00:06 <discocaml> <leviroth> There’s also ppx_log to increase the lazy-ness of the allocation if you’re using the async log module

00:23 <companion_cube> if you're writing a library that logs stuff, _please_ use `Logs` :3

00:57 <discocaml> <mbacarella> nice

00:57 <discocaml> <mbacarella> i just write to stderr in that case

01:01 <companion_cube> well, if you happen to do more serious logging… ;)

01:02 <companion_cube> (also, the cool thing is to use https://erratique.ch/software/logs/doc/Logs/index.html#val-src_log )

01:02 <companion_cube> `module Log = (val Logs.src_log your_lib_src)`

01:02 <companion_cube> a smart enough reporter can filter on sources, etc.

05:14 <famubu> Hi. I was trying to get started with ocaml in a minimal way. Without extra packages and stuff with emacs as the editor. I don't need an LSP kind of thing. I was kind of used the sml-mode of emacs and would've been happy with something similar for ocaml. Is there such a thing? Does anyone know?

05:38 <discocaml> <geoff> tuareg-mode

05:50 <famubu> discocaml: <geoff>: Thanks I installed it.

05:50 <famubu> But am having trouble figuring out key bindings.

05:50 <famubu> I got this cheatsheet https://www.google.com/url?q=https://ocamlpro.github.io/ocaml-cheat-sheets/tuareg-mode.pdf&sa=U&ved=2ahUKEwjKurvqkKH-AhWfTGwGHZvaAn8QFnoECAEQAg&usg=AOvVaw3BPFYY2mhUIFEkhyHyol_W

05:50 <famubu> Dang it..

05:51 <famubu> This link: https://ocamlpro.github.io/ocaml-cheat-sheets/tuareg-mode.pdf

05:51 <famubu> So I had been trying C-c C-s to feed buffer contents to utop.

05:51 <famubu> But this key combination only seem to start utop. Updates to the buffer doesn't show up in the utop shell.

05:52 <famubu> Any idea what I'm doing wrong?

05:53 <famubu> Oh C-x C-e seems to do it.

05:54 <famubu> Now I'm trying to figure out how to have history. As in get the last line in utop to show up. Tried M-up M-down. Neither worked..

06:30 <jmiven> famubu: have you tried M-p and M-n?

06:31 <famubu> Do you know which command it runs? M-p and M-n are mapped to something else in my OS.

06:32 <jmiven> utop-history-goto-prev

06:32 <famubu> jmiven: That works! Thanks!

06:53 <famubu> Tuareg starts complaining about non-exhaustive pattern matching before I finish writing the patterns. It changes the color of the match expression to an annoying color. Is there a way to make tuareg complain only when we explicitly ask it to check the code?

06:58 <famubu> it seem to check for errors whenever we write the buffer to file.

07:05 <famubu> I defined a new type in ocaml and a function to compare values of that type was already there. Is this `compare` function automatically generated for all user defined types?

07:21 <neiluj> Hi!

07:21 <neiluj> how do you print a string in hexadecimal? like \000\000...

07:25 <neiluj> oh, %S worked!

08:04 <discocaml> <willardthor> Hi all (noob here), I have a package definition file (ending in .opam), and I wish to install it. Which commands do I run? (details: I am trying to install VST (Verified Software Toolchain) via. opam w/o installing the Coq Platform; however, the coq-vst package is no longer available in opam it seems; but the package definition is there, and it points to the VST source)

08:08 <discocaml> <willardthor> Hi all (noob here), I have a package definition file (ending in .opam), and I wish to install the package it defines. Which commands do I run? (details: I am trying to install VST (Verified Software Toolchain) via. opam w/o installing the Coq Platform; however, the coq-vst package is no longer available in opam it seems; but the package definition is in the VST Git repository, and it points to the VST source in said repository)

08:58 <adrien> I have code like the following: MyInt.(1 + 2) and I get a warning "Warning 44 [open-shadow-identifier]: this open statement shadows ..."; I'm trying to locally disable the warning but I can't find how; is that doable?

09:42 <octachron> MyInt.( 1 + 2 )[@warning "-44"] or `let[@warning "-44"] ... = ...` work. What did you try?

09:49 <adrien> I was trying to put the block somewhere close to ".(" (before, after, in the middle)

09:49 <adrien> thanks!

10:47 <discocaml> <willardthor> I resolved my problem; I was missing a "opam repo add" command, as outlined here: https://www.cs.princeton.edu/courses/archive/spring21/cos510/sf/vc/Preface.html

12:46 <neiluj> hmm

12:47 <neiluj> when I display a string the first ascii character seems to be dropped

12:49 <neiluj> basically have a function to convert between ascii string and its representation as an hex string: https://bpa.st/J33I4

12:51 <neiluj> from my tests if c = (to_hex_string s), then to_hex_string (from_hex_string c) = c

12:52 <neiluj> but from_hex_string (to_hex_string s) is equal to s except for the first ascii character

13:16 <neiluj> shortened a bit the code https://bpa.st/7ZUCY

13:17 <discocaml> <masterbuilder> neiluj: what's the deal with the backslash?

13:18 <neiluj> the backslash is for the hex notation

13:18 <discocaml> <masterbuilder> I'm wondering if it's causing an escape sequence to be expanded by some function in `from_hex_string`

13:18 <discocaml> <masterbuilder> oh okay that makes sense

13:18 <neiluj> oh

13:19 <neiluj> hmm but the from_hex_functions removes the backslash though

13:20 <discocaml> <masterbuilder> Not if it'a expanded before it gets to the filter, then it won't find a backslash

13:20 <discocaml> <masterbuilder> s/it'a/it's/

13:37 <neiluj> :w

13:40 <neiluj> okay my bad the string was missing a starting backslash

13:41 <neiluj> thanks for pointing out the issue!

13:52 <neiluj> would you use String.equal or polymorphic equality?

13:54 <neiluj> if you have custom data structures with instances that are equal while not being structurally equal then exposing and using the equal function is safer I guess

15:30 <discocaml> <NULL> `String.equal` and `( (=) : string -> string -> bool)` should behave the exact same

15:32 <neiluj> thanks

15:33 <neiluj> was thinking of custom data structures in C where two values can be equal despite having different representations in memory because the memory representations are not put in some canonical form

15:34 <discocaml> <NULL> That can happen with sets and maps at least, but strings are safe

15:35 <companion_cube> it gets messy if you want unicode normalization, etc.

15:35 <discocaml> <Armael> yes (=) can be the wrong notion of equality even without custom data structures in C, for example with sets and maps you can have trees balanced differently that represent the same set/map

15:37 <discocaml> <NULL> companion_cube: Neither `String.equal` nor polymorphic equality do that (I'm pretty sure), but it's good to remember that they in fact do not do that

15:37 <companion_cube> for sure

15:37 <companion_cube> I was just pointing out that strings can be tricky too :)

16:20 <neiluj> thanks for the interesting info!

19:32 <olle> Say you have 5 reads and 1 write in a function. Is there anything to gain from lifting out the write so that you only have reads?

19:32 <olle> Read/write from file, db etc

19:33 <olle> I was thinking that writes complicates the contract of a function more than a read.

19:38 <olle> https://discuss.ocaml.org/t/do-io-writes-complicate-a-function-contract-more-than-io-reads/11930