Fork me on GitHub
#unrepl
<
2017-12-27
>
volrath09:12:31

SPIRAL got merged into MELPA: https://github.com/melpa/melpa/pull/5190/ 🙂 @pesterhazy I'm ready on my side to do our joined release

rickmoynihan23:12:04

@volrath: spiral’s looking really great… though I’ve noticed that sometimes it seems to crash/freeze and I can’t seem to restart it without restarting emacs.

rickmoynihan23:12:27

even when I kill all the spiral buffers

volrath23:12:04

@rickmoynihan hmm I haven't experienced that yet, would you be able to point out some steps to reproduce this behavior?

rickmoynihan23:12:32

I can try… it seems to have happened a few times to me

volrath23:12:56

when you say it freezes, do you mean the repl buffer or emacs itself?

rickmoynihan23:12:42

freezes isn’t actually the right word (though I think that might have happened earlier too but could well be my emacs)… It still seems to be in the broken state… basically running spiral-connect-to says it has connected but no REPL buffer is created

rickmoynihan23:12:54

and I can’t see any spiral buffers anywhere

volrath23:12:40

alright, do you see something like [<host>:<port>] in your mode-line for the clj buffer you are trying to connect? or maybe a [-] (brackets should be preceded by a spiral symbol)

rickmoynihan23:12:18

yes I see @[localhost:5555]

volrath23:12:26

if so, you can exec C-c q (as in spiral-quit) in that buffer..

rickmoynihan23:12:33

where the @ is your cute little spiral… nice touch

volrath23:12:56

ok, so the problem might be that when the repl disconnected, the buffer you're using didn't "disconnect"

rickmoynihan23:12:07

ok done that and I get UNREPL connection to localhost:5555 terminated

volrath23:12:17

alright, try connecting again

rickmoynihan23:12:31

ok that’s worked

volrath23:12:06

I'll check why is it not disconnecting properly, when the REPL disconnects, all related clj buffers should unlink that connection

volrath23:12:33

did you by any chance kill the REPL buffer by hand? maybe using kill-buffer or C-c C-k?

rickmoynihan23:12:53

I could well have

rickmoynihan23:12:01

but I can’t recall

volrath23:12:52

alright, maybe that was it, I'm not entirely sure I'm covering the case where the REPL buffer is killed by the user.. I'll check tomorrow for sure 🙂

volrath23:12:58

thanks for letting me know!

rickmoynihan23:12:02

ok I’ll try my best to break it some more 👹

rickmoynihan23:12:05

One other thing I noticed which I’m guessing you know about is that (read-line) doesn’t seem to work

rickmoynihan23:12:06

but I don’t really need that 🙂 was just being cruel

volrath23:12:16

haha yeah, I haven't implemented anything related to unrepl's upgrade stdin/out yet

cgrand13:12:44

Before you undertake this task we should discuss it as I’m not sure the spec is sensible

volrath23:12:03

and please do keep the bugs coming! I'm sure there might be a fair amount, the project is still very young

rickmoynihan23:12:33

yeah I know 🙂 I’ve been lurking here for a while. I’m very impressed with what I’ve seen though

rickmoynihan23:12:32

I’ve been trying to get my head around unrepl and tools.deps etc… I get the feeling the clojure tooling landscape is going to evolve a bit over the next few years

rickmoynihan23:12:17

I really love the elisions in unrepl; and the support for that in spiral is great

rickmoynihan23:12:41

I’m curious what the big philosophical differences are to nREPL though… Obviously upgrading via a socket repl (though I get the impression nREPL could probably be made to do that too)

rickmoynihan23:12:26

I’ve seen a bit about a REPL as stream vs nREPL as message-based… I’m guessing that’s the main difference but not really sure what it means in practice

volrath23:12:16

you can maybe check out this talk: https://www.youtube.com/watch?v=2P-UzXksdHs, it's about unrepl and unravel, around 23:10 you get a summary of differences between nrepl and unrepl, and also core ideas in unrepl design.

volrath23:12:49

this one is also pretty useful: https://vimeo.com/223309989

rickmoynihan23:12:51

Yes saw those talks a while ago… will watch them both again for a refresher

rickmoynihan23:12:56

thanks for the pointer

volrath23:12:19

and I think we could do a better job explaining what you are asking about unrepl, I've thought on blogging about it for quite a while, but I've been dedicating all my free time to spiral for now. It's definitely something that worth the time though

rickmoynihan23:12:00

yeah there’s never enough time 🙂

volrath23:12:12

yep, that's true

rickmoynihan23:12:43

thanks for the link — will read — watching the video again 🙂

cgrand13:12:44

Before you undertake this task we should discuss it as I’m not sure the spec is sensible