KDevelop 4 Beta 5 Artikel

[Patrick Spendrin]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Milian Wolff schrieb:
> To the KDevelop ML:
> 
> This is a reader's letter to the German magazin "Linux Magazin" that printed 
> an article about KDevelop 4 Beta 5 in their 12/09 edition.
> 
> ~~~
> 
And here the translation (not everything is my personal opinion; please
excuse all errors):

Dear Linux-Magazin Team!
> 
in the beginning: The content of this letter is my personal opinion, and
must not be the one of the whole KDevelop developer team.
> 
I find it rather nice that you write an article about our project
KDevelop 4. What your Author is writing, is mostly correct. But I must
say that I am a little bit disappointed that obviously no developer was
asked beforehand. Some aspects of the article show that the author has
no background knowledge. Doesn't it belong to journalism, to make
inquiries? And I would state that this would work with the developers
best, additionally many of them also speak German.

Here some corrections, clarifications and hints:

- - I find it negligent from the Author to bespeak the language plugins
for python, ruby and C# which are unmaintained. These plugins are really
unstable and most cannot even be compiled. This is caused only by lack
of time of the developers. If one of your readers is interested to
further develop the plugins, we are pleased to help him.

- - If "KDevelop 4 reproducable [crashes] under a testsystem running Open
Suse 11.1" one should report a bug under http://bugs.kde.org. And/Or one
searches for help on IRC (#kdevelop on Freenode).

The author seemingly did nothing like that. I don't suggest that Beta 5
(and yesterday Beta6 was released) is stable - no. It is a Beta. And
within our bug tracking system enough existing errors can be found. But
I still find it impudent to report about a bug in beta software and not
telling the developers ... How should we fix that error then?

- - The author speaks about the split between KDevplatform and kdevelop,
but he forgets to mention an important detail: This split will support
project like Quanta+ 4, when it is developed any further. Every change
on the platform also benefits Quanta. One can also think of porting
other project like Kile in the distant future. The advantages would be
great. And btw. it is not by accident that the "duo is waiting
peacefully for the KDE-server", this will stay like that forever. In the
end kdevelop is a KDE project...

- - the point that many of the features from KDevelop 3 are missing,
doesn't mean that we decided against needing these features. Rather the
opposite. Plugins for Qt Designer, AutoTools, QMake etc. do exist, even
if unmainted and instable. The problem is again that we are missing
developers. This could have also been mentioned by the author.... In the
same time one shouldn't expect that we try to port blindly every feature
from kdevelop 3. KDevelop 4 is as correctly stated a new beginning. And
we use this to throw away the less important stuff.

- - "A class diagram is still missing." There was a GSOC project with good
outcome, see also:
http://liveblue.wordpress.com/2009/08/21/gsoc-wrap-up-static-code-visualization-in-kdevelop/

- - about distributions: I find it surprising, that the author wonders
about that the current beta5 requires a recent KDE 4 as a base and the
repositories of the distributions only contain an old KDevelop 4. As he
writes himself, it is a _Beta_ version?! You can't expect from the
distributions that they package software that is not marked as stable?

That the folks of ubuntu decided to drop KDevelop 3 and package an old
Beta - this is a problem of Ubuntu and not of KDevelop.

And that we need a recent KDE 4 as base is also pretty normal: We try to
release the first stable version of KDevelop 4.0 together with KDE
4.4.0. And some of our features only can be achieved with the new
versions of Qt and KDE. Why should we support old KDE 4.X versions
without having a stable release?


In the end, I can only say that I would prefer it in the future, if an
author first inquires about the project before he writes an article
about it.

You're also not supposed to release blind positive propaganda about
KDevelop - in contrary. I just want that you give your readers a
realistic description of the project, which includes that you describe
problems and warn about instability.

But in the same time I think it is absolutely needed that the author
tries to connect with the developers of the project and goes through the
article about the project. This way such errors like the unusable
language plugins could be found early. Now you raise your readers hope
which cannot be fulfilled by us. This does lead to a bad light on our
project, which is not needed.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (MingW32)

iEYEARECAAYFAkry5FoACgkQi49rfdk/G3b6eACcCRDA/rYcc87YiSAw3inA5YQi
/bgAoIllgdrt5oebkhr99hRxCR4E7fja
=sE46
-----END PGP SIGNATURE-----