@buoyantair are you looking to use Racket for web devel (I’m taking a guess here about what line of work you’re in) or are you thinking about doing other kinds of things (systems programming, games, etc.)?

@soegaard2 thanks for submitting RN to hacker news. Currently in number 2! :tada:

No problem - I saw you forgot it.

It’s great to have something about Racket on Hacker News often.

Also a JS dev at work. I use Racket for scripting tasks folks would normally do in JS or Python (ETL, pulling data from APIs to my laptop, etc) and also build small services for my own use such as automatically exporting records from Pinboard to Pocket, etc.

I’ve been meaning to play more with https://github.com/vishesh/racketscript but the project seems kind of dormant

as a JS dev you might be interested in https://argojson.com/ | https://docs.racket-lang.org/argo/index.html

and https://riposte.in/

^ written in Racket as well

Why not make the docs part of the main foo package? Some packages do do that. That’s what binaryio, gmp, scribble-code-examples, sweet-exp, and zeromq-r do.

@pocmatos the comments about racket news are very nice, and show what a great service it is to the community

@samth it’s very exciting to see that people like it. I am having lots of fun doing it. Have lots of ideas but so much time for RN. There are other Rackety ideas that I want to nurture and I just started a contract on a very different area in compilers so things are busy at the moment. Going from working on compilers for C like languages and Javascript is a world of difference. Feel like a beginner once again. :smile:

I can also reveal I received a request to post a listing to a 100% Racket remote job. Coming up in Issue 8.

ooh

I’m open towards doing other stuff unrelated to web dev as I’ve worked too much on web dev already!

Thanks! I will check them out :3

I really like the concept of Riposte but I don’t find a lot of documentation for checking etc

@alama can help you out most likely

What kind of operations will be future safe? That’s something that I’ve had difficulty determining even in the current implementation of racket.

@ryanc @mbutterick @alexknauth and tests also go in the main package, it looks like? So main pkg = tests + docs + a dep just one other "-lib" pkg.

I think I’ve seen sometimes 4 separate pkgs — lib, doc, tests, main/meta — and I’m like, meh, that’s more ceremony and busywork than I want to spend time on in my remaining years :slightly_smiling_face: and shouldn’t submodules somehow do this for me? This seems more reasonable.

Although I guess it means you can’t use test or doc submodules. :disappointed:

@mflatt What is the expected C representation for the _string/utf-16 type? My first inclination is to assume that they are expected to be nul-terminated, just like for _bytes… but then again, I’d also expect _string/ucs-4 wouldn’t (necessarily) be nul-terminated, since the documentation says it corresponds to the native Racket string type, which can contain nuls.

(It would make sense to me if _string/ucs-4 is a special, different thing designed to correspond directly with Racket’s internal string representation, and the other types like _bytes, _string/utf-8, _path, _symbol, etc. either explicitly mention nul-handling or claim to be “plain char* strings”. But _string/utf-16 says neither, and it seems to have its own representation for the purposes of ctype->layout, so I am unsure.)

I did find scheme_utf16_to_ucs4 in the C API docs, which I assume is related, but that function requires an explicit length be specified, so it doesn’t really help with my confusion.

> Although I guess it means you can’t use test or doc submodules.

Relatedly, how can I raco test the submodule test of a submodule in a file?