06:45 <fiddlerwoaroof> luis: that PR looks good to me

06:46 <fiddlerwoaroof> the only slight issue I have is that it's theoretically possible for a rogue inferior lisp to intern a bunch of symbols in emacs and crash the editor

07:36 <luis> fiddlerwoaroof: that's already possible since any swank message can intern symbols.

07:39 <fiddlerwoaroof> luis: cool, I suspected it wouldn't be an issue

07:57 <luis> ed-rpc, epc, ec, ecall, efuncall? 🤔

07:59 <fiddlerwoaroof> ed-rpc works for me

08:01 <luis> You know the joke about the two hardest problems in Computer Science right?

08:02 <fiddlerwoaroof> :)

08:02 <fiddlerwoaroof> cache invalidation and cache invalidation?

08:03 <luis> Cache invalidation is definitely in the top 3. :)

08:06 <luis> Anyway, if you could use ed-rpc in the new contrib that'd be great. I'd merge the PR once you confirm it works for you

08:10 <fiddlerwoaroof> ok

08:11 <prite> The two hardest problems in CS: naming things, cache invalidation, and off-by-one errors. Right?

08:14 <fiddlerwoaroof> luis: is it possible to reload slime without restarting emacs?

08:15 <fiddlerwoaroof> I guess I could load-file slime.el, but is there a better way?

08:17 <fiddlerwoaroof> luis: one issue that occurs to me is that emacs puts everything into a global namespace

08:18 <fiddlerwoaroof> So, naming defslimefuns is a bit annoying

08:22 <fiddlerwoaroof> we could have swank automatically transform the symbol to package-name:symbol-name before sending it across to emacs

08:23 <luis> prite: exactly!

08:24 <fiddlerwoaroof> My cache didn't invalidate properly

08:25 <luis> fiddlerwoaroof: I see your point. It's not too different in the opposite direction, though. See slime-eval calls.

08:26 <fiddlerwoaroof> yeah, calling from slime to swank is easy because elisp symbols can contain :

08:27 <fiddlerwoaroof> Hmm, there's a slight issue with my buffer streams

08:27 <fiddlerwoaroof> (close stream) doesn't really do anything

08:29 <luis> What should it do? Close the buffer?

08:29 <fiddlerwoaroof> I'm not really sure

08:30 <fiddlerwoaroof> At least delete the marker, I'd think

08:30 <fiddlerwoaroof> On the other hand, it's not a huge deal, and it's probably UB to attempt to write to a closed stream anyways

08:33 <luis> fiddlerwoaroof: https://github.com/emacscollective/auto-compile is close to what you want, I think a little tweak to its require advice would do what you want

08:34 <fiddlerwoaroof> thanks

08:34 <fiddlerwoaroof> Anyways, your rpc stuff works nicely for my use-cases

08:36 <luis> fiddlerwoaroof: cool, I'll merge it soon then

08:36 <fiddlerwoaroof> I've updated my PR to depend on yours.

08:37 <luis> Re your PR, do you want to commit a minimal version first or add more features before a first merge?

08:39 <fiddlerwoaroof> I think getting the minimal version out would be best, unless there are other features you think the initial release should have

08:39 <fiddlerwoaroof> I think a lot of the stuff I want to do may require changes to the swank gray streams, which might be more disruptive

08:46 <luis> I suspect so, too.

08:49 <luis> I look forward to the buffer position marker(s) being visible/settable, BTW.

08:49 <luis> OK, so today I'll merge both PRs if all goes well then.

08:50 <fiddlerwoaroof> Cool

08:50 <luis> Got to go to work now. 👋

08:51 <fiddlerwoaroof> have fun :)