Op wo, 18-10-2006 te 09:53 +0200, schreef Stefan Potyra: > Hi Arjan, > > Am Freitag 13 Oktober 2006 00:35 schrieb Arjan Oosting: > > Op do, 12-10-2006 te 15:53 +0200, schreef Stefan Potyra: > > > > > The problem now is, that it FTBFS on sparc, see [1]. Not quite sure if > > > this is an ubuntu only bug though. > > > > No, it does not look like a ubuntu only bug. It seems the location of > > byteorder.h is wrong in the source. Could you try a new version from my > > personal repository [1]? I can not test it myself as I have no acces to > > a sparc machine. > > sorry for my laziness and late response... > > Since I don't have access to a sparc machine as well, I wanted to look around > if I s.o. would test-build this version for me. However in the meantime, > StevenK uploaded a different fix to ubuntu which simply strips out the > include-directive. It didn't FTBFS any longer on sparc, see [2].
But does it work on sparc? Hat is sensitive to endianness of an architecture. So while it might build it could still produce incorrect traces now. See for instance this thread on the upstream mailing list: http://www.haskell.org/pipermail/hat/2006-October/000353.html I am not sure why the header is included and at the moment have no time to find that out, so maybe you can take this upstream and mail the upstream mailing list [1]. They should be able to tell whether ripping out the include is ok. Greetings Arjan [1] http://haskell.org/mailman/listinfo/hat
signature.asc
Description: Dit berichtdeel is digitaal ondertekend
