CoCoALib Task Table (Genova)

Task table Task descriptions Recently completed tasks

The Task Table

This table summarises the state of current and imminent tasks in CoCoALib. The first tasks are considered to be the most pressing. Full descriptions of the tasks are given below.
All time estimates and completion dates are to be considered approximate.
Task Who Status Progress Last news Completion
RingTwinFloat theory and implementation Abbott active Stage 3a 2008-12
Self-saturating (and cleanup for sugar) Bigatti+Caboara+Abbott waiting Stage 2c 2009-04 2009-09
CoCoA5 language Abbott+Bigatti+Lagorio(+Figari) active Stage 3ab 2009-05 2009-06
Examples for matrix code Bigatti+Abbott COMPLETED 2009-02-13
More operations on ideals generated by monomials Bigatti resting Stage 3a 2008-11
Hilbert: optimized use of PPWithMask Bigatti inactive Stage 1 2009-01
CoCoALib documentation: txt2tags Bigatti COMPLETED 2009-02-20
Simplify CoCoAServer output Bigatti active Stage 1 2009-01
Integration with Normaliz Bigatti Stage 1 2009-04
Integration with Frobby Abbott+Bigatti COMPLETED 2009-02-11
Convert sugar into inheritance classes Bigatti active Stage 2c 2009-05 2009-07
IsHomog wrt first rows of weight Bigatti active Stage 1 2009-01 2009-09
Inheritance-ready GReductor/Do_GBasis Caboara active Stage 1 2009-04 2009-09
Homogenous lists Abbott+Bigatti inactive Stage 1 2009-02
Constructors for ideals Abbott+Bigatti inactive Stage 1 2009-02
Names for I/O fns in CoCoAServer Abbott+Bigatti inactive Stage 1 2009-05
GlobalMgr to check debug compatibility Abbott inactive Stage 1 2009-01-29
ZZ max/min bug Abbott resting Bug identified 2007-04-10
DensePoly Bigatti+Abbott resting Stage 4bd 2009-01
Ring of infinitesimals Abbott inactive Stage 1 2007-04-04
Redesign implementation of ideals Abbott+Bigatti inactive Stage 1 2006-11-09
Polys with rational coefficients to CoCoAServer Bigatti inactive Stage 1 2007-11-29
Faster eval homomorphism Abbott+Bigatti inactive Stage 1 2007-11-30
Implement IsZeroDivisor Abbott+ inactive Just starting 2008-04-07
Improve example programs Abbott+Bigatti inactive Need 2nd pass 2007-03-19
Degree dimension in myWDeg Abbott+Bigatti inactive Stage 1 2007-09-20
CoCoAServer supervisor Abbott inactive Stage 1 2008-07-22
Unique implementation for div and deriv Bigatti+Abbott inactive Stage 2007-11-29
Redesign DistrMPoly etc. Abbott+Bigatti inactive Stage 1 2006-11-13
Optimized matrix multiplication Abbott+Bigatti inactive Stage 1 2008-04-22
Doc. for creating new rings Bigatti inactive early stage 2007-02-09
Init. of myZero/myOne in rings Abbott inactive Stage 4abcd 2006-11-16
Names of GlobalInput, etc Abbott inactive Forum discussion 2006-11-17
Equality test for modules Abbott+Bigatti+Caboara inactive Stage 1 2006-11-17
Warnings Bigatti+Abbott inactive Stage 1 2007-03-09
NTL integration Abbott+Caboara inactive Stage 1 2007-01-15
Clean up OrdvArith Abbott+Bigatti inactive Stage 1 2007-02-10
Clean up OpenMathXML Abbott+Bigatti inactive Stage 1 2007-02-10
Template fns in io.H Abbott inactive Stage 1 2007-02-10
Revise CoCoALib "build system" Abbott+Bigatti inactive Stage 1 2007-05-03
Buffering when printing out power products Abbott inactive Stage 1 2007-02-29
Reconsider The "New" prefix Abbott+Bigatti inactive Stage 1 2007-03-01
Hilbert driven GBasis Caboara+Bigatti+Abbott inactive Stage 2007-12-04
Deg incompatible PP orderings Abbott+Bigatti+Caboara inactive Stage 1 2007-10-30
Unordered PPMonoids Bigatti+Abbott inactive Stage 1 2007-10-30
Conversion from double Abbott inactive Stage 1 2008-01-28
Module and EFGModule Abbott+Bigatti inactive Stage 1 2008-02-11
IsMonomial and IsHomog for DUP Abbott+Bigatti inactive Stage 1 2008-04-24
Examples for convert and IsInteger and IsRational Abbott inactive Stage 1 2007-05-21
Compatibility of IsRational and IsInteger Abbott inactive Stage 1 2007-03-26
Floor and ceiling for ordered rings? Abbott inactive Stage 1 2008-06-13
RingDUPZ (generic) Bigatti+Abbott inactive Stage 1 2007-11-29
RingDUPZ (for Hilbert) Bigatti+Abbott inactive Stage 1 2007-11-29
Hilbert final Bigatti+Abbott inactive waiting ---
Weyl algebra Bigatti+Ali active? Stage 4a 2008-11
Extend GlobalManager's Remit Abbott inactive Stage 1 2007-03-06
PPM homs Abbott inactive Stage 1 2006-11-06
Toric Mill Bigatti+Abbott inactive Stage 1 2007-01-10
Unified RingFp Abbott inactive Stage 1 2006-11-07
Requirements analysis of interactive language Abbott+Bigatti+Caboara+Figari inactive late discussion 2007-11-29
Replace SmartPtrIRC by boost Abbott+Bigatti inactive Stage 1 2006-11-07
Rounf, floor, ceiling Abbott inactive Stage 1 2008-10-29
Mysterious test failures on Digital Alpha Bigatti+Abbott inactive Stage 2 2007-12-07
Sundries Abbott inactive Pipe dream 2007-xx-xx
Task table Task descriptions Recently completed tasks

