00:00 <andarp> Looking through the docs, some source code and some videos it does seem like primitives, for all their simplicity, are useful and used in many different ways

00:01 <andarp> Pony is strange, it feels like a very functional language except everything is an object :)

00:04 <FunkyBob> I've a friend who keeps telling me to learn a functional language

00:04 <FunkyBob> I keep telling him to learn Pony :)

00:52 <mindB> I'm trying to compile the first fizzbuzz example from the tutorial (https://tutorial.ponylang.org/pattern-matching/case-functions.html), ponyc gives the following error. Any ideas?

00:52 <mindB> `src/libponyc/ast/token.c:82: token_string: Assertion `token->id == TK_STRING || token->id == TK_ID` failed.`

09:19 <andarp> mindB: I'll see if i get the same issue. What ponyc version are you using?

09:21 <andarp> im getting the same issue with ver 0.11.3-917610c

09:43 <andarp> this seems strange mindB

09:45 <andarp> here is the boiled down issue

09:45 <andarp> https://pastebin.com/5HtKdQTG

09:45 <andarp> seems to do with case functions and returning a string

09:47 <andarp> probably related to this https://github.com/ponylang/ponyc/issues/1893

10:41 <SeanTAllen> andarp: if you take a look at what happens with Pony, it all ends up being a lisp underneath. it has many levels of "strange". There's a lot of inspiration from OCaml and the ML family

10:43 <SeanTAllen> mindB: case functions are very broken right now and always have been. we've been having discussions on how to improve them (which we all want) or if we should give up the ghost and just remove them from the language like we did with delegates.

10:48 <andarp> SeanTAllen: Makes sense :) The only ML-ish language I know is F#, and I like its constructs in general - except that the syntax is a bit too chatty for me

10:49 <andarp> It seems Pony could lend itself to coding in very different styles which is interesting

18:23 <endformationage> Candle: It seems I have structured my use of HTTPHandler incorrectly as well. I am getting a hang on a second request too.

18:25 <endformationage> Candle: I think it may be due to conflicting information I took from the httpget example, in which the session is disposed of in the handler's finished behavior.

18:25 <endformationage> https://github.com/ponylang/ponyc/blob/master/examples/httpget/httpget.pony#L214

18:27 <endformationage> However, in the HTTPHandler interface code docs, it specifies that finished indicates the end of a particular HTTP message:

18:27 <endformationage> https://github.com/ponylang/ponyc/blob/master/packages/net/http/http_handler.pony#L66

18:27 <endformationage> ... and that only a single HTTP message is processed at a time:

18:27 <endformationage> https://github.com/ponylang/ponyc/blob/master/packages/net/http/http_handler.pony#L52

18:28 <endformationage> In any case, I closed the session in my finished behavior, and noticed you did as well in your gist, as per the example.

18:42 <endformationage> Thinking about this, disposing the session after the first response is handled defeats the purpose of reusing the connection.

18:45 <endformationage> I am going to restructure my handler, and without disposing the session and see where that get's me.

19:08 <endformationage> Candle: Yeah, I just don't see how to match incoming responses with their previously sent requests. Even with the handler acting on a single response at a time, the fact that responses can come in any order like you said.

19:11 <endformationage> I guess multiple clients are needed.

19:13 <endformationage> I can't see how a map would help sort things out, but I'd love to know if you can make it work :)

19:29 <Candle> endformationage: In the short term, I just made a new HTTPClient for each request.

21:36 <SeanTAllen> endformationage: could you add documentation to the HTTPClient or the HTTP package that would help folks in the future as you seem to have learned a lot