search

gap-system / gap

Main development repository for GAP - Groups, Algorithms, Programming, a System for Computational Discrete Algebra
https://www.gap-system.org
GNU General Public License v2.0
813 stars 161 forks source link

Use `LoadKernelExtension` and `IsKernelExtensionAvailable` in more packages #5761

Open fingolfin opened 4 months ago

fingolfin commented 4 months ago

These function have been available since GAP 4.12.0 and I think we should use them in more (ideally all) packages with kernel extension. So far only edim and browse do so.

Here is a list of affected distributed packages. The goal would be to either submit an issue or a PR to each of these over time.

james-d-mitchell commented 2 months ago

Some of us at GAPDays are going to work on this today.

Joseph-Edwards commented 2 months ago

I'm going to make a start on semigroups

limakzi commented 2 months ago

I will take json.

Joseph-Edwards commented 2 months ago

Working on cvec now.

limakzi commented 2 months ago

Working on orb. https://github.com/gap-packages/orb/pull/69

Joseph-Edwards commented 2 months ago

Working on ferret.

fingolfin commented 2 months ago

Thanks for the efforts, folks. A general comment: I think it is fine to just change all these packages to require GAP >= 4.12. After all: not many people will update a single package; most just update all of GAP at once. (This might change in the future if PackageManager gets more widely used but I don't think we are there yet, so we don't need to worry about it.

If you do it that way, then also update .github/workflows/CI.yml to not test against older (now unsupported) GAP versions.

limakzi commented 2 months ago

@fingolfin Lets take it as definition of done, if so. I'm fully OK with that. By the way, working on gauss. ;)

Pull-request: https://github.com/homalg-project/homalg_project/pull/614.

Joseph-Edwards commented 2 months ago

Doing float now

limakzi commented 2 months ago

https://github.com/gap-packages/hap/issues/123

fingolfin commented 2 months ago

Just to say: it might be a bit premature to mark a package as "fixed" in the list at the top of this issue right after the corresponding PR was merged: in my eyes it is only truly "fixed" if there also is a release of that package with the changes.