07:27 phillip joined 07:40 lizmat left 07:41 lizmat joined 08:01 phillip left
tonyo going to be moving the zef eco index to a binary format when i move the infra to raku *only* 16:33
alternatively i can put an API into the new server if you'd rather go that route and use an API key 16:34
(where you just request newest modules since the last request) 16:35
JRaspass I presume there would be a Raku module for parsing this index that we could use too? But an API also works, I guess what I'm saying is whatever works for you, I quite like the idea of doing diff updates though like you suggested, would stop future re-scans getting slower and slower. 17:02
jjatria Interesting! I'll be curious to see how that looks 🍿 17:05
tonyo yea for sure, the module will be a part of zef 18:30
it is essentially just an encoded binary tree index
but we're looking at that as it speeds up zef's module resolution by an absurd amount (50kX for exacts and 400% for partials) 18:31