Todo

New instructions:

1. please see New-TODOs
2. Coordinate work with other active people via the developers mailing list

Ideas for Singular on top of libpolys

Fix the compilation of libkernel.a:

1. svn rebase! (Mohamed + Hans + Oleksandr)
2. make script (just make in BASE/kernel) (Hans)
3. mod2.h is being crafted by hand + includes aux...h (Hans)
4. reintroduce currRing,rSetCurrRing, currQuotient (in kernel/polys.h)
5. deal with structs.h! (Hans)
6. install coeffs/mpr_complex.h and factory/templates/ftmpl_list.h
7. what about longrat.h and modulop.h (i placed dummy including coeffs/number for now)
8. deal with other dummy's (re-introduced stubs instead of correcting lots of code)
9. Find out all possible "HAVE_*" (which were not yet set) and define them in mod2.h for the time being! (Hans?)
10. GMP: ~~pass correct GMP_CFLAGS/GMP_LIBS: GMPrat.h, misc.cc, si_gmp.h~ + sort out kernel/si_gmp.h: removed in favour of coeffs/si_gmp.h (or factory/cf_gmp.h, which are private at the moment...)?! put a copy to kernel/...? (Oleksandr?)
11. documentation of some mixed stuff Porting instructions?!
12. p[FL]Deg -> p_[FL]Deg (Hans?)
13. pInline1.h and pInline2.h were forgotten in ROOT!/polys (Hans?)
14. rename ./for_Hans_with_love.sh into autogen.sh (or use the standard one if it supports multiple configure entries...?)
15. linking of libkernel*.a + make check in kernel (Oleksandr)
16. coeffs/longrat.h should NOT to be public! (Hans says) => rewrite with the use of number and n_*
17. TODO: a lot of code in p_polys.cc was commented out by Frank while switching to new alg./trans. extension fields :(
18. TODO: Singular interpretor has to support new alg./trans. extension fields ...!

List of files to be fixed!

Write your name next to a file if you are working on it. When you are done, use ~~ to put a strike through the file name.

See Porting instructions for instructions on what needs to be done for each file.

• digitech.cc
• eigenval.cc
• f5c.cc, f5data.cc, f5gb.cc, f5lists.cc, F5cLists.cc, F4.cc?, F5cData.cc (Christian, they are empty now)
• fast_maps.cc (Hans)
• fast_mult.cc (Martin)
• febase.cc, febase.h (Hans)
• feread.c (Hans)
• fegetopt.* (Hans)
• fglmvec.cc, fglmzero.cc, fglmcomb.cc, fglmgauss.cc (Hans)
• fglmhom.cc (empty, needs some HAVE_?)
• gfan.cc (empty, needs HAVE_FANS!)
• GMPrat.cc
• gr_kstd2.cc (Hans)
• hdegree.cc
• hilb.cc
• hutil.cc (Hans)
• ideals.cc (Hans, Oleksandr)
• int64vec.cc
• khstd.cc
• kstd1.cc, kstd2.cc, kstdfac.cc
• kutil.cc (mostly Martin + Non-Commutative: Oleksandr)
• linearAlgebra.cc (Martin: mixed with/without const ring R, interface part (with lists) went to the /Singular directory)
• misc.cc
• mmstd.c (Hans)
• multicnt.cc
• npolygon.cc (Martin)
• ratgring.cc? (empty, Oleksandr/needs HAVE_RATGRING)
• ringgb.cc (Martin)
• semic.cc (Martin, moved to /Singular)
• shiftgb.cc
• spectrum.cc (Martin, moved to /Singular)
• splist.cc (Martin, moved to /Singular)
• syz.cc, syz0.cc, syz1.cc, syz2.cc, syz3.cc (Oleksandr)
• tgb.cc, tgbgauss.cc (Hans)
• tgb_obsolete.cc (removed)
• timer.cc (Hans!/Burcin, works with HAVE_SYS_TIME_H, TIME_WITH_SYS_TIME, HAVE_GETTIMEOFDAY on 64-bit current gentoo)
• units.cc (Hans)
• walkMain.cc, walkProc.cc, walkSupport.cc
• kpolys.cc (Burcin)
• kspoly.cc (Burcin)
• mmalloc.cc
• kInline.cc -> kInLine.h

Ideas for improving libpolys:

