FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > ArchLinux > ArchLinux General Discussion

 
 
LinkBack Thread Tools
 
Old 03-12-2010, 11:20 PM
Heiko Baums
 
Default Allow comments on closed bugs?

Am Sat, 13 Mar 2010 08:39:05 +1000
schrieb Allan McRae <allan@archlinux.org>:

> I really do not see the need.
>
> If a bug is wrongly closed -> request a reopen.
> If you just want to confirm a bug has been fixed, there is no need...
> we already closed the bug report.
> If there is a "better" fix, reopen request or new bug report.

And this is forcing the reporter to begging for reopening and looking
again at the bug.

Closing a bug too early can in the reporter's sight mean: "Hey! I'm the
king and I decide if I'm willing to fix the bug. And if I don't want
to, then you don't have to say anything. Your subject to my merci."

Of course this is a bit exaggerated and of course in most cases the
developer doesn't mean it, but this is how the reporter can easily
understand it. And this can lead to such misunderstandings and to angry
reactions.

Don't see this only from your (the developer's) point of view. Try to
see it from the reporter's point of view.

Greetings,
Heiko
 
Old 03-12-2010, 11:58 PM
Dan McGee
 
Default Allow comments on closed bugs?

On Fri, Mar 12, 2010 at 4:35 PM, Aaron Griffin <aaronmgriffin@gmail.com> wrote:
> Does anyone have an opinion on this?
>
> In my eyes, I imagine the kind of people who want this feature simply
> wish to argue about the closing. I've had to deal with enough PM
> requests in the past to know that this _does_ happen.

Speaking from the pacman side more than the package-related bugs side,
I've often wanted to add comments to closed issues in the past and
have been unable to. For things like a performance bug, it is often
helpful a few days down the road to post results of the fix once a few
people have tested it. It doesn't make sense to leave the bug open if
it is fixed in git, but many people that followed the bug would find
it interesting.

Other times I get a reopen request asking "what version will this be
fixed in?" That is exactly the thing that is better done in comments,
because 1) if I do reopen, I have to retype their request and 2) then
it is there for everyone to see.

Finally, we never get a chance to see it, because on other bug
trackers people often leave nice comments like "thanks for fixing so
quickly!" or "works for me now too, thanks a lot". If you wonder why
you never feel appreciated for fixing bugs around here, it might be
because our policy leaves no way to give or receive thanks. It sounds
like an odd thing, but I can see people getting burned out with fixing
issues if they never once get positive feedback or a well-deserved pat
on the back.

-Dan
 
Old 03-13-2010, 12:22 AM
Robert Howard
 
Default Allow comments on closed bugs?

What do other distros do on their bugtrackers? We should allow comments
after closing to facilitate further user input. Lets not forget that Arch
Linux would not be in it's current state without user/dev interaction.

On Mar 12, 2010 7:19 PM, "Heiko Baums" <lists@baums-on-web.de> wrote:

Am Sat, 13 Mar 2010 08:39:05 +1000
schrieb Allan McRae <allan@archlinux.org>:


> I really do not see the need.
>
> If a bug is wrongly closed -> request a reopen.
> If you just ...
And this is forcing the reporter to begging for reopening and looking
again at the bug.

Closing a bug too early can in the reporter's sight mean: "Hey! I'm the
king and I decide if I'm willing to fix the bug. And if I don't want
to, then you don't have to say anything. Your subject to my merci."

Of course this is a bit exaggerated and of course in most cases the
developer doesn't mean it, but this is how the reporter can easily
understand it. And this can lead to such misunderstandings and to angry
reactions.

Don't see this only from your (the developer's) point of view. Try to
see it from the reporter's point of view.

Greetings,
Heiko
 
Old 03-13-2010, 02:06 AM
Heiko Baums
 
Default Allow comments on closed bugs?

Am Fri, 12 Mar 2010 20:22:35 -0500
schrieb Robert Howard <rjh0507@ecu.edu>:

> What do other distros do on their bugtrackers? We should allow
> comments after closing to facilitate further user input. Lets not
> forget that Arch Linux would not be in it's current state without
> user/dev interaction.

I used Gentoo for several years before I switched to Arch Linux a few
years ago.

On the Gentoo bug tracker I had the impression that every bug report is
taken seriously. The developers are not so easily annoyed by invalid
bug reports where a user just have missed an option in his system
configuration, because this can happen to everyone, or hasn't enough
knowledge. And the Gentoo Bugzilla isn't degenerated to a support
forum. If it turns out that a bug report is just caused by a wrong
configuration the developers or users who read the bug report usually
just explain what is wrong and how to fix it. Or they point to the
forums or the documentation and what to search for.

