Zulip Chat Archive
Stream: Lean Together 2026
Topic: Fabrizio Montesi - CSLib: The Lean Computer Science Library
Markus Himmel (Jan 20 2026 at 13:02):
Discussion thread for the talk.
Fabrizio Montesi (Jan 20 2026 at 13:32):
Thanks for all the great questions! Lovely to see interest in CSLib.
Syed Shaazib Tanvir (Jan 20 2026 at 13:34):
Is there some place where we can follow the development of Boole? The cslib repository seems to only have a placeholder README in Cslib/Languages/Boole/ right now.
Chris Henson (Jan 20 2026 at 13:35):
If anyone was wanting to look closer at the code samples Fabrizio shared, our docs are hosted at https://api.cslib.io/docs/
Fabrizio Montesi (Jan 20 2026 at 13:39):
Syed Shaazib Tanvir said:
Is there some place where we can follow the development of Boole? The cslib repository seems to only have a placeholder README in
Cslib/Languages/Boole/right now.
There's a boole-sandbox branch: https://github.com/leanprover/cslib/tree/Boole-sandbox/Cslib/Languages/Boole
But bear in mind that the language design is still under discussion, things might (probably will) change. But it should give a general idea. The directory is still not very organised yet.
Chris Henson (Jan 20 2026 at 13:42):
You can also look directly at Strata, the dependency that gives access to Boole.
Markus de Medeiros (Jan 20 2026 at 13:43):
Would all formalizations of deeply embedded languages be required to use Boole/Strata? Or is the plan to have a framework for eg. operational semantics, and have the Boole/Strata languages instantiate it?
Chris Henson (Jan 20 2026 at 13:43):
(the branch Fabrizio mentioned pins a particular revision of Strata for its usage of Boole)
Fabrizio Montesi (Jan 20 2026 at 13:45):
Markus de Medeiros said:
Would all formalizations of deeply embedded languages be required to use Boole/Strata? Or is the plan to have a framework for eg. operational semantics, and have the Boole/Strata languages instantiate it?
It's still under discussion, and sometimes tricky because of some peer dependency issues. I expect the Lean module system to help with some of this at least though.
We do have a framework for operational semantics in CSLib, and some languages that instantiate it already (also in CSLib, like lambda and CCS). That's the framework we're building a downstream compiler on.
Fabrizio Montesi (Jan 20 2026 at 13:46):
All of automata theory is built on the operational semantics framework, too.
Markus de Medeiros (Jan 20 2026 at 13:47):
Ah very good, thank you! We were considering formalizing the languages framework for Iris-Lean, but maybe we should try using the CSLib formalization instead :)
Chris Henson (Jan 20 2026 at 13:51):
Also note that if you are considering lambda calculi as a deeply embedded language, these are independent of Boole and I imagine will remain so. I will talk tomorrow briefly of the future work I have planned for unifying our treatment of lambda calculi.
Ching-Tsun Chou (Jan 20 2026 at 19:05):
Are the slides of @Fabrizio Montesi 's presentation available somewhere? It was too early in the morning for me to attend.
Fabrizio Montesi (Jan 20 2026 at 19:23):
As soon as my website's CI/CD pipeline finishes... ;-)
Edit: it's done.
Fabrizio Montesi (Jan 20 2026 at 19:23):
Slides: https://www.fabriziomontesi.com/files/slides/cslib-lean-together-2026.pdf
Shreyas Srinivas (Jan 20 2026 at 23:50):
Is there are recording of the talk available somewhere?
Ching-Tsun Chou (Jan 21 2026 at 00:05):
Last updated: Feb 28 2026 at 14:05 UTC