Task descriptions (Alphabetical order)

This section contains a short description of each task. A time estimate for a task assumes "full-time" working; as ever for software, time estimates can be wildly inaccurate (usually hopelessly optimistic).

Buffering when printing out power products Michael reported poor performance when using sockets on M$ platforms. Cause was many short messages. Review use of buffering when sending PPs. Time est.: 1 week.

Clean up OrdvArith The file OrdvArith.C is full of commented out old code. Check that this code can be deleted, and either delete it or reactivate it. Time est.: 1-2 days.

Clean up OpenMathXML In OpenMathXML.H the class definitions are publicly visible. Is this necessary? If they can be moved to the .C file then I can remove the (costly) include of iostream in the header file. Time est.: 1-2 days. The .C file is full of all sort of ancient stuff including debugging code -- needs a major overhaul. Time est.: 2-3 weeks.

CoCoA5 language We shall need a new parser (and interpreter) for CoCoA 5. Much of the design of the new language has already been completed, and a late prototype of the new lexer already exists. The new parser and interpreter are still before first prototype (i.e. approaching stages (2abc)). Expect this task to split into several more focussed smaller tasks.

CoCoAServer supervisor Exhume CoCoALib supervisor from version 0.95 (not in CVS). Compare with what Stefan Kaspar has implemented Time estimate: 2 weeks - 1 month.

Compatibility of IsRational In a fraction field should applying IsInteger to numerator and denominator give the "same" result as applying IsRational to the whole element?. Time est.: decision + 3 days.

Constructors for ideals Which constructors for ideals? Now we have ideal(ring), ideal(ring, RingElem), ideal(ring, vector). Other nice constructors ideal(RingElem), ideal("homoglist of RingElem"), ideal(vector) the last should return en error when vector is empty. Also very nice ideal(RingElem, RingElem), ideal(RingElem, RingElem, RingElem)

Names for I/O fns in CoCoAServer Change read/write into input/output for fn names in server. This will be more coherent with the naming convention used in the library.

Conversion from double Add functions allowing (checked) conversion from double to an integral type -- conversion succeeds only if the value fits and was genuinely an integer. Maybe a good idea to add a rounding function which rounds the double to an integral value, possibly allowing the caller to specify the maximum amount of rounding (from 0 up to 0.5). Time est.: 2-3 days.

Convert sugar into inheritance classes There are at least different kinds of sugar (StdDeg, wdeg, saturating,..): an abstract class representing them would lead to clean code for the reduction process, and possibly also to the GroebnerMill. ReductionCog should know/return the cofactor for computing appropriate sugar more efficiently Time est.: 1-2 weeks.

Degree dimension in myWDeg The PPMonoid (public) member functions myWDeg do not check the dimension of the supplied degree. At the very least there should be an assertion. Perhaps degree objects should be resizable?? Time est.: 1-2 days (discussion and implementation) May need a new error.

DensePoly Design and implement abstract class for Dense (univariate) polynomials. There will probably be some consequences for PolyRing and SparsePolyRing. Work for the abstract class is essentially complete.

Doc. for creating new rings The documentation about how to create new types of ring is inadequate. Make better documentation, perhaps using existing rings as illustrative examples. Time est.: 2 weeks.

Deg incompatible PP orderings The "idealization" of a graded module with a PosTo type ordering cannot use directly the same grading as used in the components of the module. One way to circumvent the problem would be to allow PP orderings which are not degree compatible. Another solution would be to extend the grading on the module components by prepending a single value (which is the position) -- but this would require creating and using a different grading. We must decide what is the cleanest way forward, and then implement it. Time est.: ???

Equality test for modules Modules have characteristics in common with both ideals and rings. Currently, equality is implemented as for rings (i.e. true only for the same identical representation in memory). Instead it should be implemented more or less as for ideals: compute two G-bases, and then check that each is contained within the other.

Examples for convert The convert function does not handle overflow and underflow correctly when producing a double value. Use scalbn and C error codes to achieve correct behaviour. Add an example program showing how to use convert and related functions (e.g. IsInteger and IsRational). Time est.: 2 days.

Extend GlobalManager's Remit A number of older global values are handled by the default C++ mechanism. Several of these would be better handled by the GlobalManager. An important example is GMPAllocator: for the allocator to work correctly it must be activated before GlobalManager is created (or in the ctor of GlobalManager). Incorrect use of the two "global" managers causes an ugly crash at the end of program execution. A number of other global values exist, whose purpose is primarily for debugging, and controlling these through the GlobalManager would improve cleanliness and comprehensibility since they do interact.

Faster eval homomorphism Produce a better implementation of the polynomial ring "evaluation homomorphism"; perhaps even a better default implementation? Time est.: 2 weeks.

