Re: Preparing for the freeze (was: Obsolete packages in main/oldlibs)
On 22 Sep 1998, Robin Stephenson wrote:
> What exactly are these problems and what on Earth makes you think
> people using Linux on Alpha don't do a lot of compiling? I'd have
> thought it was the *opposite*.
I agree with this point of yours. I expect and want people on the Alphas
to compile more (I think everyone should learn to code).
> Apropos of compilers, will the egcs bundled with slink be capable of
> compiling the kernel? It would be rather unpleasant to have to
> continue dealing with a RedHat-style gcc 2.7.3.2/egcs foo.bar compiler
> hodgepodge.
Yes, the current and all of our past egcs iterations can compile the
kernel. I'm working on getting the latest binutils packaged to solve ALOT
of the problems we've been having with jade and a few other packages.
> Well, I'm not so sure that'd be a bad thing. Maybe waiting a couple
> of months would allow things like libc6, egcs (perhaps even gnome...)
> to settle down a bit?
Not really. The x86'ers are churning out packages so fast that it's hard
to keep up as is. If we were gonna wait, we might as well wait forever
because they're never gonna stop or slow down.
> Good point. I got their 5.1 alpha distribution, and have had problems
> (understatement). The libjpeg debacle was the last straw -- I'm about
> to switch to Debian. To be honest, I don't care too much about a
> couple of rough edges -- I'll almost certainly be buying a slink
> CD-ROM in any case. It can't be worse than RedHat's 5.1. I just hate
> to think that I might have to go through downloading tens of megabytes
> of fixed stuff just because of a release schedule -- I like ID's
> attitude:
I agree. To make sure you don't have to d/l tons, PLEASE find some
testers for what we've done so far :-) Admittedly, our user base right
now is small, so the more testing we get from whom we have, the better the
dist will be.
> PS Would it be worth me bringing a UDB from home to a network to test
> installs, etc before the slink release? I'm prepared to put some
> time in if it's useful.
Absolutely. We desperately need documentation, so if you'd like to work on
that based on your install experiences, I know it would be GREATLY
appreciated.
C
Reply to: