monnier

> If I could choose between a persistent popup showing outdated information > sometimes and a popup which disappears often, I'd take the persistent popup, > but that's just me...

> ```elisp > (when (or (bound-and-true-p syntax-begin-function) > (bound-and-true-p font-lock-beginning-of-syntax-function)) > ;; We're going to modify `syntax-begin-function', so flush the cache to > ;; avoid getting cached values that used...

> FWIW the above code looks perfectly safe to me and I can't see how/why it > would cause `/` to be considered a comment delimiter. I suspect the problem...

> --- a/contrib/sly-mrepl.el > +++ b/contrib/sly-mrepl.el > @@ -333,14 +333,13 @@ In that case, moving a sexp backward does nothing." > (defun sly-mrepl--syntax-propertize (beg end) > "Make everything up to...

João Távora [2021-06-03 08:10:19] wrote: >> What am I missing? > Probably nothing, but for an appreciation of how lazily prepared my commit > was. Feel free to fix. Indeed...

> I've extracted the bit I think matters here:``` Yes, you got it right, sorry 'bout that, it was mixed with cosmetic changes to silence some local extra warnings.

> ```lisp > (defun sly-mrepl--syntax-propertize (beg end) > "Keep everything up to current prompt comment syntax." > (remove-text-properties beg end '(syntax-table nil)) This line is still redundant ;-) > (let...

João Távora [2021-06-04 11:01:01] wrote: >> This line is still redundant ;-) > But why? Because it's baked into the syntax-propertization machinery already? Yes: `syntax-propertize` removes the `syntax-table` text property...

> The bullets are currently recognized by outline-mode and could be > folded/unfolded via Outline menu. It currently has `promote-subtree` and > `demote-subtree` functionality which does nothing for bullets. Perhaps...

> If running `proof-ready-for-assistant` won't work before sourcing `pg-vars.el`, it should probably instead be something like this: > ```elisp > (if (boundp 'proof-assistant) > (if (member proof-assistant '(nil "" "Coq"...