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

List:       kde-devel
Subject:    RE: QA Team [JRT]
From:       "CARVALHO Luis Passos" <luis.passos () enabler ! com>
Date:       2004-01-30 12:56:00
Message-ID: 7190B271481B184F968027E7C0B3ADE901EB8EB2 () prt01ex02 ! enabler ! corp
[Download RAW message or body]

[Attachment #2 (unknown)]


> -----Original Message-----
> From: Stephan Kulow [mailto:coolo@kde.org]
> 
> Am Freitag 30 Januar 2004 03:58 schrieb James Richard Tyrer:
> > they they may go back.   The 3_2_BRANCH from yesterday has 
> bugs that if they
> > are not fixed, I will go back to 3.1.5.
> Enjoy it!
> 
> Greetings, Stephan
> 

I'm surprised. I'm not used to see this type of answer, particularly from you 
Mr. Kulow. I can only assume that this answer must be motivated by writing an 
e-mail at 4am.

I've read this thread up to now and don't understand what's the problem with 
what Mr. Tyrer has said up to know, that would cause all these aggressive 
reactions.

Mr Tyrer made several points:
	- Our current development relies on bug reporters to ensure that bugs get
closed which is, for the most part, true. We do have some people that go after
bugs to see if they have already been fixed but not marked as resolved (Mr. Kulow 
does an impressive job) but noone to really close the bugs.
	- He mentioned that some bugs get resolved too soon or aren't assumed to be
as critical as the person that reports them feels. That much is also true. We
have a lot of re-opened bugs, and the person who submits the bug always feels
that his/her bug is critical.
	- In his opinion we should have a QA Team, responsible for visiting the
bugs and identifying if they have been really resolved or not. I'd add also
clarifying bug reports and obtaining test cases and additional information from
the user.
	- "Regressions are showstoppers". Some of you may not work with Windows 
most of the time and may not remember what frustrated us in Windows. I have to 
work with it all the time at work. I hate it when something breaks suddenly and 
it had been working. Although some regressions aren't severe enough to be 
showstoppers, regressions do seem, for Joe User, like sloppy work. They don't 
undertand the complexity of the underlying software and they don't care. They 
could do something now they can't. You put yourself in that place and tell me how 
you'd feel.

Someone replied that he was making "airy fairy suggestions". He did offer to 
help with the QA Team didn't he? It's an offer to fulfill a role. If Mr. Kulow has
someone to help him create test cases and validate that bugs are fixed or resolved
I'm sure he'd be more than happy.

Then regarding the bug he mentioned there was a clarification that it was 
important that the bug reporter was clear with the bug description so that the
criticallity of the bug be better determined by the developer. In a perfect world
everyone expresses him/herself very clearly, so that anyone can understand. We don't
live in a perfect world. In fact, I'd say most of us have difficulty in
expressing our ideas, since most of us aren't writing/reading in our native 
languages. Being clear the first time is not always going to happen. The QA team, 
as someone referred also has the job of clarifying bug reports.

Please people. Calm down and look at this with a cool head. We have 1365 bugs in
konqueror alone! Their descriptions many times refer mis-behaviour in rendering
a specific site. How many of these are triggered by the same type of code? Wouldn't
it be great if someone visited each one of these sites and determined the offending
code? Don't we need such people? And we as developers (yes, I'm a developer too),
should be happy that someone is willing to follow bugs and ask of us that we solve
the problems we have in the code. And "nagging". I'd much rather have a person that 
is part of the project "nagging" me than release the code with a flaw that will
prevent people from upgrading.

Just as an example: nwn.bioware.com was rendering horribly. Suddenly, due to fixes
for other sites it started getting much better. Others may already be fixed. Who
knows? No developer has the time to visit every site with reported problems to check.

Regards,
Luis


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


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