If the reporter and the developer disagree the issue is discussed
before a bug is just be closed as "won't fix". How long the discussion
takes or the result of the discussion depends on the bug. In not a few
cases some or many other users and developers take part on such
discussions.

If a developer can't reproduce a bug he usually tells it in a comment
without closing the bug. A bug is usually only closed if the bug is
definitely fixed or if the fix or the invalidity is confirmed by the
reporter or another user who has this issue, too. If the developer
couldn't reproduce the bug he usually asks for testing the fix.

There are still some bugs - at least one of mine - open since several
years with several duplicates. Usually these are annoying but not
the most important issues.

If a bug is closed at once - usually this is only done by bug
wranglers, but not by developers - the bug report can easily be
reopened by the reporter. And if a bug is closed too early the bug
wrangler usually gives a reason for this in the comments, and the
reporter can easily reply with a comment. In the cases I know, then the
bug was kept open and the developer to whom it was assigned deals with
it and decides what to do.

Well, Gentoo has a lot more developers than Arch Linux, so Gentoo has
more manpower than Arch Linux. But I bet, this could be changed on Arch.

I would sum it up a bit simplified that Gentoo is more user than
developer driven while Arch Linux is currently more developer than user
driven.

It's not that the users can't file feature requests or take part on
discussions with the Arch developers. And usually the Arch developers
listen to the users. But I read too many times sentences like "Arch
is/was from developers for developers", "the developers only maintain,
what they want", etc. And too many times some developers speak openly
that they don't like Arch's growing user community. This somehow keeps
the users and their needs and wishes out.

Yes, I know, this is not quite right, but sometimes I have a bit the
impression. The early bug closing issue is one of the reasons for this
impression.

Also AUR is usually seen as unofficial by the developers, because the
packages are merely made by usual users. Sometimes I'd prefer if AUR
would be seen as unstable but official a bit like Gentoo's sunrise
overlay. Of course the AUR maintainers don't need to and shouldn't get
developer or TU status just because they maintain such a package.

The user/dev interaction is, btw., the engine of the whole OpenSource
community.

Heiko
 
Old 03-13-2010, 02:55 AM
Paul Mattal
 
Default Allow comments on closed bugs?

On 03/12/2010 06:42 PM, Andrea Scarpino wrote:

On Friday 12 March 2010 23:39:05 Allan McRae wrote:

I really do not see the need.

If a bug is wrongly closed -> request a reopen.
If you just want to confirm a bug has been fixed, there is no need...
we already closed the bug report.
If there is a "better" fix, reopen request or new bug report.

+1
we do not need comments on closed bug



+1 from me. Reopen takes care of this.

- P
 
Old 03-13-2010, 03:59 AM
Eric Bélanger
 
Default Allow comments on closed bugs?

On Fri, Mar 12, 2010 at 10:55 PM, Paul Mattal <paul@mattal.com> wrote:
> On 03/12/2010 06:42 PM, Andrea Scarpino wrote:
>>
>> On Friday 12 March 2010 23:39:05 Allan McRae wrote:
>>>
>>> I really do not see the need.
>>>
>>> If a bug is wrongly closed -> *request a reopen.
>>> If you just want to confirm a bug has been fixed, there is no need...
>>> we already closed the bug report.
>>> If there is a "better" fix, reopen request or new bug report.
>>
>> +1
>> we do not need comments on closed bug
>>
>
> +1 from me. Reopen takes care of this.

Another +1 from me.
 
Old 03-13-2010, 06:57 AM
Giovanni Scafora
 
Default Allow comments on closed bugs?

Il 12/03/2010 23:35, Aaron Griffin ha scritto:

Does anyone have an opinion on this?

In my eyes, I imagine the kind of people who want this feature simply
wish to argue about the closing. I've had to deal with enough PM
requests in the past to know that this _does_ happen.

Opinions?


I do not see the need.


--
Arch Linux Developer
http://www.archlinux.org
http://www.archlinux.it
 
Old 03-13-2010, 08:58 AM
Pierre Schmitz
 
Default Allow comments on closed bugs?

Am Samstag, 13. März 2010 01:58:02 schrieb Dan McGee:
> I've often wanted to add comments to closed issues in the past and
> have been unable to. For things like a performance bug, it is often
> helpful a few days down the road to post results of the fix once a few
> people have tested it. It doesn't make sense to leave the bug open if
> it is fixed in git, but many people that followed the bug would find
> it interesting.

