[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: QA Team [JRT]
From:       Thiago Macieira <thiago.macieira () kdemail ! net>
Date:       2004-01-31 4:57:29
Message-ID: 200401310257.37026.thiago.macieira () kdemail ! net
[Download RAW message or body]

[Attachment #2 (multipart/signed)]


This is almost turning into a flame-war, so I'll try to be careful with 
my words. I no way I am attacking your person, James, and I applaud 
your work and your effort, even if at sometimes we are at odds. Since I 
was named in the discussion, I will add my voice to it, again.

James Richard Tyrer wrote:
>You are making a generalization.  Clearly, it depends on just how bad
> the regressions are.  If you set the Focus Policy to: "Focus Strictly
> Under Mouse", set "Focus stealing prevention level" to anything other
> than: "None", and try to do any serious work with The GIMP you will
> see why I said what I did (note that there are other problems with
> the same cause). Perhaps the point from this is that a new feature
> that doesn't totally work should be turned OFF by default.

That's also a generalisation. Partial support is better than no support. 
But not in all cases. It is a serious problem if you hit the bug in 
your daily work, but it's not true for everyone involved and most 
surely not for the developer (or he'd have fixed it a long time ago).

So, yes, it's the job of the bug reporter to show that his bug is 
important.

>> So while Thiago misunderstands people like anyone else does, I would
>> always stand up and get in front of him to defend him from people
>> that think their bug reports are more worth than the words they use
>> to express them.
>
>If you mean that I thought that my bug report was worth more than
> *nothing* then I guess I see your point that you don't want people
> that report bugs to expect that they will be fixed.

That's not how I read Coolo's words. I think he meant "a bug report is 
worth the words written in it", meaning a lousily-written bug report is 
worth next to nothing; a difficult to understand bug report is worth 
less than a well-thought one -- irrespective of the severity of the bug 
being reported. 
(Example: we've got serious crashes being reported, but the report 
didn't contain any description, only a backtrace that ended before any 
function was shown. That's one example of a completely useless report.)

We can't fix bugs if we don't understand what the report is about. We 
can't fix bugs if we can't reproduce them. And we can't fix ALL bugs. 
The sheer volume of bug reports that we get (300 per day?) is too high: 
some applications can keep up, others simply can't.

Konqueror is certainly an example of one that can't. So, if you or 
anyone else report a bug on Konqueror, chances of it getting fixed -- 
or even looked at! -- are higher if you provide a good bug report and 
make your point clear. (Patches most surely welcome; as are valgrind 
logs, backtraces with symbols, etc.)

>You are wrong to defend what he did.

I don't think you have the right to say this.

>It was not that he might have 
>misunderstood, it was what he did as a result.  He dismissed it out of
> hand and then offered rationalizations to support his decision.  What
> should he have done?  If he didn't understand, he should have asked
> questions.  If he needed more information he should have asked for
> it.

Let's review what I did: 
http://lists.kde.org/?l=kde-bugs-dist&m=107491405208762&w=2

Did I close the bug report? No. 
Did I change the severity or priority? No. 
Did I misunderstand you? Yes, I did -- but of course, I didn't know it 
at that time. 
Did I know how to fix the bug? No. 
Did I provide my _opinion_? Yes.

You said "This is the wrong widget." for a widget asking for a new name; 
I answered "I don't see how this can be the wrong widget.", asked what 
was wrong about it, then I agreed ("It should, however, offer the 
option to overwrite as well. ") and then I offered my opinion. And my 
opinion was (and still would be, if the bug was restricted to split 
windows, as I had understood) that the bug is no showstopper, however 
severe.

Please tell me what was wrong with my attitude, because I still can't 
see it. I will make the effort to not make the same mistakes again.

Note that we ask that the priority field be left to the developers only 
-- and you were changing it.

>> And bugs _can_ be reopened, and if we (either the developer or the
>> bugs squad) think that bugs are fixed, then it's the responsibility
>> of the bug reporter to convince us the bug is still there (and as
>> long as KDE is a free software project and we've got no [paid] QA
>> team, this will stay the same). Subscribe to kde-bugs-dist@kde.org
>> and be part of the process before you complain about it.
>
>Then it is my understanding that you are opposed to having an unpaid
> QA team or to trying to get the bug reporters to function as a QA
> team.  I hope that this doesn't mean that you are opposed to QA in
> general.

I think you read too much into people's words. He didn't say anything 
about not wanting to have a QA Team. He said only that with the absence 
of one such team, the responsability of verifying that a bug has been 
fixed falls to the person who reported it.

By the bracketing of "paid", I understand that an unpaid team can't be 
held responsible for ALL bugs either. Besides, that team that is 
organising itself is already involving itself with code cleanups, so it 
might be that they don't take to verifying of bugs AT ALL.

If you want to do that -- and organise a team of volunteers that check 
that fixed bugs are really fixed --, by all means, go for it. The 
project will thank you many times over. If you want to go to the list 
of open bug reports on Konqueror file manager and sort them according 
to priorities to be fixed, the Konqueror devels will like it a lot.

In sum, there are many ways you can help and providing bug reports as 
you are doing is one of them. You just can't expect your opinion to be 
taken at face-value at all times.

Finally, we don't close bug reports about confirmed bugs just because we 
don't like the bug report. That doesn't happen. The closest thing that 
can happen to that is a WONTFIX, which is something else.

-- 
  Thiago Macieira  -  Registered Linux user #65028
   thiagom (AT) mail (dot) com
    ICQ UIN: 1967141   PGP/GPG: 0x6EF45358; fingerprint:
    E067 918B B660 DBD1 105C  966C 33F5 F005 6EF4 5358

[Attachment #5 (application/pgp-signature)]

>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<


[prev in list] [next in list] [prev in thread] [next in thread]