On 2010-Aug-16 14:32:14 -0700, Bill Hart <goodwillh...@googlemail.com> wrote: >I'm also really wondering why cephes is in Sage. Unless there are two >packages with the same name, it was last updated circa 1994 and uses >non-portable long doubles. We've booted packages from Sage for much >less felonious offences.
cephes is used to provide C99 math functions for systems that do not provide complete C99 math functionality natively. Currently, this is cygwin and FreeBSD, though other proprietary Unix systems may also need it. Whilst it was _written_ in the early 90's, it has been updated since then, with patches dated up to around 2005. In any case, age of code is not inherently an indication of bugginess. Assuming code has no known bugs and achieves its stated functionality, there is no need to touch it (though I don't know this is true of cephes). As for "long doubles" - whilst they may be non-portable, AFAIK, they are part of C99. They are also extensively used in Sage. If you want to get rid of long doubles completely, feel free. It's being unreasonable to have long doubles in Sage and then remove code that is needed to support them. On 2010-Aug-16 16:24:24 -0700, William Stein <wst...@gmail.com> wrote: >I would rather have a windows port than not having a Windows port. >Problems may be created as a result. But since Cephes is only used >on Windows, the problems will be less than the current problem which >is that Sage doesn't work at all on Windows. See http://trac.sagemath.org/sage_trac/ticket/9543 Is there no longer any interest in a FreeBSD port? -- Peter Jeremy
pgpPtLrNSEAFi.pgp
Description: PGP signature
