Hi lists, In case it slipped past people's notice, I just wanted to mention that Cernlib's PAW data analysis program <URL:http://packages.debian.org/paw> should work OK on Debian/AMD64 as of version 2005.05.09.dfsg-3. I've applied Harald Vogt's AMD64 patches from <URL:http://www-zeuthen.desy.de/linear_collider/cernlib_2005.html> and forced PAW to be statically linked on 64-bit arches.
PAW on AMD64 now runs almost all of the PAW tests and examples (in the paw-demo package) successfully. The only one it fails at is example 29, where COMIS for some reason is unable to execute the HPLNUL subroutine - I don't yet understand why. Rigorous tests by PAW mavens with AMD64s are welcome! Sorry, it still doesn't seem to work very well on Itanium or Alpha, due to the fact that those systems store data and text above the 32-bit boundary in their virtual address spaces, defeating the workarounds for 64-bit uncleanliness used by Professor Vogt. Please follow up to debian-science if you wish. On another note, I've found an interesting set of applications called "OpenScientist" on the web: <URL:http://openscientist.lal.in2p3.fr/>. It is GPL and apparently aims to be a competitor to ROOT (about which the main OpenScientist author seems to have a bit of a chip on his shoulder). The relevant part to my email is that OpenScientist contains a re-implementation of PAW called "Open PAW": <URL:http://openscientist.lal.in2p3.fr/v15r0/html/openpaw.html> Since a lot of it was re-written from scratch, it may very well be that this could work as a PAW substitute on 64-bit arches. Seems to be in the early stages yet; anyone have any experience with it? Happy Holidays to all, -- Kevin B. McCarty <[EMAIL PROTECTED]> Physics Department WWW: http://www.princeton.edu/~kmccarty/ Princeton University GPG: public key ID 4F83C751 Princeton, NJ 08544 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]