Floor and ceiling for ordered rings? Might it be useful to have floor and ceiling functions for ordered rings (as a complement to IsInteger)?

GlobalMgr to check debug compatibility Perhaps add code to GlobalMgr so that it can check it a user has tried to link non-debugged code with a debug version of the library (or vice versa). Possibly remove the existing code which performs the same task from MemPool. Time est.: 1 week.

Hilbert driven GBasis Complete implementation of Hilbert driven GBasis. Time est.: 4 weeks??

Hilbert: optimized use of PPWithMask Change Hilbert code so that it uses PPWithMask instead of C type eterm. Might require some changes in the implementation of PPWithMask. Use PPVector. Time est.: 4+ weeks.

Hilbert final Convert to C++ the main code of Hilbert in CoCoA4. Waiting for the special univariate polynomials and the verification of Hilbert's requirements for PPWithMask. Time est.: ???

Homogenous lists It appears it can be useful to have a (template?) class for "Homogenous lists", i.e. vectors (?) of RingElems/PPMonoidElems having the same owner. It should be reference counted. It would be quite useful for the generators of an ideal: to avoid making copies, to guarantee that all elements are in the same ring. A bit like PPVectors....

Implement IsZeroDivisor Implement a function IsZeroDivisor for ring elements. Time est.: 1/2 week.

Improve example programs The example programs could be clearer and easier to navigate through. For instance, one program could refer to others which are in some way related to it. Some of the examples should be especially simple, focussing on a single idea. Ideas for which examples to add can be taken from questions on the forum (e.g. "How do I do this in CoCoALib?"). A first hasty pass was completed on 2007-03-09, but a second pass will be needed.

Inheritance-ready GReductor/Do_GBasis Compare Do_GBasis-like functions in order to prepare code for migration to inheritance class GMill. Time est.: 1-2 weeks.

Init. of myZero/myOne in rings All concrete ring ctors currently "naughtily" create the zero and one elements of the ring. This is naughty because the ring has not been completely constructed at that point, so conceivably calling a virtual member function might behave unexpectedly. The member functions myZero and myOne could construct the cached value upon demand (but they must also decrement the reference count when construction occurs). The current code seems to work fine on every platform tried so far, but some "good style" guides say one shouldn't call virt.mem.fns. in the ctor. Decide what to do, do it, AND DOCUMENT IT.

Integration with Normaliz Normaliz is not designed as a C++ library, but it should not be difficult to make a prototype for doing so, and then link it to CoCoALib. Time estimate: 1 week for first prototype.

IsHomog wrt first rows of weight Add partial IsHomog. Time est.: 1-2 days.

IsMonomial and IsHomog for DUP Decide whether IsMonomial and IsHomog are useful for Dense Univariate Polys. Adjust impl accordingly. Time est.: 1 week.

More operations on ideals generated by monomials more operations on ideals generated by monomials. Now we have Gbasis/interreduction, and intersection.

Names of GlobalInput, etc The names GlobalInput, GlobalOutput, GlobalLogput and GlobalErrput are cumbersome (perhaps even ugly). Devise better names, and ideally remove the need for the annoying parentheses.

NTL integration Without doubt NTL offers some impressive implementations of algorithms for dense univariate polynomials and matrices of integers (or rationals?). Some NTL functions should be integrated into CoCoALib as a pilot project for understanding how conditional builds can be achieved -- CoCoALib should build equally easily with or without NTL cooperation (of course, some operations may be missing from a CoCoALib with NTL support). Time estimate: 3-4 weeks, perhaps longer if difficult decisions are encountered.

Redesign DistrMPoly etc. The designs of DistrMPoly, DistrMPolyInlPP, and DistrMPolyInFpPP are poor (due to scope creep during development). While the code does appear to work well, it is fragile and hard to understand/maintain. Given that this code forms a major part of the foundations of CoCoALib, it is essential that this code be redesigned and rebuilt before version 1.0. The new design must be cleaner, easier to understand, and no slower (in any circumstance). Probably the best period for undertaking this work is early 2007; it is work which produces "invisible progress". It is quite likely that other work on the library will be suspended during the rewrite. Time estimate: 1-2 months.

Redesign implementation of ideals The implementation of [ideal] in CoCoALib is old and seems to have severe problems. We need a new analysis of requirements for the implementation. Some "policy decisions" may be needed (e.g. consult the forum?). The conclusions will (probably) appear in an internal CoCoA Report. Probably a complete rewrite will result. Analysis will take 1-2 weeks. Consequential rewrite will take 2-3 weeks.

Revise CoCoALib "build system" Review and update the Makefiles and configuration scripts after the new source code reorganization. Time est.: ~1 week

Reconsider The "New" prefix A current naming convention uses the prefix "New" for certain functions which create new values (e.g. NewFractionField). However, the use of the prefix is not universal (e.g. RingElem ctor), and the prefix is somewhat C++-ish. An alternative could be to use the prefix "Make". We should also review precisely when it is appropriate to use the prefix, and document why. Time est.: 1-2 weeks.

RingTwinFloat theory and implementation Revise "twin float" implementation in accordance with revised theoretical ideas. Time estimate: 2+ weeks.

Ring of infinitesimals A ring for computing with first order infinitesimals would be handy for some. Time est.: 2 weeks (for design, impl, and testing).

Module and EFGModule Decide whether module and EFGModule should be merged or kept distinct. Note that the documentation for EFGModule is largely absent. Time est.: 2 weeks.

Mysterious test failures on Digital Alpha Investigate mysterious bugs which cause some tests (ComputeSyz) to fail on Digital Alpha when compiled with optimization. No idea when we might get it fixed. JAA thinks it is probably a compiler bug.

Optimized matrix multiplication Matrix multiplication is currently done in a "dense" way (maybe because the programmer is a bit dense :-) make it work more efficiently when at least one arg has a special structure (e.g. sparse, diagonal, completely zero, etc). No sure idea how to do this in a general way -- maybe each matrix could give an indication of how "clever" it is a multiplying (perhaps an estimate of run-time cost???) Time est.: 2 weeks.

Polys with rational coefficients to CoCoAServer Modify CoCoAServer to allow polys with integer coeffs, and to allow fractions when coeffs are rationals. Time est.: 2 weeks.

PPM homs Implement homomorphisms for PPMonoids; maybe pseudo-homs too.

Requirements analysis of interactive language Review of the requirements for the new interactive language and how they interact. The aim is to produce a report which analyses the various compromises and then indicates a final choice to proceed with for the production of a parser. All decisions should be justfied so that a later reconsideration could be effected cheaply. Time est.: 3-4 weeks.

Replace SmartPtrIRC by boost The BOOST library offers a sophisticated smart ptr implementation. Can we use that instead of SmartPtrIRC? If so, there will be several consequential changes to the rest of the library. Analysis will require several days; no time estimate for consequential changes.

RingDUPZ (generic) Concrete class for univariate polynomials over any ring. Inspired by the efficient implementation of unipoly for Hilbert from CoCoA-4. Time est.: 1-2 weeks.

RingDUPZ (for Hilbert) Special concrete class for univariate polynomials over Z; inspired by the efficient implementation of unipoly for Hilbert from CoCoA-4. Time est.: 2-3 weeks.

Rounf, floor, ceiling Make ordered rings offer round, floor & ceiling functions (since these cannot be implemented efficiently using the existing interface). Time est.: at most 1 week.

Self-saturating (and cleanup for sugar) Buchberger's algorithm for non-homogeneous input can be implemented as a self-saturating process (similar to what was done for Toric by Bigatti): Input polynomials will be homogenized with h; then every power of h which is a factor of a polynomials in the running computation may be cancelled out. This might be a mathematically clean way to interpret sugar (for GradingDim 1). Theory is all set. Clean implementation waits for task sugar and GReductor/Do_GBasis

Simplify CoCoAServer output Simplify the way to send output in CoCoA4io.C: some functions print the object itself, others print the assignment to the communication variable. All objects should have a direct printing function (then called by the assignment function).

Sundries Here are some ideas for algorithms to implement in CoCoALib. Factorization over the reals, complex root finding.

Template fns in io.H The presence of the template functions in io.H requires that iostream be included, but iostream is rather lengthy, so compilation becomes slow. How to rectify the situation? Time est.: < 1 week!

Toric Mill Design and implement Toric Mill using inheritance to "choose" between the different concrete algorithms. This will be pilot design for the Groebner Mill. Time est.: almost 1 month.

Unique implementation for div and deriv Like for other functions, also div and deriv should probably have a unique implementation in SparsePolyRing.C instead of 3 copies in the DistrMPolyXX.C files. Time est.: 1-2 days.

Unified RingFp Unify the three different implementations of RingFp, RingFpLog, RingFpDouble -- perhaps use a SmallFpArith object to choose the implementation?

Unordered PPMonoids Consider modifying existing polynomial code so that it tolerates PPMonoids which are not arithmetically ordered (e.g. PPMonoidModSquares). The elements of the PPMonoid must be ordered so that polynomials have canonical forms. Multiplication of a polynomial by a PP will be more complicated: there may be zero divisors, terms may merge, terms may swap ordering, etc. Decide how invasive such a change would be, and whether implementing it is worthwhile.

Warnings We have sometimes wandered whether some commands should return an error or not, for example "content(f)" when the coefficient ring is a field. Mathematically "1" is a correct answer, but probably it means that the program is not doing what is intended by the user. Maybe a "warning" mechanism could be usefully applied to such cases.

Weyl algebra Weyl algebras have been revived: need proper testing.

ZZ max/min bug Max found a bug in the use of max/min with temporary ZZ values. Found out what really happens; fixing it will not be simple. Time est.: >1 week.

Development stages

To describe the progress of each subtask (contributing C++ code to the library), I suggest using the following scheme for identifying what has been done, and what is still to be done. The intention is that normally all lower numbered stages be completed before progressing to the higher numbered stages. For other types of task, the task description should give an indication of progress.
  • (1) subproject start (incl specification of subproject)

  • (2a) abstract algm design, choice of data-structures, etc.
  • (2b) initial test set
  • (2c) early prototype, mostly works, maybe some bugs

  • (3a) late prototype, fully working, almost no bugs
  • (3b) comprehensive test cases

  • (4a) final C++ version (incl. relevant build commands)
  • (4b) maybe extra tests
  • (4c) portability and quality control issues
  • (4d) documentation

  • (5a)-(5d) independent verification of stages (4a)-(4d)

Status

For each task, the table indicates the current "status". Here are the main choices for the status:
  • active being worked on currently
  • inactive not being worked on currently (but work should begin fairly soon)
  • waiting progress suspended until some other task has progressed enough
  • resting temporarily suspended (e.g. holiday)
  • other explanation of actual status is with task description
Task table Task descriptions Recently completed tasks

Recently Completed Tasks

Tasks we believe to have been completed are moved from the table to here. Older items in this list will be removed from time to time (to avoid gathering too much clutter). The dates are only approximate.

