inconcistency in build failures between buildds
Dear all buildd maintainers, especially those of s390, powerpc and armel,
please kindly inspect
https://buildd.debian.org/status/package.php?p=ball
which shows successes or failures for an implicit conversion from the
constant integer -1 to char.
I am in contact with upstream about it, which may indeed help to clarify
if a -1 or a 255 is expected to be stored. That aside, I have already
prepared another version with -fpermissive in its CFLAGS. But since this
is fine for some platforms (i.e. for my local unstable amd64 and the
i386 buildd) and not so fine for others, I thought you might be
interested to compare compiler versions between your buildds. Mine here
is 4.6.1.
With many thanks and kind regards,
Steffen
Reply to: