The arch:all part of theano FTBFS with sphinx 1.7.5-1
(#901662). This has now been fixed in sphinx.
dw theano_1.0.2+dfsg-1 . all . -m 'sphinx-common (>= 1.7.5-2)'
Thanks.
scilab:all FTBFS on x86-bm-01 but succeeds on x86-csail-02 (#926180):
https://buildd.debian.org/status/logs.php?pkg=scilab
The error message looks like this might be a baseline violation.
Checking when it first appeared in a few packages' build logs suggests
x86-bm-01 is newer than my laptop (
The buildd status pages for statsmodels, pandas and mender-client say
ppc64 (and in the last case powerpc) tried to build them, but do not
give the usual link to the build log.
All of these were attempted in the last 24 hours. The failures were on
kapitsa2, the only success (statsmodels) does
The pandas build was retried and succeeded, but again with no log.
I mostly want to know whether the failure was an instance of
https://bugs.debian.org/940679 (random test crashes - as this is an
intermittent bug, it makes sense that the retry succeeded), and if so,
which test crashed.
The build logs have now appeared. (The pandas failure was #940679.)
numexpr 2.8.5-2 didn't actually fix enough of #1049326 to let pandas
build. numexpr 2.8.5-3 should do so, but hasn't been built yet.
(DMs can't use the self-service method.)
dw pandas_1.5.3+dfsg-5 . ANY all . -m 'python3-numexpr (>= 2.8.5-3)'
Please give back pandas_1.5.3+dfsg-11 (unstable) on i386: that looks
like a rare random failure.
(DMs can't use the self-service method.)
