Diff emails to debbugs?

Diff emails to debbugs?

[Lars Ingebrigtsen]

The diff mailing thing knows what debbugs the diff relates to:

fixes bug: http://debbugs.gnu.org/13274

So perhaps it would be nice if it Cc'd the debbugs address, so the bug
submitter could easily see what the resulting diff is?

-- 
(domestic pets only, the antidote for overdose, milk.)
  http://lars.ingebrigtsen.no  *  Lars Magne Ingebrigtsen

Re: Diff emails to debbugs?

[Glenn Morris]

Lars Ingebrigtsen wrote:

> So perhaps it would be nice if it Cc'd the debbugs address, so the bug
> submitter could easily see what the resulting diff is?

Or even if it could close the bug. Don't hold your breath:

https://bugs.launchpad.net/bzr-hookless-email/+bug/902643

Re: Diff emails to debbugs?

[Lars Ingebrigtsen]

Glenn Morris <rgm@gnu.org> writes:

> Or even if it could close the bug. Don't hold your breath:
>
> https://bugs.launchpad.net/bzr-hookless-email/+bug/902643

Do we have to wait for the Canonical people to fix this upstream, or
would it be possible to install local hacks on the GNU machine(s)?

-- 
(domestic pets only, the antidote for overdose, milk.)
  http://lars.ingebrigtsen.no  *  Lars Magne Ingebrigtsen

Re: Diff emails to debbugs?

[Glenn Morris]

Lars Ingebrigtsen wrote:

>> https://bugs.launchpad.net/bzr-hookless-email/+bug/902643
>
> Do we have to wait for the Canonical people to fix this upstream, or
> would it be possible to install local hacks on the GNU machine(s)?

I imagine the latter would be fine. Come up with a patch and open a
Savannah support request.

Re: Diff emails to debbugs?

[Stefan Monnier]

>> So perhaps it would be nice if it Cc'd the debbugs address, so the bug
>> submitter could easily see what the resulting diff is?
> Or even if it could close the bug.

Closing the bug wouldn't be a good idea: I regularly install patches
which fix part of a bug but not all of it.


        Stefan

Re: Diff emails to debbugs?

[Fabian Ezequiel Gallina]

2012/12/29 Stefan Monnier <monnier@iro.umontreal.ca>:
>>> So perhaps it would be nice if it Cc'd the debbugs address, so the bug
>>> submitter could easily see what the resulting diff is?
>> Or even if it could close the bug.
>
> Closing the bug wouldn't be a good idea: I regularly install patches
> which fix part of a bug but not all of it.
>
>
>         Stefan
>

This sounds nice,

Perhaps we can say something like "closes #12345" and that would mean
the bug must be closed, else if a bug number is found it would only CC
to debbugs.

WDYT?


Regards,
Fabián

Re: Diff emails to debbugs?

[Stephen J. Turnbull]

Stefan Monnier writes:
 > >> So perhaps it would be nice if it Cc'd the debbugs address, so the bug
 > >> submitter could easily see what the resulting diff is?
 > > Or even if it could close the bug.
 > 
 > Closing the bug wouldn't be a good idea: I regularly install patches
 > which fix part of a bug but not all of it.

Somewhat off-topic because it couldn't be done automatically: what I
do with partial fixes is to close the original bug, and mark it
superseded by a new bug the describes the reduced scope concisely.

Dunno if anybody else thinks it helps, but it does help me personally.

handling partially fixed bugs [was: Diff emails to debbugs?]

[Drew Adams]

> Somewhat off-topic because it couldn't be done automatically: what I
> do with partial fixes is to close the original bug, and mark it
> superseded by a new bug the describes the reduced scope concisely.
> 
> Dunno if anybody else thinks it helps, but it does help me personally.

It would be great if it were quick & simple to do that and also have the new bug
automatically reference the closed bug.

Depending on how it's implemented, the same manipulation could perhaps be used
to split a multiple-topics bug into several smaller bugs (automatically closing
the original and adding references to the new ones).

And in bug threads (e.g. at http://debbugs.gnu.org/), related threads would
hopefully contain links to their relations.

Re: Diff emails to debbugs?

[Glenn Morris]

Fabian Ezequiel Gallina wrote:

> Perhaps we can say something like "closes #12345" and that would mean
> the bug must be closed, else if a bug number is found it would only CC
> to debbugs.
>
> WDYT?

Parsing the commit message sounds like too much work. Just check for a
--fixes debbugs:123 entry, and if there is one, cc the commit to
123@debbugs with some short explanatory note.

Re: handling partially fixed bugs [was: Diff emails to debbugs?]

[Xue Fuqiao]

On Sat, 29 Dec 2012 08:35:53 -0800
"Drew Adams" <drew.adams@oracle.com> wrote:

> And in bug threads (e.g. at http://debbugs.gnu.org/), related threads would
> hopefully contain links to their relations.

I think so, too.
-- 
Best regards, Xue Fuqiao.
http://www.emacswiki.org/emacs/XueFuqiao

Re: Diff emails to debbugs?

[Lars Magne Ingebrigtsen]

Glenn Morris <rgm@gnu.org> writes:

> Parsing the commit message sounds like too much work. Just check for a
> --fixes debbugs:123 entry, and if there is one, cc the commit to
> 123@debbugs with some short explanatory note.

Parsing the commit messages is what I did at work to tie the VC and the
bug tracker together.  There we use magical strings in the commit
messages to control the bug tracker.  So we say "bug 123 fixed" and
stuff.  If there isn't recognised keyword after the bug report number,
the patch only refers to the bug (and the diff gets filed with it), but
it doesn't change the status.

An equivalent thing for Emacs would be to look for strings like
"bug#123 done 24.4".  The extra nice thing about this is that you can
refer to many bug reports at once.  Just say "(bug#123, bug#124 done)"
to have the diff go out to both the 123 and 124 bug reports, and close
the 124 bug report.

This might be over-engineering things, though.  And at the same time
under-engineering, since it relies on "magical strings".  :-)

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/