3949
Comment:
|
4313
|
Deletions are marked like this. | Additions are marked like this. |
Line 5: | Line 5: |
David Harvey, Martin Albrecht, Joel Mohler, Robert Bradshaw, William Stein | [:DavidHarvey:David Harvey], [:Martin Albrecht:Martin Albrecht], Joel Mohler, Robert Bradshaw, [:WilliamStein:William Stein] |
Line 12: | Line 12: |
* Write a benchmarking module. | * Write a benchmarking module or improve on the existing one. |
Line 18: | Line 18: |
* Wrap NTL finite fields and polynomials over finite fields. * Benchmark omalloc. (''Martin'') |
* Make GivaroGFq the default finite extension field implementation ([:MartinAlbrecht:malb]) * build [:MallocReplacements:omalloc] as a shared library (or link it statically to memory.so the Pyrex wrapper) and make sure as much stuff as possible uses it (SAGE_malloc macros, GMP, etc.) * Wrap NTL finite extension fields (order $\geq 2^{16}$) and polynomials over finite fields. |
Line 21: | Line 22: |
A page to summarise and coordinate SAGE development activity.
Core arithmetic team
[:DavidHarvey:David Harvey], [:Martin Albrecht:Martin Albrecht], Joel Mohler, Robert Bradshaw, [:WilliamStein:William Stein]
Develop new __mul__ semantics for rings/algebras using something like ToyQuarternionAlgebraElement as a test case.
Migrate _add_, _mul_ etc to new _add_sibling, _add_sibling_cdef, framework etc. Intention is to eventually remove _add_ everywhere.
Change all __add__ methods (etc) to use new fast code pathways, pyrex tricks, as studied at SAGE Days 2.
Try changing a.parent() == b.parent() conditions to a._parent is b._parent in various places and see what breaks.
Write up more entries in WritingFastPyrexCode. (everyone)
- Write a benchmarking module or improve on the existing one.
- Optimise pyrex in coercion module.
Make Algebra _base_ring into a cdef attribute.
Move univariate Polynomial class down to Pyrex, and make it inherit from CommutativeAlgebra instead of CommutativeRing. (David)
Rewrite polynomials over Z and Z/nZ using direct NTL calls. (David)
- Optimise new integers mod n code using fast pyrex tricks.
Make GivaroGFq the default finite extension field implementation ([:MartinAlbrecht:malb])
build [:MallocReplacements:omalloc] as a shared library (or link it statically to memory.so the Pyrex wrapper) and make sure as much stuff as possible uses it (SAGE_malloc macros, GMP, etc.)
Wrap NTL finite extension fields (order \geq 2^{16}) and polynomials over finite fields.
- Linear algebra (perhaps William or Robert want to say something here?)
Lower priority issues
- Make generic dense polynomial class in pyrex using C arrays to store coefficients?
Making SAGE Releases (Packaging and Integration)
- William Stein
- Getting sage-1.4.1 ready:
- New maxima, pari, clisp.
- Editing and including axiom interface.
Interfaces
- William Stein, Bill Page
- Axiom -- Bill Page (a 0.2 version is done; Stein is integrating it)
- MATLAB -- William Stein -- work in progress
Notebook/web team
- Tom Boothby, Alex Clemesha, Dorian Raymer, William Stein
- Change the Notebook over to use Twisted
- Add SSL support for the Notebook (with Twisted/pyOpenSSL)
- Make public Notebook more secure
no cross-side scripting by passing cell output thru MoinMoin
- send passwords (at minimum) over SSL
- Sage spawns sage0 in a changeroot jail which is firewalled to only listen on the port assigned to it. This sage0 serves the notebook, which spawns an additional sage0 for each worksheet -- these, in turn, are in smaller changeroot jails which have no internet access, and can only read/write from within their worksheet directory.
- Add notebook help browser
- Wikify some parts of the Notebook
- Add ability to edit plain text Notebook cells and then re-upload
3D Graphics
Alex Clemesha, Robert Miller, Tom Boothby ...
- Add 3D Graphics built on top of matplotlib
Fix & re-optimize tachyon plotting
Generalize interface with plot3d(f, ..., style='tachyon') to place camera & light sources apppropriately, etc.
Graph Theory Implementation
Emily Kirkman, Robert Miller, Bobby Moretti, Tom Boothby (eventually)
[http://sage.math.washington.edu:9001/graph Main Project Wiki]
[http://sage.math.washington.edu:9001/graph_survey Existing Software] list; includes links to several algorithm pages also; list is now complete
[http://sage.math.washington.edu:9001/graph_benchmark Initial Benchmarking] now beginning; purpose is to test construction, storage, retrieval and very basic functions
- currently interfacing NetworkX with SAGE
- experts will be contacted regarding a) the completeness of the list, b) the best benchmark tests
thorough benchmarking to begin soon: we will set up another wiki for results
- after benchmarking, the real fun begins...
GSL
Josh Kantor, William Stein, Tom Boothby
- RDF, CDF are done. Are matrices / vectors?
- Wrap more special functions, add to sage.misc.functions...