Re: Modular X broken

[Jimmy Johansson <jimmy%Update.UU.SE@localhost>]

On Sun, May 19, 2013 at 05:16:49AM +0300, Jukka Ruohonen wrote:
> On Sat, May 18, 2013 at 04:42:01PM +0200, Thomas Klausner wrote:
> > On Sat, May 18, 2013 at 02:12:00PM +0300, Jukka Ruohonen wrote:
> > > The recent modular-X11 updates broke the Intel-driver at least for the
> > > 945G/M-chipsets that have worked continuously for four years. Was these
> > > changes even tested on NetBSD? I recall a reason why xf86-video-intel was
> > > deliberately kept behind, but can't remember the details.
> > 
> > They were compile-tested only.
> > 
> > We had to get to a newer version of X. The version of Mesa was chosen
> > so it still had the backwards compat we needed. There's no way a
> > single developer can test all X servers, so we need help here.
> 
> There was a "need" to get a new X even though it is known to cause
> regressions?  This is quite unfortunate since the Intel 945-965 is probably
> the most popular chipset among the laptops that actually currently work in
> NetBSD (the KMS etc. mess notwithstanding).  So prepare for new bug reports
> and more angry people with unsolveable problems with this supposedly needed,
> compile-tested update.
> 
> > Worst case, there's always the X server in base which is supposed to
> > work too, or the vesa driver.
> 
> True enough; I since reverted to the base-X too. The fundamental problem
> however remains.

The fundamental problem seems to be that you are running the wrong
branch of pkgsrc? It is like complaning when NetBSD-current breaks...

If you want stability then run a stable branch of pkgsrc.

Problem solved.

Regards,

Jimmy
-- 
If you don't shoot the bearers of bad news, people will keep bringing it to you.