2009-05-20 There is now a global option for specifying whether elements of a quotient of ZZ are printed as least non-negative residues or as least magnitude residues. The default is least magnitude (a change from the previous deault behaviour). The residues are now always printed without brackets -- previously brackets were used in some cases and not in others.

2009-05-20 Modified the Makefiles to use the archiver ar in a more economical manner; this make partial recompilation usefully faster. The Makefiles have been cleaned/improved as well.

2009-02-20 CoCoALib documentation is written in pure text. Using txt2tags we can keep it clean/readable while allowing pretty formatting and linking into HTML and TeX. Proper polishing will take some time.

2009-02-13 The example ex-matrix1.C has been reorganized and extended with examples for (Const)MatrixView Abbott+Bigatti

2009-02-11 configure arguments Version 0.9928. (1) Configuration for optimization/debugging using --flags=ARG. (2) Configuration for linking frobby using --with-libfrobby=ARG and added tests (just to check the linking has been successful) in cocoalib and cocoa4. Abbott+Bigatti

2009-02-09 New: abstract class for sugar Version 0.9927. First prototype for abstract class for sugar. Still unused, but makefile, dependencies and library.H had to be changed. Bigatti

2009-01-30 Enum instead of bool arguments in GReductor GReductor constructor had many boolean arguments: they have been converted into more meaningful "enum"s. Bigatti, Caboara

2009-01-30 Frobby integration Version 0.9926. First check-in for Frobby files. Configuration process not properly defined yet, but all code is in the right place and compilation flag is set. Makefile, dependencies and library.H had to be changed. Bigatti, Roune

2009-01-14 Improve configure script Configure behaves badly when gcc exists but g++ does not. Fix the script so that it works gracefully in both cases. One or two problems still remain (e.g. 64-bit on Mac G5) Abbott

2008-12-17 Machine integers in place of RingElems Now using MachineInteger type whenever a machine integer value is used in place of RingElem. This avoids nasty surprises when C++ silently converts signed values into unsigned (or vice versa). New version 0.9925 to signal this wide-ranging change. Abbott

2008-12-16 Makefiles version 0.9924. Changed all Makefiles for compatibility with Solaris make which has no -C option: make -C dir blah now is (cd dir; make blah) [Thanks to Paul Whitener for testing on Solaris] Bigatti

2008-12-11 Revised symbol Now using MachineInteger type in ctors to specify the values of symbol indexes (but we check that the values fit into a long). Corrected two bugs in SymbolRange (an undetected overflow and an off-by-1 error). This improvement needed the update to MachineInteger values. Abbott

2008-12-11 Updated MachineInteger Revised MachineInteger. Fixed a bug, and added the new function IsSignedLong which simplifies use of MachineInteger values. Abbott

2008-11-24 Reorganized SOI/NBM, server changed too SOI/NBM now reside in their own separate file. Implementation has been tidied considerably, and enjoys a better structured organization. Each algorithm now offers three coherent interfaces, while the actual work is done by a single implementation of each underlying algorithm. Abbott

2008-11-24 Preprocessing is now rational, server changed too The preprocessing algorithms are now rational, so give exact results (rather than the rational aproximations produced when all input data were converted into doubles). The speed of computation remains essentially unchanged. Abbott

2008-11-18 Fixed ideal inheritance code for RingWeyl version 0.9922. The missing ideal constructor in RingWeyl caused the ideal to be generated by the default constructor in SparsePolyRing (the code is identical). So the ideal did not call the specific RingWeyl::ideal operations. Now test-RingWeyl1 checks that 1 is in GBasis(ideal(x, dx)). Bigatti

2008-10-08 Error Codes version 0.9921. Error messages need to be expressive, but also short. To help non-english speakers, each error should produce an alphanumeric key/code which can be used to search in a manual of errors. We want to create an "index" of all coded errors with a more verbose explanation. Possibly it should also include some information about the most common occurences of a given error and some suggestions for debugging and/or fixing it. See CoCoAErrorCodes.html: CoCoA web page or local link in doc/.
Moreover the way to check the type of error in an ErrorInfo object has been changed (and herein lies the backward incompatibility). The new way should be simpler to use. Abbott

2008-09-22 Updated communication with CoCoA-4 version 0.9920. cocoa-4 and CoCoAServer communication has changed (you need the new cocoa5.cpkg for CoCoA-4): OperationCommunication allows variable number of arguments and sends the information to the server. So the function myReadArgs in RegisterServerOps.C now takes an extra argument, NumArgs. (added function linkability for the Frobby library, updated code for TmpMayerVietorisTree, added NBM) Bigatti

2008-09-22 Updated communication with CoCoA-4 version 0.9918.

