gnus bugs tracking

gnus bugs tracking

[Alexander Kotelnikov]

Hi.

Sometimes bugs reported to ding@gnus.org are forgotten without a fix.
It's a pity. And it is a trouble, since gnus does not have its own bug
tracking system.

Any comments?
-- 
Alexander Kotelnikov
Saint-Petersburg, Russia

Re: gnus bugs tracking

[Vasily Korytov]

[-- Attachment #1: Type: text/plain, Size: 645 bytes --]

On 22 Mar 2003, Alexander Kotelnikov wrote:

> Sometimes bugs reported to ding@gnus.org are forgotten without a fix.
> It's a pity. And it is a trouble, since gnus does not have its own bug
> tracking system.
>
> Any comments?

AFAIK, bugs are not meant to be reported on the ding list.

,----[ C-h f gnus-bug RET ]
| `gnus-bug' is an interactive compiled Lisp function
|   -- loaded from "gnus-msg"
| (gnus-bug)
| 
| Documentation:
| Send a bug report to the Gnus maintainers.
`----

-- 
       I accept RFC3156 and RFC1991-compatible encrypted mail.
PGP key fingerprint: 123A 7CCE 6E26 6233 0D87 E01A A0F8 3524 FCD8 1841

[-- Attachment #2: Type: application/pgp-signature, Size: 188 bytes --]

Re: gnus bugs tracking

[Alexander Kotelnikov]

>>>>> On Sat, 22 Mar 2003 15:42:13 +0300
>>>>> "VK" == Vasily Korytov <deskpot@myrealbox.com> wrote:
VK> 
VK> On 22 Mar 2003, Alexander Kotelnikov wrote:
>> Sometimes bugs reported to ding@gnus.org are forgotten without a fix.
>> It's a pity. And it is a trouble, since gnus does not have its own bug
>> tracking system.
>> 
>> Any comments?
VK> 
VK> AFAIK, bugs are not meant to be reported on the ding list.

Then, how do I check if a bug I intend to submit is not already
reported? How do I see if some actions are performed to fix a bug?


-- 
Alexander Kotelnikov
Saint-Petersburg, Russia

Re: gnus bugs tracking

[Frank Schmitt]

Alexander Kotelnikov <sacha@myxomop.com> writes:

> VK> AFAIK, bugs are not meant to be reported on the ding list.
>
> Then, how do I check if a bug I intend to submit is not already
> reported?

Well, you look in news://news.gnus.org/gnus.gnus-bug

> How do I see if some actions are performed to fix a bug?

You look if a developer had answered "Yupp, fix in CVS" or "I'll look
into it"

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.

Re: gnus bugs tracking

[Simon Josefsson]

Alexander Kotelnikov <sacha@myxomop.com> writes:

> Hi.
>
> Sometimes bugs reported to ding@gnus.org are forgotten without a fix.
> It's a pity. And it is a trouble, since gnus does not have its own bug
> tracking system.
>
> Any comments?

Adding meta-data management to the current gnus.gnus-bug is one useful
improvement I can see.  Like, e.g., making it possible to enter the
summary buffer and list all "open" bugs, and be able to "close" them
somehow.  I'm not sure how to best implement this though.  With IMAP
it would have been simple, by using "gnus-bug-open", "gnus-bug-fixed",
"gnus-bug-wontfix" etc flags on the messages.

Web based bug tracker systems are just a time sink.  The bugs are
tracked, but less frequently fixed because developers cannot
prioritize.  In mail, if a problem is serious many people will report
it many times until it is fixed.  So it has a built-in flexible
adaptive priority system which most web BTS's lack.

Re: gnus bugs tracking

[Frank Schmitt]

Simon Josefsson <jas@extundo.com> writes:

> Adding meta-data management to the current gnus.gnus-bug is one useful
> improvement I can see.  Like, e.g., making it possible to enter the
> summary buffer and list all "open" bugs, and be able to "close" them
> somehow.  I'm not sure how to best implement this though.  With IMAP
> it would have been simple, by using "gnus-bug-open", "gnus-bug-fixed",
> "gnus-bug-wontfix" etc flags on the messages.

Well, you could do it the hard way by allowing the developers to
supersede the messages (only changing a X-bug-status header), but this
most probably won't work because of cancel-lock (well, we could tell
Gnus not to generate cancel-locks when creating a gnus-bug message).

