An user point of view on KDE and the glib issue

[Adam Treat]

On Tuesday 11 March 2003 10:00, M. Fioretti wrote:
> This (and the underlying assumption that the active KDE developers
> know what is always better for all users, and want to deliver it as an
> "all or nothing" bundle) is as Microsoftish as it can get. Quite
> scary, if you ask me

That is a crazy statement.  Consider:

"... and the underlying assumption that the active MacOS developers
know what is always better for all users, and want to deliver it as an
"all or nothing" bundle"

"... and the underlying assumption that the active X11 developers
know what is always better for all users, and want to deliver it as an
"all or nothing" bundle"

"... and the underlying assumption that the active Java developers
know what is always better for all users, and want to deliver it as an
"all or nothing" bundle"

I could go on.  The KDE projects decides (for better/worse) to use certain technologies to build
KDE applications.  Others can choose to use these, decide not to, fork the technologies, do
anything they wish.  What you are asking is the equivalent of telling the X11 developers that they
are Microsoftish because they have not provided an option of GDI in the core part of xlib so that
users can reuse Win32 code.

>I have read some rather worrying arguments in this thread that basically seem 
>to come down to "KDE applications have strong integration with the KDE 
>environment, to give KDE applications a competive advantage such integration 
>should be withheld from non-KDE applications". I see that slightly 
>differently.

KDE apps have strong integration because they are KDE apps.  That makes sense to me.  If non-KDE
apps wish to integrate into KDE then let them do so.  We have *open* interfaces and they are
welcome to use them or develop more suitable interfaces if needed.  Just because they do not wish
to make some effort to integrate with KDE does not mean that we have to bend over backwards,
redesign our infrastructure just to give them an interface that they wish for political reasons.  

You know, we could choose to use Java infrastructure and chuck kdelibs then we can have great
integration with *some* non-kde apps.  Of course KDE apps will suffer and this won't do anything
for Motif apps or GNOME apps or Wine apps....  Not very fair to them.  Common specs and reasonable
integration are great and I won't argue with them at all, but ripping apart our great
infrastructure just to accomodate a subset of non-KDE apps at the expense of KDE apps and all the
other non-KDE apps doesn't make sense to me.

>I see KDE based upon three pilars: 
>1) The development framework (Qt/kdelibs)
>2) The runtime environment: Desktop, WM, panel etc. (kdebase)
>3) Applications

<snip>

>It might be nice for KDE applications if 1) and 2) play 
>favourites towards KDE applications for political reasons, but 1) and 2) 
>don't become any better by doing so, on the contrary.

Changing our successful 1) and 2) which are part and parcel what KDE is all about just to
accomodate non-KDE apps that don't want to use 1) and 2) for political reasons will make 1) and 2)
worse IMO, and KDE apps will suffer for it.  If we are going to change 1) and 2) for easier
accomodation of non-KDE apps then I vote for Java apps ;)

Adam

__________________________________________________
Do you Yahoo!?
Yahoo! Web Hosting - establish your business online
http://webhosting.yahoo.com