On 10 June 2011 13:59, Bill Hart <goodwillh...@googlemail.com> wrote: > On 10 June 2011 10:36, Burcin Erocal <bur...@erocal.org> wrote: >> Hi, >> >> On Fri, 10 Jun 2011 09:27:47 +0100 >> Martin Albrecht <martinralbre...@googlemail.com> wrote: >> >>> I'm wondering whether we should do a Sage-FLINT days thingy/coding >>> sprint (either somewhere in the big blue room or online) to make use >>> of all the cool new features in FLINT 2.x? >>> >>> Or should we wait until everything from 1.5 are in 2.x? >> >> We could do it gradually if it was possible to install the two side by >> side. Since the API changed drastically and 2.x will not provide all >> the bells and whistles of 1.x in the near future (by fast-paced full >> of impatience Sage measures), this makes perfect sense. >> >> Could anybody modify the build system so we can use both of these >> together? >> > > Do you mean the build system of flint or of Sage? If the former then I > think the only thing I can think of is to rename libflint.so to > libflint2.so. This is easy to do of course. Simply change the relevant > line in Makefile.in and reconfigure. > > I believe the only name clashes should be between fmpz and fmpz_poly > between the two flints. One library order should give you flint 1.5 > fmpz access, the other library order should give the other fmpz. > > Note that in order to get all the features of flint 1.5 into flint2 we > need the new FFT which I started writing nearly 1.5 years ago. This > will be going into MPIR, the next release for which is scheduled in > early September. How is Sage doing with keeping up with the MPIR > releases? I haven't been keeping track lately. flint2 needs a pretty > recent release of MPIR and similarly for mpfr. > > I assume Sage did not upgrade to flint 1.6. This is a good thing as it > will be a very long time before we are able to cover factorisation of > polynomials over the integers in flint2. It'll be worth waiting for > though. >
Of course all the new fmpq_poly and fmpz_poly_mat stuff relies on the new fmpz_poly stuff and fmpz_mat on fmpz and fmpq on fmpz, and fmpq_mat on fmpq, and arith on fmpz etc. So you may have disable some (most) modules in flint2 until you are using the new fmpz and fmpz_poly. This can be done by removing the names of the modules that you want to disable from the last line of Makefile.in. At the moment fmpz_poly relies only on fmpz, fmpz_vec, nmod_vec, nmod_poly and ulong_extras. The test code may also rely on long_extras as well. So this is probably a good set of initial modules to enable in flint2, with the only conflicts with flint1.5 being in the fmpz and fmpz_poly "namespace". Bill. -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