The other option would be the following:
1) Every message created by M-x gnus-bug gets a special header like
X-Gnus-bug: report

2) If you e.g. fixed the bug you call a function gnus-bug-fixed with
point over the message which creates a mail buffer with  
X-Gnus-bug: Report and the tag [fixed] in the subject

Other possible bug statuses would be
e.g. [fixed], [in-work], [postponed], [infos-missing], [no-bug]

3) If you want to see all open bugs, you look for every message with 
X-Gnus-bug: Report
if it has a child with 
X-Gnus-bug: fixed

This shouldn't be too hard to implement, but would probably mean two
runs over all headers (however, gnus-bug isn't very crowded, so this
should be not too slow).

-- 
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.

Re: gnus bugs tracking

[Alex Schroeder]

Simon Josefsson <jas@extundo.com> writes:

> Web based bug tracker systems are just a time sink.  The bugs are
> tracked, but less frequently fixed because developers cannot
> prioritize.  In mail, if a problem is serious many people will report
> it many times until it is fixed.  So it has a built-in flexible
> adaptive priority system which most web BTS's lack.

I like this approach.  Web based bug tracking systems only work well
when the number of bugs and reports per bug is small enough to make a
tool like reportbugs (Debian) feasible.  The current approach in
Emacs and Gnus development works well enough.

Alex.

Re: gnus bugs tracking

[Alexander Kotelnikov]

>>>>> On Sat, 22 Mar 2003 17:05:34 +0100
>>>>> "SJ" == Simon Josefsson <jas@extundo.com> wrote:
SJ> 
SJ> Alexander Kotelnikov <sacha@myxomop.com> writes:
>> Hi.
>> 
>> Sometimes bugs reported to ding@gnus.org are forgotten without a fix.
>> It's a pity. And it is a trouble, since gnus does not have its own bug
>> tracking system.
>> 
>> Any comments?
SJ> 
SJ> Adding meta-data management to the current gnus.gnus-bug is one useful
SJ> improvement I can see.  Like, e.g., making it possible to enter the
SJ> summary buffer and list all "open" bugs, and be able to "close" them
SJ> somehow.  I'm not sure how to best implement this though.  With IMAP
SJ> it would have been simple, by using "gnus-bug-open", "gnus-bug-fixed",
SJ> "gnus-bug-wontfix" etc flags on the messages.

but gnus.gnus-bug is an nntp, not an imap group, isn't it?

SJ> Web based bug tracker systems are just a time sink.  The bugs are
SJ> tracked, but less frequently fixed because developers cannot
SJ> prioritize.  In mail, if a problem is serious many people will report
SJ> it many times until it is fixed.  So it has a built-in flexible
SJ> adaptive priority system which most web BTS's lack.

I am almost sure that most of existing and working BTS have notion of
priority. Even I would not call them "web BTS", since frontend is not
the main part of most databases.

-- 
Alexander Kotelnikov
Saint-Petersburg, Russia

Re: gnus bugs tracking

[Alexander Kotelnikov]

>>>>> On Sat, 22 Mar 2003 17:05:34 +0100
>>>>> "SJ" == Simon Josefsson <jas@extundo.com> wrote:
SJ> 
SJ> Alexander Kotelnikov <sacha@myxomop.com> writes:
>> Hi.
>> 
>> Sometimes bugs reported to ding@gnus.org are forgotten without a fix.
>> It's a pity. And it is a trouble, since gnus does not have its own bug
>> tracking system.
>> 
>> Any comments?
SJ> 
SJ> Adding meta-data management to the current gnus.gnus-bug is one useful
SJ> improvement I can see.  Like, e.g., making it possible to enter the
SJ> summary buffer and list all "open" bugs, and be able to "close" them
SJ> somehow.  I'm not sure how to best implement this though.  With IMAP
SJ> it would have been simple, by using "gnus-bug-open", "gnus-bug-fixed",
SJ> "gnus-bug-wontfix" etc flags on the messages.

BTW, there is no good way to perform searching on bugs.

Ad the thing I forgot to ask: is it ok to post bug reports directly to
gnus-bug?

-- 
Alexander Kotelnikov
Saint-Petersburg, Russia

Re: gnus bugs tracking

[Lars Magne Ingebrigtsen]

Simon Josefsson <jas@extundo.com> writes:

> Adding meta-data management to the current gnus.gnus-bug is one useful
> improvement I can see.  Like, e.g., making it possible to enter the
> summary buffer and list all "open" bugs, and be able to "close" them
> somehow.

Yup.

> I'm not sure how to best implement this though.

I think the most logical thing is to just do it by following up with
articles in a formally defined format, and having the news client do
something clever.  That requires that all the maintainers use the same
client, though.

> Web based bug tracker systems are just a time sink.

Yup.  I just can't understand how people can work with those things.
They're so unwieldy and primitive and time-consuming.

Bug tracking should be done from within Emacs, like nature intended.
:-)

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen