Re: Build system race conditions

[Martin Husemann <martin%duskware.de@localhost>]

On Fri, Sep 12, 2014 at 01:49:10PM +0200, Rhialto wrote:
> Maybe the parallelism got a bit out of control there with the "cannot
> fork" errors?

I have seen that, but unlimiting the build user always worked around it
for me. There is also PR 49182.

I've been told failure in netbsd-7 and -current should be less likely
than in older branches (due to reworked library dependency handling).
We'll see once -current and -7 are buildable again.

And at least a few recent failures are bogus: the build succeeded but the
scripts transfering the results failed, and the status page noted this as
failure.

Martin