06:58 <flip214> luis: yeah, thanks!

11:03 <luis> flip214: https://github.com/slime/slime/issues/600#issuecomment-762173333

11:20 <flip214> luis: my draft uses local-addr ;(

11:28 <luis> flip214: naming might be one of 2 or 3 of computer science's most difficult problems, but we're not really doing computer science here!

11:38 <flip214> well, but that means that my approach is doomed, right from the start! it will _never_ be accepted upstream, and has to lead its empty life in a small, unaccounted repository!

11:41 <flip214> luis: https://github.com/phmarek/slime/commit/6c9634aa4975462d7be936daddbf3483c57c67eb, but I didn't even run or test that yet

11:57 <flip214> luis: I wrote that I didn't even run that yet! do you think it's fair to let others stumble over my bugs??

11:59 <luis> flip214: added your disclaimer :)

12:00 <flip214> grrr, stop changing the text ;)

12:01 <flip214> I just wanted to show _you_ what I'm about to do... NO PUBLICITY, please ;)

12:05 <luis> flip214: sorry, feeling the pressure? :D

12:10 <flip214> luis: yes. Well, I'll get even with you with LOTS of pull requests ;)

12:14 <flip214> luis: BTW, is it okay to use generic functions in swank? Then I could just have a defmethod to parse S-expr resp. JSON, dispatched on the first character (and the connection)

12:14 <luis> flip214: I think we still run into some locking in SBCL's PCL if we do that

12:15 <luis> although SBCL no longer has a global compiler lock, it still has some global PCL locks

12:30 <flip214> You mean during debugging? Well, the method definitions would (typically) only occur during _loading_ swank, and not at any time afterwards, so I don't think there's a problem

13:06 <luis> flip214: yes, but they're compiled when they're first executed, though maybe this was only a issue for custom method combinations. It's probably fine in your case, you're right (maybe!).

13:07 <flip214> let's hope so