Roadmap and status report for Sage-Combinat

This page is an attempt at drawing a road map for  Sage-Cominat, and in particular the migration from  MuPAD-Combinat.

Todo: extend the history, and include links to the relevant trac tickets.

Overview

All Sage-Combinat tickets

Timeline

• February 2007: First contact with Mike Hansen who wanted to port some features of MuPAD-Combinat, which we very much encouraged. Since then Mike translated 30k lines of code, which accounts for most of the basic combinatorics (tableaux, permutations, ...), and symmetric functions.

• June 2007: design discussions between Nicolas and Mike at the Axiom Workshop 2007

• January 2008: presentation of MuPAD-Combinat at the AMS meeting in San-Diego; meeting and discussions with the Sage team

• February 2008: Sage Days 7 (Los Angeles)
  • Technical experimentation with Sage to see how fit it is for our purposes.
  • Partial port of the crystals library (#2742, AnneSchilling? and NicolasThiéry?)
  • Implementation of Xin's Omega algorithm by Jason and Greg.

• Spring 2008, in particular at MSRI (Berkeley, USA):
  • Long discussions within the community about the opportunity to switch
  • Port of root systems (#2808, #2809, #2864, #2868, #2874, #2964, #3660, #3664, NicolasThiéry?, DanBump?, JustinWalker?, MikeHansen?, TomDenton?)
  • Further port and extensions to the crystals library (#2868,#3032,#3417,#3418,#3660, AnneSchilling?, DanBump?, JustinWalker?, BrantJones?)
  • Basic setup for FreeModule?'s
  • Posets (FrancoSaliola?)

• June 19th 2008: Visit of Florent to Davis. Final decision to migrate!

• June 24th 2008, FPSAC (Valparaiso, Chile):
  • Official announcement of the migration
  • Goal: elementary combinatorics users can start directly with Sage

• Summer 2008: port of decomposable objects (from MuPAD-combinat) / species (from aldor-combinat) by Mike Hansen, funded by Google

• September 08: announcement that Sciface is purchased by Mathworks (Matlab).
  • MuPAD does not qualify anymore as a "reasonably priced high quality computer algebra system".
  • Sciface cancels its formerly liberal licence policy for MuPAD-Combinat developers.
  • Plan for a last stable release of MuPAD-Combinat dropped.

• October 08: Sage Days 10 (Nancy, France)
  • Get the core MuPAD-Combinat developers started with Sage
  • Design, prioritization, planning
  • Design of the categories and (Hopf) algebra framework using the new coercion system

• Winter 08-Spring 09
  • Most new basic users can start directly with Sage
  • Cleanup and refactoring of the combinatorics (#4549, #5200, #5308, #5487, #5534, #5551, #5781, #5600, #6000) (FlorentHivert?, NicolasThiéry?, ...)
  • Cleanup and refactoring of root systems (#4326, #4327, #4608) and crystals (#4311,#5729,#3663, #5002, #5729, #5879) (AnneSchilling?, DanBump?, NicolasBorie?, StevenPon?, NicolasThiéry?, ...)
  • Implementation of the categories and Hopf algebra framework (#5891) (NicolasThiéry?, ...)
  • Refactoring of symmetric functions (#5457) (NicolasThiéry?, JasonBandlow?)
  • Implementation of QSym, NCSF (NicolasThiéry?, LenniTevlin?, MikeZabrocki?)
  • Words, ...
  • Families: ..., #5538
  • Technical patches: #5120, #5405, #5449, #5783, #5843, #5920, #5967, #5979, #5985, #5986, #5991 (NicolasThiéry?, ...)

• July 2009:  FPSAC'09 (RISC, Linz, Austria)
  • Goal: Most features ported
  • Goal: Most of the research done with Sage-Combinat
  • Goal: All new users can start directly with Sage

Progress

Topic	Progress	Priority	Comments
Basic combinatorial classes	75%		2007-08 by Mike
Decomposable objects / Species	75%		Google summer of code 2008 by Mike
Trees	0%
Posets	%
Words	%
Copy-on-write data structure	30%
Symmetric functions	90%
k-Schur & the like	50%
Root systems / ...	75%
Crystals	75%
Category framework	50%		Beta patch in sage-combinat
Hopf algebra framework	10%
Free modules & such	20%
Algebra (desosseur, ...)	0%
* with several bases	0%
Operads	0%
Linbox interface	100%		(compares to 10% in MuPAD)
GAP interface	100%		(compares to 1% in MuPAD)
Interface for fast Gröbner basis	100%		(compares to 0% in MuPAD)
Nauty	100%		(compares to 50% in MuPAD
Symmetrica	60%
GLIP	50%		Deprecated by Robert's generic code (also funded by Google)
dot2tex	10%		Delegated to the SAGE graph theorists
Database access	100%
MachineIntegerListsLex?	0%		Will be easy via cython
Rigged configurations	0%
Basic abstract data structures	100%		(fast stacks, AVL, dancing links) compares to just the basic ones in MuPAD + no real way to implement some with serious speed ourselves