thofma/Hecke.jl

Names for group types

Closed this issue · 3 comments

thofma commented

I am writing down new names for the types related to groups: https://hackmd.io/ONVPdlfARgWzkfkNkcJkgA?both#Something-with-groups

I would appreciate some help with it @ThomasBreuer @fingolfin @fieker. I guess we can live with FinGenGrpAb? But what about the thing which is now called GrpGen, a group which is represented using a multiplication table. GenericGroup does not sound right. Maybe CayleyTableGroup? MultiplicationTableGroup?

ThomasBreuer commented

If we want to keep the names PermGroup, PcGroup, FPGroup then FinGenAbGroup (as in the HackMD) or FPAbGroup is o.k.
Concerning GrpGen, I would prefer MultTableGroup to CayleyTableGroup.
Concerning the morphisms, appending Mor to the common type name of domain and codomain (as in NumFieldMor) is o.k.; then GAPGroupHomomorphism can be renamed to GAPGroupMor.
And if the domain has type X and the codomain has type Y then choosing the name XToYMor looks o.k.

fingolfin commented

I agree with all that @ThomasBreuer said, and would add that

  • I'd stress the "if" in "if we want to keep the names" (I mean, I kinda do want to keep them, but of course they are not set in stone)
  • ring types end with ...Ring so I think group types should end with ...Group
  • we just changed a lot of ring morphism types to end with ...RingHom so I wonder why use ...GroupMor and not ...GroupHom?
thofma commented
  • Are you planning to rename the group types? I have not been at all meetings and I can't find it in one of those HackMD documents.
  • I must have missed that with the RingHom. Is the convention to use ...Hom written down anywhere? (We will probably have SchemeMor, so I wonder why we would want to go for this mismatch.)

OK, so FinGenAbGroup, FinGenAbGroupHom, MultTableGroup and ...Hom it is. Thanks!