaos.  Another possibility is to modify 
CONTRIBUTE to add "Fixes: debbugs:NNNNN" and "Fixes: bug#NNNNN" and etc. as 
alternative ways to spell "Bug#NNNNN".  It's no big deal either way.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-21  3:41       ` Paul Eggert
@ 2015-04-21 14:48         ` Stefan Monnier
  2015-04-24 18:35           ` Glenn Morris
  2015-04-21 14:49         ` Eli Zaretskii
  1 sibling, 1 reply; 40+ messages in thread
From: Stefan Monnier @ 2015-04-21 14:48 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325

>> I like "fixes: debbugs:123" and see no reason to change it.
> Hmm, that's news to me.  Until recently CONTRIBUTE said "Fixes: bug#NNNNN"
> because Eli suggested it
> <http://lists.gnu.org/archive/html/emacs-devel/2014-12/msg00892.html>, and
> you said that it sounded good and that you were also OK with "(bug#NNNNN)"
> <http://lists.gnu.org/archive/html/emacs-devel/2014-12/msg00917.html>.

I also like "Fixes: bug#NNNNN" and "(bug#NNNNN)".
But I don't see any reason to change from one to the other.
IOW, I'm not interested in changing the color of the bikeshed.


        Stefan





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-21  3:41       ` Paul Eggert
  2015-04-21 14:48         ` Stefan Monnier
@ 2015-04-21 14:49         ` Eli Zaretskii
  1 sibling, 0 replies; 40+ messages in thread
From: Eli Zaretskii @ 2015-04-21 14:49 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325

> Date: Mon, 20 Apr 2015 20:41:26 -0700
> From: Paul Eggert <eggert@cs.ucla.edu>
> Cc: 20325@debbugs.gnu.org
> 
> I don't care what format is used, and it's fine with me to revert my attempt to 
> bring order to this little corner of chaos.  Another possibility is to modify 
> CONTRIBUTE to add "Fixes: debbugs:NNNNN" and "Fixes: bug#NNNNN" and etc. as 
> alternative ways to spell "Bug#NNNNN".  It's no big deal either way.

Indeed.  But please, let's decide soon, and make that decision hold
for at least a year.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-21 14:48         ` Stefan Monnier
@ 2015-04-24 18:35           ` Glenn Morris
  2015-04-24 18:55             ` Paul Eggert
  0 siblings, 1 reply; 40+ messages in thread
From: Glenn Morris @ 2015-04-24 18:35 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: 20325, Paul Eggert

Stefan Monnier wrote:

>>> I like "fixes: debbugs:123" and see no reason to change it.
[...]
> I also like "Fixes: bug#NNNNN" and "(bug#NNNNN)".
> But I don't see any reason to change from one to the other.
> IOW, I'm not interested in changing the color of the bikeshed.

I don't even know what colour it is now. :(





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-24 18:35           ` Glenn Morris
@ 2015-04-24 18:55             ` Paul Eggert
  2015-04-24 19:13               ` Glenn Morris
  2015-04-24 20:27               ` Stefan Monnier
  0 siblings, 2 replies; 40+ messages in thread
From: Paul Eggert @ 2015-04-24 18:55 UTC (permalink / raw)
  To: Glenn Morris, Stefan Monnier; +Cc: 20325

On 04/24/2015 11:35 AM, Glenn Morris wrote:
> I don't even know what colour it is now. :( 

You could file a bug report about this, and mark it as "Bug#23196". :-)

For what it's worth, on the master branch since I changed CONTRIBUTE to 
suggest "Bug#NNNNN" on April 18, there have been

  7 commit messages saying "Bug#NNNNN" (all were parenthesized)
  4 commit messages saying "bug#NNNNN" (3 were parenthesized)
  3 commit messages saying "Fixes: debbugs:NNNNN" (all were on a 
separate line)

The unparenthesized "bug#NNNNN" was a separate line "Fixes bug#20266.".

It's early days, but it appears that "(Bug#NNNNN)" is the most popular 
color.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-24 18:55             ` Paul Eggert
@ 2015-04-24 19:13               ` Glenn Morris
  2015-04-24 19:19                 ` Glenn Morris
  2015-04-24 20:27               ` Stefan Monnier
  1 sibling, 1 reply; 40+ messages in thread
From: Glenn Morris @ 2015-04-24 19:13 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325

Paul Eggert wrote:

> It's early days, but it appears that "(Bug#NNNNN)" is the most popular
> color.