Yes, there are sometimes situations where I wanted to add a more or less
important comment to a bug reports but it was just closed.

In the other hand I see the problems of never ending discussions and flame
wars.

But as we don't know which of these cases occur more often we could just
enable that feature for a month and decide afterwards.

--

Pierre Schmitz, https://users.archlinux.de/~pierre
 
Old 03-13-2010, 11:55 AM
Heiko Baums
 
Default Allow comments on closed bugs?

Am Sat, 13 Mar 2010 10:58:24 +0100
schrieb Pierre Schmitz <pierre@archlinux.de>:

> Yes, there are sometimes situations where I wanted to add a more or
> less important comment to a bug reports but it was just closed.
>
> In the other hand I see the problems of never ending discussions and
> flame wars.

In other bug trackers I haven't seen never ending discussions and flame
wars on closed bugs, yet. Sometimes long discussions happen on open
bugs, but usually in a factual manner, and usually there's a reason for
this, because it's a controversial issue. If a comment is written on a
closed bug it's just to give some more details, because the fix is not
quite sufficient or could be made still better or the like.

And in cases in which it's possible to reopen a bug directly without
sending a request, I also haven't seen exploitations. Usually the bug is
reopened once or twice to give another argument or another aspect. But
then it's alright.

How many bug reports are actually invalid because of an imperfect
knowledge or (search) laziness of the reporter? How many bug reports
are of the type Aaron Griffin has mentioned before (Feature request,
closed as "won't implement", reopend with "but it's a good feature",
denied with "we won't implement this, wait for upstream", reopened,
denied)? Are there really so many of them? I have my doubts. Or is this
more the developer's fear that this could happen?

And as I've written in other e-mails, don't see only your developer's
point of view. See also the user's and reporter's point of view and how
a certain bug handling (early closings, forcing reopening requests
(begging), etc.) is or at least can be received by the user/reporter.

Btw., comments on closed bugs is more or less a minor issue. The bigger
issue is the early bug closing. ;-)

Heiko
 
Old 03-13-2010, 12:54 PM
Allan McRae
 
Default Allow comments on closed bugs?

On 13/03/10 22:55, Heiko Baums wrote:

Am Sat, 13 Mar 2010 10:58:24 +0100
schrieb Pierre Schmitz<pierre@archlinux.de>:


Yes, there are sometimes situations where I wanted to add a more or
less important comment to a bug reports but it was just closed.

In the other hand I see the problems of never ending discussions and
flame wars.


In other bug trackers I haven't seen never ending discussions and flame
wars on closed bugs, yet. Sometimes long discussions happen on open
bugs, but usually in a factual manner, and usually there's a reason for
this, because it's a controversial issue. If a comment is written on a
closed bug it's just to give some more details, because the fix is not
quite sufficient or could be made still better or the like.

And in cases in which it's possible to reopen a bug directly without
sending a request, I also haven't seen exploitations. Usually the bug is
reopened once or twice to give another argument or another aspect. But
then it's alright.


You obviously do not visit the glibc tracker. Then again, perhaps there
is a reason for such things in that particular project... And I have
seen similar things on other distros trackers where reopens by users are
allowed (yes, including Gentoo), although they obviously are not the
usual. Then again, most bugs do not get re-open requests in Arch.



How many bug reports are actually invalid because of an imperfect
knowledge or (search) laziness of the reporter? How many bug reports
are of the type Aaron Griffin has mentioned before (Feature request,
closed as "won't implement", reopend with "but it's a good feature",
denied with "we won't implement this, wait for upstream", reopened,
denied)? Are there really so many of them? I have my doubts. Or is this
more the developer's fear that this could happen?


Yes, it occurs and not infrequently. The question should be does it
occur more often than that bugs are closed to early? I.e. which
solution would cause less total annoyance. Note that one option focuses
the annoyance on a small number of devs, while the other spreads it out
across many users.



And as I've written in other e-mails, don't see only your developer's
point of view. See also the user's and reporter's point of view and how
a certain bug handling (early closings, forcing reopening requests
(begging), etc.) is or at least can be received by the user/reporter.


We are also users... According to flyspray, I have opened 126 bugs and
I do not recall ever requesting or reopening one after it was closed.


I do not see all reopen requests, but the need to "beg" seems
overstated... I do know that it is much, much easier to get a bug
reopened if the request is clear and well justified. A large portion of
reopen requests provide no information to properly judge their merit in
which case they are more likely denied.


Allan
 

Thread Tools




All times are GMT. The time now is 02:28 AM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright ©2007 - 2008, www.linux-archive.org