-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I wrote the current endianness test Yhc uses for telling the ffi library it compiles. It uses C (compiled from Python/Scons), and it's an endianness test, so it's not portable. Seeing as there has been a bit of interest in a Haskell-based build system, whether or not that happens, I wrote a test in Haskell based around (Ptr WordX), and it works correctly on my big-endian machine at least (and its comments reference a modern source on endianness, namely, wikipedia :). It didn't take long. If you're interested: darcs get http://isaac.cedarswampstudios.org/2007/Endian/ (or don't even bother with darcs, just web-browse) Isaac P.S. Following the tradition of releasing little libraries, I feel tempted to do that, but... ;) -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFF7epRHgcxvIWYTTURAvorAJ4ns8TXlqt0pYPrJtzcWpZbWp9CzQCfW+Vf uw4xepQ/58n5Ly/L4xyZpns= =ysGx -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Isaac Dupree wrote:
I wrote the current endianness test Yhc uses for telling the ffi library it compiles. It uses C (compiled from Python/Scons), and it's an endianness test, so it's not portable. Seeing as there has been a bit of interest in a Haskell-based build system, whether or not that happens, I wrote a test in Haskell
What I'm forgetting, of course, is where the configure tests (and build system? maybe not for cross-compiling, if that is possible) are supposed to be able to run. Possibly places where there isn't a haskell compiler yet, and doesn't need to be one (only yhi-with-ffi being desired). Isaac -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFF7pwGHgcxvIWYTTURAmcJAJ9+H2rTcLCI5+FWeRs+lNBV7T4w7wCfWd9N 9Wm55ayFuEevXRuVrOXts1I= =wJye -----END PGP SIGNATURE-----
participants (1)
-
Isaac Dupree