[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: alpha buildd

On Fri 20 Apr 2001, Ivan E. Moore II wrote:

> > By my daily list of out of date alpha packages, it looks like it's
> > not really coping that well... Where are the build logs, so that
> > manual fixing of stuff can be done without getting in the way of
> > stuff that *does* get built correctly by the build daemon?
> hmmm...i think it's coping quite well.  So far all the packages that do
> fail to build have been packing problems except those listed on the web
> site (http://build.tdyc.com/)  which is still incomplete...and of the 


> 225 currently unbuilt packages, some are non-US which aren't built by the

Hmm, my list (available via http://auric.debian.org/~paul/alpha/diff/ )
shows about 280 out of date packages in main. Of course, that's
binary packages, not source packages (which is estimated at 116,
including packages already uploaded but not yet installed).
There are also 85 source packages not yet built for alpha at all in
main, but that's probably for a large part down to arch non-alpha.

My concern was raised by the fact that my list has steadily grown in
length from about 490 lines a month ago, to over 800 the last couple
of days.

> which the buildd shouldn't be doing anyways...others have problems do to
> them bing c++ w/ optimization...

I have a nifty little wrapper for c++ which first exec'ed the real
binary as-is, and if that exited with exit status != 0, exec'ed it again
with -O0. Works pretty well :-)

> and then of course the buildd just got kicked off for todays run right after
> I got that 225 number...  a week ago we were at 50 packages needing built...
> so with all the compiler issues we have been having I think that the buildd
> is coping quite well.

Hmm, OK... It's just that I saw my list getting longer and longer,
whereas it was getting shorter and shorter when we were still building
by hand. I'll root through it and at least filter out the non-alpha
stuff that's new since I last did that.

Paul S.

Reply to: