IRC channel logs
2026-02-05.log
back to list of logs
<theesm1>re goblins shepherd again: also switched my work laptops, which uses x86_64 ISA, home shepherd to the goblins one; even though I don't think ISA is too relevant here but it's running on two different ISA now <sneek>Welcome back dthompson, you have 2 messages! <sneek>dthompson, ArneBab says: shepherd on goblins sounds like goblins is becoming more critcal (fleet management), so it could be time for pentests. Sadly the security testing capacity of https://nlnet.nl/NGI0/review/ has been exhausted … <sneek>dthompson, ArneBab says: just listening to the beyond javascript talk again: Python on Guile by stis suddenly sounds much more powerful than before -- maybe finally a way to fix brython? <dthompson>ArneBab: yeah security audit stuff will be nice to have in the nearish future <dthompson>re: python on guile, I don't know anything about it <ArneBab>dthompson: is there a simple way to get support for other languages into compile-wasm? <ArneBab>in #guile I wrote results from tests I did with hoot and brython and pyodide: hoot requires roughly a third of the download size of brython or pyodide for the same usecase. And a 3x advantage on download size pretty huge. <ArneBab>so getting python-on-guile (an implementation of Python as a guile language) integrated with hoot could make Guile hoot a better runtime for existing applications. <dthompson>ArneBab: wow that's a significant statistic. pretty cool! <dthompson>there's no way to compile non-scheme languages yet <dthompson>do you know what python on guile lowers to? is that scheme or tree-il or something else? <ArneBab>this means tree-il, right? #:compilers `((tree-il . ,compile-tree-il)) <ArneBab>(seems that gitlab is currently completely overloaded) <dthompson>I think to get other languages into hoot we need to formalize the concept of whole-program compilation for each language that can be compiled this way <dthompson>each language needs to implement something that outputs tree-il representing an entire program rather than a single module <dthompson>for scheme, that's the library-group stuff. python would need its own strategy <ArneBab>that would mean that for wisp it should be almost trivial, right? <ArneBab>Is there a timeframe to merge hoot back into Guile? <dthompson>ArneBab: was talking to Andy about this recently. no real time frame but what we do know is that hoot is still changing at a rate that is much faster than guile so merging it into guile now would slow hoot down