00:37 <companion_cube> you can keep some parts of the stdlib, too

00:37 <companion_cube> I suggest Str, Digest, Buffer…

07:22 <discocaml> <lukstafi> Ctypes is very actively developed, as the author maintainer points out it had a release 3 weeks ago.

07:54 <discocaml> <emiletrotignon> You have the same thing with `true` and `false` which I think they could have went with `False` and `True` to fit with the rest of the syntax. You can actually do

07:54 <discocaml> <emiletrotignon> ```ocaml

07:54 <discocaml> <emiletrotignon> type 'a list = true of 'a * 'a list | false

07:54 <discocaml> <emiletrotignon> ```

07:54 <discocaml> <emiletrotignon> You have the same thing with `true` and `false` which I think is weirder, they could have went with `False` and `True` to fit with the rest of the syntax. You can actually do

07:54 <discocaml> <emiletrotignon> ```ocaml

07:54 <discocaml> <emiletrotignon> type 'a list = true of 'a * 'a list | false

07:54 <discocaml> <emiletrotignon> ```

08:14 <reynir> If I have a type supported_hashes = [ `MD5 | `SHA512 ] how can I assert that supported_hashes is included in Digestif.hash' ? If I write type supported_hashes = [< Digestif.hash' > `MD5 `SHA512 ] I get complaints about the 'a type variable

08:45 <octachron> You can have the assertion as a function `let f (x:supported_hashes) = (x:>Digestif.hash')`

08:47 <octachron> you could write your initial example as `type supported_hashes = private [< Digestif.hash' > `MD5 `SHA512]` but this means that supported hashes is some unknown set of variant between Digestif.hases and [`MD5|`SHA512]

09:10 <reynir> interesting

09:11 <reynir> let _ (x : supported_hashes) = (x:>Digestif.hash') is a syntax error >:(

16:40 <discocaml> <luck33> What is the process for finding "standard" packages for X? Coming from Rust, there's definitely a lot of "best in class" libraries. The Rust cookbook has strong recommendations.

16:58 <discocaml> <yawaramin> try https://ocaml.org/cookbook

17:48 <discocaml> <luck33> Who could've foreseen the analogue to "<language> cookbook" for OCaml would also be "OCaml cookbook"?

17:48 <discocaml> <luck33> @yawaramin : thank you 🙏

17:52 <discocaml> <luck33> Okay, I won't lie the cookbook in its current state is ***extremely*** bare.

18:01 <discocaml> <contificate> be the change you want to see 🙏

18:32 <discocaml> <luck33> this would be a textbook case of the blind leading the blind 🙈

18:32 <discocaml> <luck33> i think in a month i'll feel more confident to try my hand 🙂

18:32 <discocaml> <luck33> I think in a month I'll feel more confident to try my hand 🙂

19:26 <discocaml> <yawaramin> there's also awesome-ocaml and the Ocamlverse wiki

20:33 <discocaml> <nsmmrs> You will probably have decent luck if you take examples from a published cookbook for a different language, and ask an LLM to translate the example to OCaml.

20:33 <discocaml> <nsmmrs>

20:33 <discocaml> <nsmmrs> Alternatively, there is Rosetta Code, where you can compare lots of examples of "language you know" with OCaml.

20:35 <discocaml> <nsmmrs> The LLM strategy will probably work best if your input code is also a statically typed language (doesn't necessarily have to be one you know).

22:25 <discocaml> <enza7059> https://tenor.com/view/cat-mitosis-gif-5938797153464928683