Skip to content
Kazuhiko Sakaguchi edited this page Nov 12, 2025 · 2 revisions

Attendees : Reynald Affeldt, Cyril Cohen, Pierre Roux, Kazuhiko Sakaguchi, Quentin Vermande

2.5.0 release

now only waiting for an infotheo release, then nixpkgs will be done and backporting on OPAM (which is lacking CI)

Splitting ssralg.v (#1275)

  • Currently, ssralg.v is about 7.7k lines long. We will probably need to split it soon.
  • The change is not urgent, unless someone wants to add more structures very soon, e.g., unitSemiRingType, semiFieldType, unitSemiAlgType, and Euclidean domains (#742).
  • How should we split the file, e.g., creating a new file dedicated to unit(Semi)RingType and below? What should be the names of the new files?
  • will need to add new structures, not even counting the ones currently in CoqEAL
  • no urgent need, except maybe discrete rings (from CoqEAL) to do tensor products for differential geometry
  • how to restructure the Theory and so on modules? should it be done before or after the split?
  • could take inspiration from the ssrnum.v restructuring
  • would loose the GRing.... qualification, except maybe using NES
  • we should test more NES, so that the day it work we can implement (and test it) in Rocq
  • should we test on a leaf (like Analysis) first? why not
  • names of files splitted from ssralg.v? structures without/with multiplicative inverse?
    • algebra.v? (structures without multiplicative inverse) maybe not a good name for a file
    • three components: commutative algebra (Z-modules, ring, modules), multiplicative inverse, third component?(characteristic and decfield?)
    • keep the morphisms in same files as structures (structures are almost useless without morphisms)
    • -> split in three

a semiring morphism structure with a commutativity condition #1493

  • needed for polynomial evaluation (need to commute coefficients and variables)
  • could we have a more general condition?
  • could be an application of Quentin's work (set subtypes recurring issue)
  • needs more discussion in Lyon

Also missing the join of ring morphism and order morphism (order preserving ring morphisms)

Question about website, metaquestion: how do we publicize/communicate?

  • documentation is scattered in Rocq website, mathcomp, analysis, opam?
  • we should organize a new mathcomp school (last one was in 2022 with buggy MC2 initial version)
  • need some standardization (standard library, common between MC and stdpp for instance, including unifying framework between typeclasses and canonical structures)

Sharing day on November 26th

  • Reynald to {impr,rem}ove the page and announce it

Next chair

Cyril

Clone this wiki locally