  • moved code for monomial ideals into TmpPPVector
  • added code TmpMayerVietorisTree for irreducible decomposition of monomial ideals (E. Saenz de Cabezon)
  • added stable order ideal algorithm (SOI)
  • fixed multivariate GCD, same as CoCoA-4.7.4
  • exponent overflow (and underflow) should now be detected when CoCoA_DEBUG is set.
Abbott, Bigatti

2008-04-22 No more IsPositive/IsNegative for ZZ Removed these two silly little fns from ZZ.H. In compensation, the usual infix comparison operators now handle specially the case when one of the args is 0. Note that MachineInteger does still have IsNegative! Abbott

2008-04-21 A few more matrix ops The adjoint function now works over every ring (added AdjointByDetOfMinors). Added some matrix norm functions (Frobenius, 1-norm, infinity-norm). Added PseudoInverse for full rank matrices. Moved fns related to PPOrderings into the new file MatrixForOrdering.H/C. Abbott

2008-04-18 Major revision of matrices New version: 0.9917 Added the concept of "matrix views". This should offer a good compromise between flexibility and efficiency. Matrices now have a coherent design principle (which I hope is "correct"). Minor changes may still occur as experience with the new design grows. Abbott

2008-02-12 Improved configuration improved GMP identification during configuration Abbott

2007-12-06 Bug in RingDenseUPolyCleanImpl ctor A very well hidden bug has been fixed in RingDenseUPolyCleanImpl ctor. This depends on the creation of RingElems (0, 1, x) before the completion of the constructor.... so, just remember this: myRefCountZero(); must be the last line, and RingElems and RingHoms must be just before it. Bigatti

2007-12-05 StdDeg, log, and exponent Main changes are:

  • log/myLog(pp, i) renamed into exponent/myExponent(pp, i)
  • log/myLog(poly, i) renamed into MaxExponent/myMaxExponent(poly, i)
  • return type for StdDeg/myStdDeg changed from size_t to long
CoCoALib version is now 0.9913. Bigatti

2007-11-28 Fix Eval in CoCoA-4.7 The Eval tried to be too general and failed. Fixed the code and the manual -- should be mathematically cleaner too. Abbott+Bigatti

2007-10-30 Move to GPL-3+ All major parts of the source of CoCoALib are henceforth released under the GNU Public licence version 3 or later. Previously the code was distributed under version 2 only. CoCoALib version is now 0.9912. Abbott

2007-10-10 Hilbert-Poincare version 0.9911. Another step in rewriting the Hilbert-Poincare code:

  • added TmpPoincare.[CH] using DenseUPolyRing instead of unipoly (the old C code)
  • modified TmpHilbert.C so that it can call the new code (HilbertNumeratorMod) as well as the old one (HilbertNumeratorMod_C).
Bigatti

2007-09-28 CoCoA-4.7.2 Double quotes for string constants As part of the "gentle migration" to a better language, change all strings in CoCoA-4 packages to be between double quotes instead of single quotes. Check other places too (e.g. manual). Abbott

2007-09-25 CoCoAServer version handling The contructor for ServerOpBase requires a LibraryInfo. This way the server can print all libraries which have been linked in (or the list of all available operations with the library they are defined in) Bigatti

2007-09-24 Release of CoCoA-4.7.3 Bigatti

2007-06-06 Resuscitate RandomBitStream Make RandomBitStream use GMP random number generator (which in turn uses the Mersenne Twister). Turned out to be easier than expected, which probably means there's a nasty bug hidden away somewhere. Code still needs a little tidying. Not sure how efficient it is either. Version is now 0.9907. Abbott

2007-06-06 Check indet names properly It should not be possible to construct rings where a symbol has an ambiguous interpretation. Perhaps even the arity of a symbol should be fixed in a ring (i.e. you cannot have x[1] and x[2,3] in the same ring). One version of NewPolyRing_DMPI seems to check this correctly but other pseudo ctors do not. Abstract a checking fn, and rectify the implementations. Much the same applied to PPMonoids. Bigatti

2007-06-01 CoCoA-4.7.2 matrix facilities Added functions CharPoly, Eigenvectors, and modified Det to work better on matrices of rationals. Updated the manual accordingly. (implementation subsequently updated twice) Abbott

2007-05-31 CoCoA-4.7.2 string constants CoCoA4 now warns about obsolescent syntax for string constants inside single quotes. The idea is to encourage all CoCoA4 users to migrate to double quotes in readiness for the transition to CoCoA5. All CoCoA4 packages and tests have been brought up to date (easier said than done). Anna has updated the CoCoA4 manual entry about strings. Abbott+Bigatti

2007-05-31 Zero to power zero gives error Not sure what the code did before, but now any attempt to compute the zeroth power of zero gives error. Abbott

2007-05-30 Snapshot releases Bigatti has a script to create a snapshot release of CoCoALib-0.99xx. Then the table in the page ChangeLog-0.99 has to be manually updated with release notes.

