CUPS, printcodes, stale UTMP entries, and terminal hangs

[Fairlight]

With neither thought nor caution, Bill Vermillion blurted:
> 
> I wish I'd save the e-newletter I got a couple of weeks ago with
> some of the testing done on the 2.6 based systems.  It looks to be
> a tremendous winner for web-servers, and the different test
> results I've seen run from 30 to 50% inprovments.
> 
> It really shine in the multi-cpu environment.

And have new drivers, etc.,etc. 


> Since 2.6 was released in December of 2003 you are saying that
> a kernel should be in use for at least 1.5 years before you trust
> it?  

No.  It could happen far faster--used to, in fact.  But historically,
"stable" trees haven't been truly stable until around .18 or so.  Usually
that's about magic time.  You can point to 2.6.10 and all its great
features.  They said the same thing about 2.4.*--until 2.4.15 started
wiping whole filesystems on a proper shutdown--or sometimes a boot.  Then
people sang a different tune.  But it was stable by about .18 as well.  2.2
was starting to go into maintenance mode with only security patches and
new feature backports at about .18.  It's almost always been this way since
I can remember.  The time is immaterial.  They're just developing far
slower these days.  It's about maturity and stability.  And trust me, those
poor sods that got bitten by 2.4.15 believed 2.4 was stable until then.
My best friend got bitten by that one, in fact--and had been touting that
tree until that time.

> > > box without incident.  For those of a technical bent, it has to do with 
> > > the 2.6 kernel using pseudo-tty numbers in excess of 999.  If you'd 
> 
> > *cough*  The prosecution rests.
> 
> I'm confused by that last statement.  What exactly are you tyring
> to say?

I'm saying that he's saying that (while they appear to have modified
openssh to get around it) the problem is being reported as related
specifically to the 2.6 kernel.  I've been saying not to use it until .18
since before it came out.  I've been into linux long enough to know how
this stuff usually goes.

I'm saying his bug report is a direct indictment of something that's Not
Quite Right with the 2.6 tree.  And it just helps prove my case that people
should wait until it's right, not just close.

mark->
-- 
Bring the web-enabling power of OneGate to -your- filePro applications today!

Try the live filePro-based, OneGate-enabled demo at the following URL:
               http://www2.onnik.com/~fairlite/flfssindex.html