Review Days 3 is a one-day online event held February 7, 2017. The Review Days will officially run from 9-17, GMT+1, or until everyone runs out of steam.

We will communicate using Slack and Framapad. See below.

Organizers

Daniel Augot, David Lucas, Julien Lavauzelle, Johan Rosenkilde, Clément Pernet.

Please contact Johan Rosenkilde (jsrn[at]jsrn.dk) for any question.

Motivation and Focus

This Review Day follows up on SageDays75 on Coding Theory, six months after conclusion of the ACTIS-project on developing Coding Theory for Sage.

The main goal is to finish and review many of the open tickets on Coding Theory functionality. See CodingTheoryRoadmap on Trac.

How to get together

This is a virtual meeting, so every-one is invited to join from her/his preferred location. To communicate, we propose the two following virtual rendez-vous:

• A Slack public chat room with possibility to create sub-channels and private discussions if necessary;

• A Framapad collaborative pad to keep track of the working agenda, and to be updated with status reports

Participants

Feel free to add yourself to this list if you plan to participate for some of or the entire day.

• Daniel Augot (daugot)
• David Lucas (dlucas)
• Johan Rosenkilde (jsrn)
• Clément Pernet (cpernet)
• Julien Lavauzelle (jlavauzelle)
• Luca De Feo (defeo) — afternoon only
• Bruno Grenet (bruno)
• Matthew Rennekamp (mrennekamp)
• Arpit Merchant (arpitdm) - joining intermittently

= Aftermath: Copy-paste of our Framapad =

Reviews being done:

• # 20335: BCH codes: solved the conflicts, doctests pass, but need a deeper review (Julien) --> we almost finished, it just remains to add some doc to explain the link between the BCH code and its underlying GRS code.

Rebase and Review

• #20849: improve grs.py docstring (cpernet) -> merge conflicts solved

Improve and Review

• # 20938: Decoding algorithm for low-order q-ary or binary Reed-Muller codes (Johan) #21088: Merged conflicts, resolved doctest errors and fixed deprecated statements (Arpit)

Other ideas

• CyclicCode : overwrite dual_code()

  the ZeroCode should be available (Sage supports VectorSpace of dimension 0)

Done during Sage Review Day 3

• # 21328: Parity check code: just need a quick final review => Done (David)

  #21131 Interpolation, Minimum vanishing polynomial and Multi-point evaluation -> Positive Review

  #20899: AbstractLinearCode should throw sensible error messages on printing

  #20087: __call __ (message) on linear code or encoder should return a codeword -> Positive Review

  #20898: Syndrome decoder for a linear code sometimes sets wrong decoder type : -> Fixed, positive review # 20908: Reage index and catalogs for sage.coding (David, Johan) # 21333: Asymptotically fast (GF (q) [x]) [y] root-finding. (Johan) # 21024: Faster k [x] matrix reduction (could be closed # 16742 as invalid). (Johan)

  # 20087 (__call __ (message) on linear code or encoder should return a codeword): cpernet => done (positive review)

  # 20001: Wrote the table for decoder types and their definitions (David) => (positive review)

  #21159: Cached generator matrices and parity check matrices should be immutable: cpernet => fixed (positive reviewed) # 20953: minimum distance and gap (Julien)

  #21156: dimension() uses _dimension even if it does not exist (bruno) => Positive review

  #20787: GolayCode looks OK, but we are stuck in the --long coverage because of olf chinen_polynomial() stuff.