Just a thought that popped into my head

Gerard Beekmans gerard at linuxfromscratch.org
Wed May 8 11:45:38 PDT 2002

On Wed, May 08, 2002 at 06:37:53PM +0100, Ian Molton wrote:
> > We never said we were not going todo a proper test.  You implied this.
> hm. gerard said he was going to use 3.1 on the strength of his current
> CVS gcc in one post. that does sound to me rather like any testing would
> be irrelevant and it will be used anyhow.
> How do you interpret that?

Come on now. How long have you been around on this list? If I'm not
mistaken since September 10th of last year. That means you've seen a bunch
of LFS releases already and you should very well know by now that when a
package is added to LFS-CVS, it never results in a next stable LFS release
for quite some time. This is done on purpose to give ourselves a couple of
weeks/months to test out LFS-CVS. LFS-CVS is only meant to be used by
adventurous people who don't mind testing the latest software and possibly
having unstable systems during that period. When the testers report no
problems we can then assume it's stable and a package freeze is put into
effect. We then hammer out the last minor bugs, typo's, grammar, download
locations, dependencies, descriptions and such for another week or so, then
release the next LFS book version.

This simply means that even if we don't run any specific GCC tests, it (the
new gcc) will be in LFS-CVS for a month at the very least (if not closer to
two) during which a lot of testing is done by a lot of people. So any
instability problems would be reported, anyways. Add some specific GCC
testing on top of it and I believe that will give us enough of a testing
window (one to three months from the looks of it now) I would say.

And, you have also seen that sometimes a package is added to CVS, after a
few weeks problems were found, it's removed from CVS and reverted to the
last known working version. The same will happen with GCC, it's just one of
the packages that gets the same treatment.
> Thats what Im trying to find out - I want to know exactly how he is
> planning on doing this upgrade as he currently sounds like he has 'pre
> appproved' gcc 3.1

I have, until otherwise proven. If GCC quality stays as it is, it will be
just fine for the book. If it improves, even better. If it worsens, I will
reconsider it.
> Why couldnt someone say that in the first place?

You didn't ask.

Gerard Beekmans

-*- If Linux doesn't have the solution, you have the wrong problem -*-
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message

More information about the lfs-dev mailing list