Instead of that `xml` library and format, many people use the SXML format defined by Oleg Kiselyov:<p><a href="https://docs.racket-lang.org/sxml-intro/" rel="nofollow noreferrer">https://docs.racket-lang.org/sxml-intro/</a><p>SXML became the de facto standard in the Scheme community after Oleg and others developed some powerful libraries for it, starting with Oleg's SSAX parser. SXML is also better-suited to efficient manipulation of large HTML/XML when using immutable lists.<p>Also, if you're looking at code examples in TFA, and thinking that, even when composing HTML using lists rather than strings, it's still susceptible to injection programming errors, then one SXML-ish improvement is:<p><a href="https://docs.racket-lang.org/html-template/" rel="nofollow noreferrer">https://docs.racket-lang.org/html-template/</a><p><a href="https://docs.racket-lang.org/rws-html-template/" rel="nofollow noreferrer">https://docs.racket-lang.org/rws-html-template/</a>
Almost 20 years ago I worked at company that developed a commercial product on top of MzScheme (prior to DrScheme and Racket). There was an article about it in Russian, unfortunately now only available via web archive <a href="https://web.archive.org/web/20210506123442/http://fprog.ru/2009/issue2/alex-ott-using-scheme-in-dozor-jet/" rel="nofollow noreferrer">https://web.archive.org/web/20210506123442/http://fprog.ru/2...</a>
I've been working on porting golang's net/http and gorilla/websocket libs to Racket for the past few months (I don't like the servelet model.) Does anyone have insight into how that model might not be optimal in a Scheme enviroment or how to better optimize it?
Realistically, how much data is it practical to stash inside these continuations? It seems like it could get out of hand quite quickly for non-toy examples.<p>If you were propagating e.g. game state, you'd still want to stick that in a db, so the continuation would just be a session id?