SageMath Accepted for GSoC 2026

[Travis Scrimshaw]

Hi everyone,

   I'm sorry for posting about this somewhat late after receiving notice. SageMath was accepted again this year for Google's Summer of Code (GSoC).

However, we need to have people willing to mentor for this to continue to be a successful program for SageMath. If you would be willing to mentor this year, please email me (tcscrims(a)gmail.com) or post here. No commitment is necessary at this time. You do not need to be an expert in SageMath to be a mentor (e.g., you can co-mentor with someone while you focus on the math or programming aspects).

Another thing that is good to have is more ideas, they can be things you (or a colleague) want implemented, something somewhat larger you want to be improved, or just some wishlist item you're not even sure how to do (yet). If you have such ideas, please also either post them here or add them to the GSoC ideas page: https://wiki.sagemath.org/GSoC/2026

Let me know if you have any questions. Please also ask colleagues you might be interested for ideas and/or to be a (potential) mentor.

Thank you,
Travis Scrimshaw

[Martin R]

Thank you so much for doing this!

I'd be very happy to co-mentor "Refactor the diagram algebras/monoids and add new ones" and "Improve (free) module implementations".

Martin

[David Coudert]

Hi Travis,

thank you for taking care of this.

I'm willing to mentor a student this year, if we have strong proposals and enough slots.

Best,

David.

[Michael Orlitzky]

On 2026-03-01 18:23:32, Travis Scrimshaw wrote:
>
> Another thing that is good to have is more ideas, they can be things you
> (or a colleague) want implemented, something somewhat larger you want to be
> improved, or just some wishlist item you're not even sure how to do (yet).
> If you have such ideas, please also either post them here or add them to
> the GSoC ideas page: https://wiki.sagemath.org/GSoC/2026

A relatively straightforward but still large-enough project would be
to create a replacement for AA and QQbar within Sage based on Fredrik
Johansson's Calcium library. There is an old ticket for this,

https://github.com/sagemath/sage/issues/30222

and in the meantime the task has become simpler, because Calcium was
absorbed into Flint:

https://flintlib.org/doc/introduction_calcium.html

(You wouldn't have to get a new library adopted into Sage and every
linux distro before you can start.)

The new classes can be written behind the scenes, without risk of
breakage. All software estimates should be judged with suspicion, but
I think you could get them to act like fields within the "medium
difficulty" time frame. Replacing AA and QQbar in the namespace with
the new implementations would ratchet up the difficulty a bit, but if
the project ends 80% completed we would still have something useful.

[Vincent Delecroix]

Thanks Travis for handling the process again.

I will not have time myself for GSOC this year. But I would like to
enhance Michael proposal (if there is a strong enough candidate) :

* flint has two implementations of qqbar : one working with
"multivariate ideals" and one working with "univariate minimal
polynomials". Depending on the usage, one is more suited than the
other. It is not clear how to interface both in sage (the drop in for
AA/QQbar replacement would be the former for sure).

* the former calcium is much more than algebraic numbers as it can
handle things like sqrt(exp(2) + sqrt(5)). I think that if the project
goes well there could be a SymbolicRealField/SymbolicComplexField on
top of calcium

Best
Vincent

> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/sage-devel/aaY5Dk3OGhQPwEby%40mertle.