[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: pkgsrc HEAD selective NetBSD 5.99.49/x86_64 2011-10-07 03:35
On Sat, Oct 08, 2011 at 01:37:43AM +0000, David Holland wrote:
> pkgsrc bulk build report
So for the record, this is a limited build done only on (1) packages
that are known to fail, and (2) packages that have been substantively
changed since they last built ok. The point of this is to get rapid
turnaround on new commits; it's intended to complement, not replace,
full builds. In particular it won't necessarily catch cases where
package Y stops building because package X gets updated.
For some reason I don't understand a number of packages that fail with
compile errors in Joerg's build work in mine; mine is done on a
somewhat older vintage of -current, and even though they're both
supposedly gcc 4.1, I think the warning flags in packages using BSD
makefiles aren't quite the same. Also, my build isn't sandboxed as
carefully, so some stuff may be succeeding by finding undeclared deps
in the host /usr/pkg or /usr/local. Anyway, direct comparison to
Joerg's builds isn't likely to be helpful.
I'm also excluding several known hopeless cases:
as well as the following packages that are known to need major
attention (someone please take these up...):
plus (for now at least) some stuff that fails on quota2 in -current,
plus as many of the failing python31 packages as I can shut up.
Having unfixable noise in the reports isn't useful.
The goal is to get down to zero or nearly zero failures, and then,
hopefully, stay there. Some of the currently failing packages have
been broken for ages, and this isn't much use to anyone.
David A. Holland
Main Index |
Thread Index |