If you hadn't changed dir-locals, I assume "Fixes: debbugs:NNNN" would
remain the most popular.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-24 19:13               ` Glenn Morris
@ 2015-04-24 19:19                 ` Glenn Morris
  0 siblings, 0 replies; 40+ messages in thread
From: Glenn Morris @ 2015-04-24 19:19 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325


But I added the sed one-liner that does what I asked for in this report,
so just forget about it.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-24 18:55             ` Paul Eggert
  2015-04-24 19:13               ` Glenn Morris
@ 2015-04-24 20:27               ` Stefan Monnier
  2015-04-24 21:41                 ` Paul Eggert
  1 sibling, 1 reply; 40+ messages in thread
From: Stefan Monnier @ 2015-04-24 20:27 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325

>  7 commit messages saying "Bug#NNNNN" (all were parenthesized)
>  4 commit messages saying "bug#NNNNN" (3 were parenthesized)
>  3 commit messages saying "Fixes: debbugs:NNNNN" (all were on a separate
> line)

> It's early days, but it appears that "(Bug#NNNNN)" is the most
> popular color.

I think we shouldn't look at recent commits but at all commits in
the repository, since we want to reduce variation overall.


        Stefan





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-24 20:27               ` Stefan Monnier
@ 2015-04-24 21:41                 ` Paul Eggert
  2015-04-25  7:17                   ` Jan D.
  2015-04-25 14:47                   ` Stefan Monnier
  0 siblings, 2 replies; 40+ messages in thread
From: Paul Eggert @ 2015-04-24 21:41 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: 20325

On 04/24/2015 01:27 PM, Stefan Monnier wrote:
> I think we shouldn't look at recent commits but at all commits in
> the repository, since we want to reduce variation overall.
>

Then the counts are closer.  If one looks at the current 1,021,440-line 
output of 'git log', 3515 lines match the regular expression 
'bug#[0-9][0-9][0-9]' and 2906 lines match 'debbugs:[0-9][0-9][0-9]', 
ignoring case for both searches.

My impression is that most of the 'Fixes: debbugs:NNNNN' lines come from 
"bzr commit --fixes NNNNN", i.e., that they were not originally part of 
the commit message but were in Bzr metadata that was transferred into 
Git's commit messages because Git doesn't have metadata of this form.

I expect that most of the "by-hand" instances use the "Bug#NNNNN" form 
because it's easier to type, and now that we're doing this stuff more by 
hand the "Bug#NNNNN" form would become more popular even if CONTRIBUTE 
weren't recommending it.





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-24 21:41                 ` Paul Eggert
@ 2015-04-25  7:17                   ` Jan D.
  2015-04-25 14:47                   ` Stefan Monnier
  1 sibling, 0 replies; 40+ messages in thread
From: Jan D. @ 2015-04-25  7:17 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325


> 24 apr 2015 kl. 23:41 skrev Paul Eggert <eggert@cs.ucla.edu>:
> 
> My impression is that most of the 'Fixes: debbugs:NNNNN' lines come from "bzr commit --fixes NNNNN", i.e., that they were not originally part of the commit message but were in Bzr metadata that was transferred into Git's commit messages because Git doesn't have metadata of this form.

Indeed.  In the trunk log it looks like I have written 172 'Fixes: debbugs:NNNNN’, but I have not.
They all come from vc inserting that from the ChangeLogs Bug#NNNNN.

	Jan D.






^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-24 21:41                 ` Paul Eggert
  2015-04-25  7:17                   ` Jan D.
@ 2015-04-25 14:47                   ` Stefan Monnier
  2015-04-27 18:34                     ` Glenn Morris
  1 sibling, 1 reply; 40+ messages in thread
From: Stefan Monnier @ 2015-04-25 14:47 UTC (permalink / raw)
  To: Paul Eggert; +Cc: 20325

>> I think we shouldn't look at recent commits but at all commits in
>> the repository, since we want to reduce variation overall.
> Then the counts are closer.  If one looks at the current 1,021,440-line
> output of 'git log', 3515 lines match the regular expression
> 'bug#[0-9][0-9][0-9]' and 2906 lines match 'debbugs:[0-9][0-9][0-9]',
> ignoring case for both searches.

No need to count.  The convention was very clear and followed fairly
consistently: use (bug#NNN) in the ChangeLog, which gets turned into
"Fixes: debbugs:NNNN" in the *VC-Log* buffer (and was then turned into
a "--fixes debbugs:NNNN" argument to bzr which finally turned it into
a "Fixes: URL" message in the "bzr log" output).

Some/many (Bug#NNN) ended up in the commit log, and IIUC Eric's
conversion to Git turned Bzr's internal "Fixes" into "Fixes: debbugs:NNNN".

So, AFAIK, in terms of "pseudo-header syntax", the only thing we've used
consistently is "Fixes: debbugs:NNNN".


        Stefan





^ permalink raw reply	[flat|nested] 40+ messages in thread

* bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog
  2015-04-25 14:47                   ` Stefan Monnier
@ 2015-04-27 18:34                     ` Glenn Morris
  0 siblings, 0 replies; 40+ messages in thread
From: Glenn Morris @ 2015-04-27 18:34 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: 20325, Paul Eggert

Stefan Monnier wrote:

> No need to count.  The convention was very clear and followed fairly
> consistently: use (bug#NNN) in the ChangeLog, which gets turned into
> "Fixes: debbugs:NNNN" in the *VC-Log* buffer (and was then turned into
> a "--fixes debbugs:NNNN" argument to bzr which finally turned it into
> a "Fixes: URL" message in the "bzr log" output).
>
> Some/many (Bug#NNN) ended up in the commit log, and IIUC Eric's
> conversion to Git turned Bzr's internal "Fixes" into "Fixes: debbugs:NNNN".
>
> So, AFAIK, in terms of "pseudo-header syntax", the only thing we've used
> consistently is "Fixes: debbugs:NNNN".

Sounds to me like you just want to revert
http://lists.gnu.org/archive/html/emacs-diffs/2015-04/msg00301.html
then. And "s/bug#/debbugs:" in the resulting CONTRIBUTE file.
(No idea where "Fixes: bug#" came from, other than a misunderstanding.)

(But Emacs developers as a whole seem incapable of following
conventions, so it's likely pointless.)





^ permalink raw reply	[flat|nested] 40+ messages in thread

end of thread, other threads:[~2015-04-27 18:34 UTC | newest]

Thread overview: 40+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-04-14  6:01 bug#20325: convert "fixes:debbugs:123" in log to (Bug#123) in ChangeLog Glenn Morris
2015-04-17  6:46 ` Paul Eggert
2015-04-17 11:25   ` Dmitry Gutov
2015-04-17 15:19     ` Paul Eggert
2015-04-17 15:32       ` Dmitry Gutov
2015-04-17 15:56         ` Paul Eggert
2015-04-17 16:02           ` Dmitry Gutov
2015-04-17 16:25           ` Jan D.
2015-04-17 16:20         ` Jan D.
2015-04-17 16:30           ` Dmitry Gutov
2015-04-17 17:06             ` Jan D.
2015-04-17 17:15               ` Dmitry Gutov
2015-04-18  7:46                 ` Jan D.
2015-04-18  8:08                   ` Eli Zaretskii
2015-04-17 21:45     ` Michael Albinus
2015-04-17 21:48       ` Dmitry Gutov
2015-04-17 22:02         ` Michael Albinus
2015-04-17 22:44           ` Dmitry Gutov
2015-04-18  1:02             ` Paul Eggert
2015-04-18  6:28               ` Michael Albinus
2015-04-18  7:25               ` Eli Zaretskii
2015-04-18  9:25                 ` Ivan Shmakov
2015-04-18  9:46                   ` Eli Zaretskii
2015-04-18  7:14             ` Eli Zaretskii
2015-04-19  2:59               ` Paul Eggert
2015-04-19 14:23                 ` Eli Zaretskii
2015-04-20 21:54   ` Glenn Morris
2015-04-21  1:17     ` Stefan Monnier
2015-04-21  3:41       ` Paul Eggert
2015-04-21 14:48         ` Stefan Monnier
2015-04-24 18:35           ` Glenn Morris
2015-04-24 18:55             ` Paul Eggert
2015-04-24 19:13               ` Glenn Morris
2015-04-24 19:19                 ` Glenn Morris
2015-04-24 20:27               ` Stefan Monnier
2015-04-24 21:41                 ` Paul Eggert
2015-04-25  7:17                   ` Jan D.
2015-04-25 14:47                   ` Stefan Monnier
2015-04-27 18:34                     ` Glenn Morris
2015-04-21 14:49         ` Eli Zaretskii

Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).