00:06 <discocaml_> <NULL> The usual integration for emacs is tuareg

00:10 <technomancy> oh, I see; tuareg covers font-lock and other stuff too, not just repl

00:10 <technomancy> I'm used to those being separate

00:10 <technomancy> thanks

00:11 <companion_cube> Btw, hi technomancy :)

00:11 <discocaml_> <NULL> I don't use emacs so I couldn't have helped more, but I'm glad you found what you needed

00:12 <technomancy> companion_cube: hey, nice to see a friendly face

00:14 <technomancy> my kids are taking a CS class and I'm thinking of adding another bonus unit on the end that covers some other more interesting languages

00:17 <companion_cube> "see", but yeah :)

00:17 <companion_cube> a CS class at what level? High school?

04:35 <technomancy> they're in high school but the class is https://www.edx.org/cs50

06:59 <discocaml_> <Bozhidar> That's that CCBot? Is this some IRC bridge? At least to me it always looks super weird that some conversations are coming out of nowhere.

07:01 <discocaml_> <Bozhidar> (it's also not very clear how to interact with people on the IRC channel)

07:10 <Armael> from my perspective you're the one talking through an IRC bot ;-)

07:11 <Armael> the bridge bots could definitely be better integrated, but I think they are better than no bridge at all

08:25 <discocaml_> <lukstafi> (BTW the discussion of a `Typedtree` pass for PPX happened in 2018: https://github.com/ocaml/ocaml/pull/2161)

08:28 <Armael> /w 43

08:28 <Armael> (oops)

09:28 <discocaml_> <Bluddy> I wish a typed pass for PPX was practical, but there's a good reason nobody does it.

09:52 <discocaml_> <Bozhidar> Fari enough. 🙂

09:54 <discocaml_> <Bozhidar> Actually they both cover pretty much the same things, but tuareg is the better option. There are some plans to eventually sunset caml-mode altogether.

10:30 <discocaml_> <NULL> (The messages you reply to aren't passed through the bridge, you should prefix with `Armael:` and `technomancy:` instead)

11:40 <discocaml_> <froyo> the previous bridge changed its name depending on who sent the message from IRC, I liked that so I wrote a small js plugin to discord to restore the functionality:

11:40 <discocaml_> <froyo> https://cdn.discordapp.com/attachments/436568060288172044/1077555313466675321/image.png

11:40 <discocaml_> <froyo> much nicer to read this way

13:54 <discocaml_> <Drup> @kit_ty_kate do you have update package for llvm 15 ?

13:54 <discocaml_> <Drup> @kit_ty_kate do you have an updated package for llvm 15 ?

13:56 <discocaml_> <Drup> (it would be nice to have a repository with the last version of the opam files, to be able to modify it/pin it easily)

14:44 <discocaml_> <Kate> @Drup https://github.com/ocaml/opam-repository/pull/22142 and https://github.com/kit-ty-kate/llvm-dune

14:46 <discocaml_> <Kate> the two reasons this hasn't been merged is that it's a significant change from previous build system for this package (breaks previous assumptions when used with ocamlfind) and llvm 15 wasn't available on any distributions until recently

15:08 <discocaml_> <Bluddy> dune (include_subdirs qualified) is in!

16:19 <companion_cube> Does it namespace according to directories?

19:56 <discocaml_> <Drup> @kit_ty_kate great ! thanks a lot

19:57 <discocaml_> <Drup> @kit_ty_kate btw, llvm.moe is extremely outdated, I think it would be best to remove it (and maybe change it to a link to the online doc on opam doc, or something)

19:58 <discocaml_> <peval> will the GC survive if i modify the size field of a block to get "lazy compaction" when it runs?

20:11 <discocaml_> <Kate> I'll keep it for now. The API didn't change much since LLVM 7.0 imo, and the documentation in ocaml.org is not built for llvm

20:41 <discocaml_> <lukstafi> I decided that the most fun way forward for me for now is `ppx_minidebug` that's a single-file poor approximation of `ppx_debug`. It will trace only bindings and functions which are type-annotated with single-identifier types. It will log to a file using `sexp_of_<t>`. I'll advance to `ppx_debug` when this starts feeling impoverished.

20:41 <discocaml_> <lukstafi> I decided that the most fun way forward for me for now is `ppx_minidebug` that's a single-file poor approximation of `ppx_debug`. It will trace only bindings and functions which are type-annotated with single-identifier types. It will directly log to a file using `sexp_of_<t>`. I'll advance to `ppx_debug` when this starts feeling impoverished.

21:44 <discocaml_> <lukstafi> (Using `sexp`, it might be easy to implement an interactive debugging mode where one can substitute a different value for a traced value and let the computation proceed, but I'm not planning on that.)

21:44 <discocaml_> <lukstafi> (Using `sexp`, it might be easy to implement an interactive debugging mode where one can substitute a different value for a traced value and let the computation proceed, but I'm not planning on that. Maybe someday.)

22:31 <darchitect> hey guys

22:32 <darchitect> does anyone know why when I try to do 'dune --help' I get a blank screen with no information whatsoever on a macos

22:32 <darchitect> ?

22:34 <discocaml_> <NULL> Does `man` work for you ?

22:39 <darchitect> yeah

22:39 <darchitect> it does :/

22:40 <discocaml_> <NULL> That's weird, I thought it just launched it but apparently there's more to it if it breaks for you

23:36 <dh`> all you really need to know about STREAMS is Dennis Ritchie's famous quote: "I designed streams, they build sewers"

23:36 <dh`> bah

23:36 <dh`> also s/build/built/