3_3_0_BETA_! Build problems [RESOLVED]

[James Richard Tyrer]

Kurt Pfeifle wrote:
> James Richard Tyrer wrote:
> 
>> I finally resolved my build problems after making at least one serious 
>> mistake myself :-( while trying to do it (I somehow checked out the wrong
>> copy of the kdebase module when I was rebuilding if form scratch).  But, I
>> also wasted a lot of time trying to build code that wouldn't build.
> 
> 
> See? We all experience this from time to time. But the only one to *always*
> make a lot of noise about it and *first* and *formost* attacking fellow KDE
> developers is you, James Richard.

I don't see this as an "attack".  I apologize for the fact that I screwed up and
didn't manage to find the problem before  it was released.  But, there should be
others testing it before the tarballs were released.  Obviously nobody tested
the build.

>> There ARE errors in: "kdebase/kwin"; now documented:
>> 
>> http://www.kde.org/info/3.3beta1.php
>> 
>> I see this as a MAJOR quality failure.
> 
> That may or may not be the case. So what?

The project has had several such problems over the last few months.  The "so
what" is that we need to do something to prevent such problems.

> It can be fixed. That's why we have "public" betas....
> 
>> No testing at all before it was released.
> 
> It is not *released*. It is a beta1. Do you know what that means?

This is semantics.  Tarballs were posted on the FTP servers.  This is what I
mean by it was released.  Before that is done at least one person should confirm
that they will build.

>> But, I wonder why none of the developers noticed this.
> 
> I wonder why you dont realize just how *arrogant* your attitude does come
> over.

I believe that you have chosen the wrong word.  Yes, I an overbearing
perfectionist (in short, I am Germanic).  But I do not meet the dictionary
definition of arrogance:

http://www.m-w.com/cgi-bin/dictionary?book=Dictionary&va=arrogance&x=9&y=14

>> Logically, the only conclusion is that nobody ever updated their CVS with
>> the tag: KDE_3_3_0_BETA_1 and then did a build.
> 
> 
> I am very glad that *you* did it now for all of us. Thank you very much! We
> are very happy about that.
> 
> What we still dont like is your arrogant way to talk down to us.
> 
> You are an engineer? Well, I am not. You are a software developer? Well, I am
> not.

Yes, I have the attitude of an engineer.  The first step in problem solving.  I 
have done this, I have identified the problem.  Or, at least I have started. 
Perhaps others have other ides about the problem.  We should discuss this if you 
think that there are other explanations.

> You have a short memory? Well, I have not. Because I still remember very well
> how you promised some time ago to now start tackling the task to make
> KDEPrint fully support LPRng, and at the same time to better architect and
> software-design the complete filtering, print and PPD-system.

I ran into the fact that the GNUlpr is dead and I contacted the LPRng developer 
and he advised me that he would be making major changes to the program.  This 
leaves me up a creek.  I don't like LPRng, but I will have to get back to him 
and see what he is doing.

> I am still waiting to see the first working code on that field. Until that
> happens, I only feel sorry for you if I read your rants.

Actually, I was going to start with debugging the existing code which didn't 
seem to work with LPR.  It was suggested that since LPR was dead that I 
shouldn't do that.  Probably correct, but GNUlpr which is what I said that I 
would work on is now a dead issue except that I still use it and it still 
doesn't work with KDEPrint.

So, I volunteered to do a job that I was told didn't need doing.  I felt 
somewhat stupid. :-\  When the next major release of LPRng is released, I will 
install it and if I find it usable, I am willing to help with the KDE interface 
for it.

I still like the idea of making out own print spooler based on LPR.  Currently, 
the configuration system on LPRng is a mess.  They seemed to find my ideas of 
how to improve it as having some merit.

>> I tried to do this, but I had some CVS problem (I am not sure of what 
>> exactly, only that it is now fixed) so I was not able to do this. However,
>> to conclude that none of the developers did this indicates that we have a
>> corporate culture that must be changed.
> 
> Once you change your attitude to a more positive, constructive, 
> "lead-by-example" (instead of a "talk-from-above") one, you would have done
> the first step to a better *community* culture from your side. If you want to
> change the KDE project to become a cold *corporate* culture: go away. Try
> this elsewhere. Not here.

You misunderstand what "corporate culture" means.  We already have one.  Every 
organization that has existed for a while has one.

Quite the opposite, the changes I see as needed are not "cold" I would like to 
see a warmer corporate culture.  Don't you see how "cold" our current corporate 
culture is.  Specifically, I believe in the ideas of the Quality Guru Tom 
Peters.  I believe that some of his ideas would greatly improve the project.

But, even if I ranted too much --

However, I must ask, did you build the KDE-3.3.0BETA1 from source.  Perhaps I am 
just venting my frustration of working on it for hours only to finally find out 
that there were errors in it.  I expect this with HEAD but not with a release, 
even if it is just a BETA.

-- I have identified the problem, which is that stuff is kicked out the door 
without adequate quality testing.  This project has become too large not to have 
some organized quality control.  People don't like the message so now you want 
to kill the messenger.

--
JRT