IRC channel logs

<dthompson>jfred: :)

<dthompson>btw everyone, wingo will be speaking about hoot this saturday at the scheme workshop at icfp https://icfp24.sigplan.org/details/scheme-2024-papers/6/Scheme-on-WebAssembly-It-is-happening-

<dthompson>it's sure to be a hoot

<jfred>Ooh: "All the talks from ICFP and the colocated workshops will be streamed online in real-time on SIGPLAN’s YouTube Channel for video streaming and Discord for Q&A."

<jfred>though, dang, 3 AM my time haha

<jfred>I'll catch the recording

<dthompson>ooh thanks for catching that. I must have missed it or it wasn't published yet

<dthompson>q&a is through discord though 🤮

<dpk>did they announce yet where ICFP 2025 is taken place?

<dpk>*taking

<dthompson>not sure

<dthompson>haven't paid attention to that level

<dpk>grumble

<dpk>the year it’s reasonably close to me i can’t afford it. i bet next year will be in Timbuktu or something

<dthompson>jfred: also 3am for me 🙃

<dthompson>dpk: 🙁

<dthompson>looks like the acm youtube channel is keeping the unedited stream recordings up, so maybe I don't have to get up at 3am if I can just wait until the stream ends to catch the VOD

<dckc>last night I was thinking about the (%r k1 v1 k2 v2) idiom from rockt and the js {shorthand, syntax}...

<dckc>... it evolved into some devious ideas for rockit-like macros.

<dckc>for ref: rockit = https://github.com/cwebber/rockit

<dckc>I happened by https://docs.scheme.org/surveys/property-lists/ and https://docs.scheme.org/surveys/keyword-syntax/ ... oh my! so much diversity!

<dpk>symbol properties are not a great idea

<dpk>usually the implementations that support them implement them using a hash table on the side anyway

<dpk>(the traditional Lisp implementation was to store them directly on the symbol objects)

<dpk>and the good news is that hash tables are portable

<dpk>also nice @ rockit, i hadn’t seen that before

<dckc>I'm not so much interested in symbol properties as just scheme data structures that are used for stuff like the result of parsing a JSON object

<dthompson>an alist is the most convenient thing for this

<dpk>oh, there’s a quasi-standard representation of JSON in Scheme

<dpk>this one: https://srfi.schemers.org/srfi-180/srfi-180.html#json-read-port-or-generator-object

<dpk>it’s used by pretty much every Scheme JSON implementation by default, although don’t ask me what those are :D

<dthompson>yeah that's a good converstion table

<dckc>json-null? ... sigh... scheme is stuck in emacs style pfx-blort where you have to prefix every identifier with your initials (ok, project name). Any critical mass on modules isn't likely to ever emerge.

IRC channel logs

2024-09-05.log