Linked by Thom Holwerda on Tue 23rd Dec 2008 20:20 UTC, submitted by AdamW
Mandriva, Mandrake, Lycoris The first pre-release of Mandriva Linux 2009 Spring is now available. This alpha concentrates on updating to the major desktop components of the distribution, including KDE 4.2 Beta 2, GNOME 2.25.2, Xfce 4.6 Beta 2, X.org server 1.5, and kernel 2.6.28 rc8. It is also the first distribution to introduce the major new Tcl/Tk release, 8.6. The alpha is available only in the DVD Free edition with a traditional installer and no proprietary applications; future pre-releases will add the live CD One edition with proprietary drivers. Please help test this first pre-release and report bugs to Mandriva.
Thread beginning with comment 341604
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[9]: KDE 3?
by lemur2 on Sat 27th Dec 2008 04:07 UTC in reply to "RE[8]: KDE 3?"
lemur2
Member since:
2007-02-17

"It is indeed a problem because many people, yourself included apparently, have apparently totally got the wrong end of the stick here.

No. I have the right end of the stick. That's why you have to type, and type, and type so many paragraphs to "prove" me wrong. And I can just point to the obvious truth.

KDE4 jumped on the wrong boat with this particular design decision... despite any spin that you or Aseigo care to put on it after the fact. It was, perhaps, based upon a reasonable guess about the future of commonly available, hardware accelerated Xrender support at the time the decision was made. But it was based upon a *wrong* guess.
"

How was the KDE team supposed to know that Nvidia would introduce a severe performance bug into their binary driver (for Linux only) and then refuse to fix it (or even acknowledge it) for over two years?

You seem to think that this functionality wasn't available when the KDE4 team was making design decisions. You seem to assume that the KDE4 team bet the house on something that was supposed to appear in the future. That is an entirely incorrect assumption on your part. This bug appeared in the nvidia binary driver for Linux only AFTER KDE4 was designed, and was approaching a first release.

http://en.wikipedia.org/wiki/Kde_4#KDE_4.0
On May 11, 2007, KDE 4.0 Alpha 1 was released marking the end of the addition of large features to the KDE base libraries and shifting the focus onto integrating the new technologies into applications and the basic desktop.


That dates the design decisions for KDE 4 at about the late 2005 timeframe. The Nvidia driver bug has been present, for 8000 and 9000 series cards only, for about two years ... since late 2006 timeframe. At first it wasn't that severe, and it seemed to affect only the rendering of anti-aliased fonts.

The bug doesn't affect nvidia cards earlier than the 8000 series, for example. They work fine with KDE4, and perform far better than the 8000 or 9000 series cards that are affected by the bug.

http://en.wikipedia.org/wiki/Nvidia#Market_history
Nvidia released the 8-series chip towards the end of 2006


So, the KDE4 design wasn't at all "based upon a reasonable guess about the future of commonly available, hardware accelerated Xrender support at the time the decision was made". Rather, it was based on what actually worked at the time. Xrender itself has been around since 2000, as I previously pointed out, and at the time KDE4 was designed it worked on all accelerated graphics hardware that was commonly available, representing almost all of the hardware still running.

I think you still have entirely the wrong player down as the one who gummed up the process here.

Edited 2008-12-27 04:25 UTC

Reply Parent Bookmark Score: 2

RE[10]: KDE 3?
by sbergman27 on Sat 27th Dec 2008 04:30 in reply to "RE[9]: KDE 3?"
sbergman27 Member since:
2005-07-24

How was the KDE team supposed to know that Nvidia would...

Lemur2, it's not about what they could or could not have known. It is about what *is*. None of the paragraphs you have written change today's reality.

You keep implying that I claim that the KDE guys made a stupid decision during the KDE4 design process. And maybe they did make some. But I am not claiming that this is one of them. This was simply a *wrong* decision that has had, and continues to have substantial negative consequences for them. It doesn't matter whether they could have predicted it. It's like in "What Ever Happened to Baby Jane":

Blanche: If only I weren't in this wheelchair!

Jane: Butcha are Blanche! Ya are in that wheelchair!

Reply Parent Bookmark Score: 2

RE[11]: KDE 3?
by lemur2 on Sat 27th Dec 2008 04:42 in reply to "RE[10]: KDE 3?"
lemur2 Member since:
2007-02-17

" How was the KDE team supposed to know that Nvidia would...

Lemur2, it's not about what they could or could not have known. It is about what *is*. None of the paragraphs you have written change today's reality.
"

Correct. The only thing that will change the reality is when Nvidia finally release the version of their driver that fixes nvidia's bug.

You keep implying that I claim that the KDE guys made a stupid decision during the KDE4 design process.


I'm not implying that all. I'm not even infering that you thought it.

Here, I'll quote you directly:
"It was, perhaps, based upon a reasonable guess about the future of commonly available, hardware accelerated Xrender support at the time the decision was made. But it was based upon a *wrong* guess."


Sorry, but that was dead wrong. Complete bollocks. Hardware accelerated Xrender support required by KDE4 was present at the time the decision was made, and the decision to use it was not at all based on a guess of any kind.

And maybe they did make some.


Name one.

Edited 2008-12-27 04:43 UTC

Reply Parent Bookmark Score: 2