• we have a lot of easy to-get-get-rid-of-warnings (in compile time): eliminate them.
• create a single libpolys.a (and libpolys.so!?) out of all our libraries (libpolys/misc.a .... etc)
• try to make use of clang and Sun-Studio-compiler for libpolys: don't work as of now :(
• find some tool to check test coverage

##List of files to be fixed in Singular directory

• attrib.cc (no changes necessary)
• bbcone.cc (needs HAVE_FAN)
• bbfan.cc (needs HAVE_FAN)
• bigintm.cc (empty)
• blackbox.cc (no changes necessary)
• calcSVD.cc (needs HAVE_SVD) (Hans, O.)
• claptmpl.cc (no changes necessary)
• cntrlc.cc (no changes necessary)
• denom_list.cc (no changes necessary)
• eigenval_ip.cc (Burcin)
• emacs.cc (removed by Hans)
• extra.cc (O.)
• fehelp.cc (Burcin)
• feOpt.cc (Martin, Burcin)
• fglm.cc (Burcin)
• gentable2.cc (no changes necessary)
• gentable.cc (no changes necessary)
• gms.cc (Oleksandr)
• grammar.cc (Hans)
• interpolation.cc (Burcin, uses coeffs/longrat.h!!!)
• iparith.cc (Burcin, Hans, O.)
• ipassign.cc (Burcin, Hans, O.)
• ipconv.cc (Burcin)
• ipid.cc (Hans?)
• iplib.cc (Hans?)
• ipprint.cc (Hans?)
• ipshell.cc (Burcin, Hand, O.)
• janet.cc (Martin)
• libparse.cc (Hans?)
• linearAlgebra_ip.cc (Burcin)
• lists.cc (no changes necessary)
• maps_ip.cc (Burcin, Hans, O.)
• Minor.cc, MinorInterface.cc, MinorProcessor.cc (Burcin)
• minpoly.cc (no changes necessary)
• misc_ip.cc (Oleksandr)
• mpsr_Error.cc (needs HAVE_MPSR)
• mpsr_Get.cc (needs HAVE_MPSR)
• mpsr_GetMisc.cc (needs HAVE_MPSR)
• mpsr_GetPoly.cc (needs HAVE_MPSR)
• mpsr_Put.cc (needs HAVE_MPSR)
• mpsr_PutPoly.cc (needs HAVE_MPSR)
• mpsr_sl.cc (needs HAVE_MPSR)
• mpsr_Timer.cc (no changes necessary)
• mpsr_Tok.cc (needs HAVE_MPSR)
• ndbm.cc (empty!?)
• newstruct.cc (Hans)
• pcv.cc (Burcin)
• pipeLink.cc (Hans?)
• pyobject.cc (Oleksandr, makefile needs python-config for this!)
• pyobject_setup.cc (no changes necessary)
• scanner.cc (Hans?)
• sdb.cc (Hans?)
• silink.cc, ssiLink.cc (Hans?)
• sing_dbm.cc, sing_win.cc (empty)
• slInit_Dynamic.cc, slInit_Static.cc
• subexpr.cc (Burcin)
• tesths.cc (Burcin, Hans, O.)
• utils.cc (O.)
• walk.cc (Martin) (fixed up to currRingHdl)
• walk_ip.cc (Burcin)
• wrapper.cc (Hans)

Singular TODOs

qring/currQuotient (O.)

• ring r; ring->cf set in rDefault (O.)
• tests in kernel & Singular needed special linking options for using modules (Burcin, O.)
• parameter handling?
• minpoly?

ToBe fixed in numbers/ directory:

• *.h (Burcin)
• *.cc (O.)
• Makefile.am + made libraries: libnumeric{.a,_g.a} out of this directory (O.)

Defects in the new Singular

• declare a ring with params (rInit) (Hans, O.)
• output the Frank's extension fields in the old Singular way (in order to avoid changing the testsuite result) (O.)
• use ring names for building towers as in ring NEW = OLD, (VARSTR), (ORDSTR); where OLD is a ring or qring (O.), BUT: interpreter should be tought more about this since there can be more valid symbol names than variables and the immediate parameters!
• minpoly = (number); seems to work now (O.)
• a**0 was a in Frank's extensions :( (fixed by Hans)
• TODO: add *KillChar to coefficient domains! esp. Frank's extensions (they should also --ref.count for ext.Ring)
• why alg. extention use minideal instead of qideal??!?