2007-05-29 CoCoA-4.7.2 matrix facilities Added functions RowMat, ColMat, NumRows, NumCols. Updated the manual. Bigatti

2007-05-23 Release CoCoALib-0.99 The first public version with the new code organization. To be synchronous with first public release of ApCoCoALib by Dortmund. Bigatti

2007-05-22 PushFront/PushBack Modify the interface to the PushBack family of functions to make it more friendly: now they all accept zero coefficients and make sanity checks in debugging mode. Bigatti

2007-05-21 CoCoALib-0.9904 in CVS New version to "celebrate" the new MachineInteger class. Also fixed a long languishing bug in the linear system solver. Abbott

2007-05-21 Proper handling of machine integers Added new class MachineInteger in an attempt to circumvent undesirable automatic type conversions between C++ machine integer types. Class ZZ modified accordingly to act as a "proving ground". Abbott

2007-05-03 New source code organization Implement the new code organization agreed upon Genoa. The configuration and make files still need some tweaking. Bigatti

2007-05-03 CoCoAServer Mark III Redesign the server to accommodate properly the new code organization. The new design must improve modularity still further, and add suitable safety checks. Getting the design right is of particular importance here. Bigatti+Abbott

2007-05-03 PPMonoidEvZZ A new PP monoid with essentially unlimited exponent size. It seems some people need it. The code is new, and largely unchecked, but does pass a non-trivial test. Abbott

2007-04-04 Release CoCoA-4.7 Of course, this is not strictly work on CoCoALib, but it is related. Many of the new features in CoCoALib-0.98 will become accessible through the new CoCoA-4.7. A modified "internal" version of CoCoA-4.6 already exists, but we still need to document and check everything thoroughly. Bigatti,Abshoff

2007-04-04 Release CoCoALib-0.9801 Conduct all the usual pre-release checks (portability, documentation, etc). Release CoCoALib-0.9801 (with revised server, and some minor bug fixes). Abbott,Bigatti,Abshoff

2007-04-04 Integrate Gin5 and preprocessing into C4.7 Modify server to offer Gin5 and preprocessing fns; extend C4.7 interface package. Add tests for new CoCoALib fns (preprocessing and Gin5). Update the C4.7 manual. Bigatti, Abbott

2007-04-04 Integrate numerical routines into C4.7 Complete integration of the "numerical" routines into the C4.7-CoCoALib interface: minor revisions to the CoCoALib server, minor revisions to the C4.7 packages. Produce a test suite for the numerical functions (NB aim is to verify our code, not the underlying floating point matrix library). Complete manual for the numerical fns; Abbott can help with the language polishing. Abshoff, Heldt

2007-03-31 CoCoA4 manual XML tags Conclude dictionary of XML tags for the CoCoA4 manual. Bigatti, Hiddemann

2007-04-04 Revise CoCoAServer CoCoAServer has been redesigned so that it is easier to add new functions. The new design has vastly improved modularity, and should "scale well". It should make it easier for non-members of the CoCoA Team to add their functions. While the new design is surely a great improvement over the old, evolutionary one, it cannot yet be regarded as stable. After more experience has been gained, the design will be reviewed. Bigatti

2007-03-21 New code management structure After a review of existing code management structure and some consequent difficulties, a new management structure has been drafted allowing greater autonomy (and hopefully reducing "email synchronization"). The transition to the new structure should occur immediately after Easter; we expect the draft to be approved shortly.

2007-03-19 CoCoA Meeting in Genoa Last week Genoa welcomed almost everyone involved in the project. It was the second intense week in a row: important meetings, presentations about the various aspects of the project, numerous mini-meetings to make good use of the physical proximity of collaborators, etc.

2007-03-09 CoCoALib-0.98 Released After a long week of very intensive work by Abbott, Bigatti, Abshoff and Heldt, the internal version of CoCoALib was brought up to a level which would permit public release. Everyone learned that eight days' notice is definitely not enough for a full public release. This task was actually completed almost on time (i.e. only about 7 hours late).

2007-03-08 Sugar strategy in Buchberger's algorithm The implementation of the sugar strategy was incomplete: certain cases could not be handled. A stop-gap solution has been found and implemented; now everything works properly. The whole structure will be revised at a later date.

2007-03-07 Remove magic numbers from MemPool The MemPool code contains some magic numbers (minimal loaf sizes). Abstracted these into named values, and improved documentation.

2007-03-07 Update coding conventions The CoCoALib Coding Conventions need to be improved after discussions during and after the minicourse in Dortmund. Especially guidelines for when to use inline and when not. The guidelines are now publicly accessible on WWW at Genova. A further update will occur after 0.98 has been released.

2007-03-07 Improved compilation flags Modified compilation flags (with g++): more stringent warnings, and position independent code (almost never harmful, and sometimes beneficial).

2007-03-06 Dynamic Buchberger Algm First prototype of the Dynamic Buchberger's Algorithm. Uses a new way to compute the pairs, but there is still a problem with the GM criteria. Added procedures for the lex comparison of Hilbert functions, for use in the Dynamic Buchberger's Algorithm. Implemented the simplex algorithm, stage one. This implementation uses the CoCoALib matrices. This implementation uses GMP integers, and is for use in the Dynamic Buchberger Algorithm, NOT for huge optimization problems. Max.

2007-03-06 Sugar and PosWDegTO in the case of PosWDegTO orderings, Grading dim is now set to 0 and the sugar is now used. This is the case for most operations. This fixes a bug, since we cannot (yet) use WSugar as it does not work in these cases. Max.

2007-03-06 Interreduced non-homog GBasis In non-homogenous case, a new final stage interreduces the resulting Groebner basis. Max.

2007-03-06 FacetComplex name clash The name complex clashes with a name from the standard library. Changed our name to FacetComplex. Max.

2007-03-06 Eliminate RingZ type There was a special C++ ring type for the ring of integers, viz. RingZ. It did not appear to be of any great use in CoCoALib, and removing it did not make any existing code significantly harder to read/comprehend. Combined with making RingZ a global value, the updated code actually becomes simpler and more natural.

2007-03-06 Unique ring of integers After due deliberation we decided that Z, the ring of integers, and Q, the field of rationals should be unique "pre-existing" structures in the CoCoALib C++ run-time environment. This will make life rather easier (and more natural) for users of the library. Rather than use an ad hoc technique for managing these global values, we designed and implemented a coherent scheme for managing them (see next task).

2007-03-06 Management of Global Values A consequence of deciding to make the ring of integers a global value was the need to manage such a global value properly. The tricky part was ensuring the proper destruction of RingZ before main exits. Recall that C++ normally destroys globals after main has completed, and that the order of destruction cannot easily be governed; destroying values in the wrong order can cause to the program to crash just before it terminates. Another advantage of forcing destruction before main exits is that it makes debugging very much simpler. And of course it is simply good manners to clean up properly at the end of the program.

2007-03-05 Fix test-factor1 Randomization means that test-factor1 can appear to fail, though it is behaving correctly. Fixed it so that the random effects cannot affect the validity. Several other tests were cleaned up too.

2007-03-05 Remove "New" Prefix for some homomorphisms The New prefix on some homomorphism pseudo-constructors is not useful and reduces readability of code using them. Change the names of the pseudo-constructors to eliminate the prefix, and ensuring that the resulting code will be easily readable.

2007-03-05 Canonical Homomorphisms Implement CanonicalHom which allows easy creation of certain canonical homomorphisms (and hopefully more readable code). Document clearly which cases are covered by CanonicalHom.

2007-02-27 Printing for DistrMPolyXXX objects The three DistrMPolyXXX implementations repeat virtually identical code for printing out the polynomials. This code could be moved up to SparsePolyRing so that there is only a single copy BUT then DistrMPoly objects would no longer be directly printable (for debugging). Perhaps this would not matter if they were not proper objects...?

2007-02-20 Index of examples The "examples/" directory needs an index page to help find and remind what the various examples illustrate. There are two levels of index: a brief "one-line" summary for each example, and a fuller description.

2007-02-08 Generate BuildInfo.C automatically Rather than generate BuildInfo.C automatically (which was deemed to be a bit too dynamic), an equivalent effect can be achieved simply by using special compilation flags for that one file. This removes the need for the ugly and complicated CXXFLAGS which we previously used for all files. In fact, CXXFLAGS was simplified further than expected: all the tricky stuff is concentrated inside the special case for BuildInfo.C. CoCoAServer.C needed some minor adjustment. A big benefit is that the command for compiling against the CoCoA library has become considerably simpler. Added a new example, updated the corresponding documentation.

2007-01-29 One line defns are obsolete Modify CoCoA4 so that one line fn defns produce an "obsolescent" warning, and replace all one line defns in the packages by proper fn defns.

2007-01-24 Unify three PolyRingHoms The implementations of the three PolyRingHoms (in the three different sort of SparsePolyRing) are virtually identical. Now we have polynomial iterators, it should be possibly to unify them into a single implementation in SparsePolyRingBase. Same thing for coeff embedding homs.

2007-01-18 Make TmpHilbertDir Make a separate directory for the Hilbert stuff. Perhaps following the scheme used for the factorizer code.

2007-01-16 Rename "raw" args Improve readability/comprehensibility of code by using "suggestive" names for fn args which are "raw" values. This conversion generally happens in parallel with other reviews of the code. It will eventually require a final "closing" assessment. While the task is simple, it is time-consuming to be sure that absolutely all changes have been effected. It does not seem possible to get the computer to check that the task is complete.

2007-01-15 SmartPtr in RingHom Use SmartPtrIRC in RingHom, probably as a data member. Was not as quick and simple as predicted, but did lead to an overhaul of the code (esp. regarding printing of homomorphisms).

2007-01-10 Use SmartPtrIRC in ring Put SmartPtrIRC in rings, modules; substitute ad hoc reference counting with reuse of SmartPtrIRC.

2007-01-11 Clean includes in header files Sutter's coding conventions suggest using forward declarations instead of including header files. This makes (re)compilation faster, and expresses more clearly the true dependencies in the code. Task should take at most 1 week, unless more unpleasant surprises surface.

2007-01-11 Fix typedef ring The use of a typedef for [ring] causes trouble as a typedef cannot be forward declared as a class can. [ring] should become a trivial class forwarding all its operations to a SmartPtrIRC data member. Everything can be inline. Task should take at most 1 day.

2007-01-10 Attach old factorizer Attach old CoCoA4 factorizer to CoCoALib; mostly involves fiddling with makefiles and the "build system". A working factorizer was put in CVS for version 0.9713. This stopgap solution should be enough for a few months.

2006-12-22 Minicourse Prepare CoCoALib minicourse (to be held in Dortmund); Bigatti might help. This task will occupy almost all of Abbott's time until the end of 2006.

2006-11-16 Migrate into SparsePolyRing The code for ideal in the multivariate polynomial rings is essentially repeated three times: simply copied for each of the specific ring implementations. This code should be migrated into SparsePolyRing, so that just a single copy exists. The migration should be possible, but this must be verified. It may also be possible for ideals in RingWeyl to be covered by this same migrated generic implementation -- again this must be assessed. Time estimate: 3-7 days.