Andrea

mmm, it would be easier if you start from a command (I am not sure how to handle the on-clicl menu). This command should implement in the Nyxt prompt something...

Ah! Wonderful and useful ideas. I am rereading it, but just to say this is inspiring. My thoughts were to use `:examples` as type definitions. I didn't explore the idea...

> This would reduce the maintenance burden of manually defining every path through the graph possible, and might surprise the user at what's possible with their current buffer. I find...

> I think it should provide the type information of the data in increasing specificity (e.g. this is an emacs-lisp buffer, which means it's code, which means it's a tree,...

> it would need to walk the list of known molds, V={M0, M1, ..., Mn}, and for each mold, Mi: ... I think we could make use of a principle...

> This approach changes molds a little. Instead of defining molds as BeginStateToEndState, e.g. "CSVToOrgTable", molds should only be named to refer to their end state, e.g. "ToOrgTable" (or I...

> Molds should become as small as possible to promote reuse. Any suggestions about this is welcome! I tend to make private molds first and add to core.el and contrib.el...

Very cool you are hanging around, again pretty interesting ideas! I really like you can express in formulas the algorithm complexity, I just had a rough intuition it could pop...

This is wonderful, thanks for the initiative! And I guess this is a good time to learn more about Guix too. Do you use Guix the package manager or the...

@kat-co sorry it took longer than expected because I was playing around with guix. It seems things work from my latest commit. Just wondering if you could help me understand,...