✧✦Catherine✦✧
✧✦Catherine✦✧this question is coming up a lot so i'll pin it
- replacement for github: https://codeberg.org
- replacement for github pages: https://grebedoc.dev
- easy ci runners for codeberg: https://codeberg.org/whitequark/nixos-forgejo-actions-runner
✧✦Catherine✦✧
✧✦Catherine✦✧something that i'm concerned about and which may prove itself to be true is that it's economically impossible to run a git forge offering free CI runners. it's obvious that github's are subsidized in much the same way early uber subsidized rider costs, which is concerning because they'll crank down restirctions, yes, but also because it might just not be a ... serviceable offer
✧✦Catherine✦✧was i aware that i'm relying on subsidized and potentially unsustainable CI infrastructure while investing a lot of effort into that infrastructure? yes, fully, for a good number of years
i just think that testing your software today and tomorrow is important even if five years from now you'll stop being able to test it entirely
i expect it'll be snowing in hell long before any OSS-focused forge will offer windows and macos runners, for example
✧✦Catherine✦✧the writing was on the wall years ago actually, but also: there aren't any really good ways to backup a github repo with complete metadata anyways. there's an unmaintained python script that collects JSON API responses into a bunch of files, which works correctly and can probably run in cron too (making it "above average") but viewing the results? you're fucked if you want that
✧✦Catherine✦✧something that i'm concerned about and which may prove itself to be true is that it's economically impossible to run a git forge offering free CI runners. it's obvious that github's are subsidized in much the same way early uber subsidized rider costs, which is concerning because they'll crank down restirctions, yes, but also because it might just not be a ... serviceable offer
✧✦Catherine✦✧every time i had to do an emergency infrastructure migration in the recent years (which is almost exclusively due to matrix.org people being not good at what they're doing) it had a noticeable temporary impact on my health. migrating several hundred (i don't even know how many precisely. around 200 i think?) repositories because of corporate restructuring will fucking kill me
✧✦Catherine✦✧the writing was on the wall years ago actually, but also: there aren't any really good ways to backup a github repo with complete metadata anyways. there's an unmaintained python script that collects JSON API responses into a bunch of files, which works correctly and can probably run in cron too (making it "above average") but viewing the results? you're fucked if you want that
✧✦Catherine✦✧most likely i'll bring up a complete mirror (one that includes repository metadata like issues, comments, PRs, etc as a seccondary channel) somewhere and then leave when each individual project i maintain hits actual practical problems
✧✦Catherine✦✧every time i had to do an emergency infrastructure migration in the recent years (which is almost exclusively due to matrix.org people being not good at what they're doing) it had a noticeable temporary impact on my health. migrating several hundred (i don't even know how many precisely. around 200 i think?) repositories because of corporate restructuring will fucking kill me
✧✦Catherine✦✧personally i'm not planning to migrate from github any time soon. the people who i'd've lost on ethical grounds have self-selected out long ago, i have basically no lock-in besides very heavy reliance GH Actions (which have no viable replacement besides gitlab maybe) and i'm counting on the process of corporate death to take a few years until it starts to cause actual immediate problems
✧✦Catherine✦✧most likely i'll bring up a complete mirror (one that includes repository metadata like issues, comments, PRs, etc as a seccondary channel) somewhere and then leave when each individual project i maintain hits actual practical problems
✧✦Catherine✦✧personally i'm not planning to migrate from github any time soon. the people who i'd've lost on ethical grounds have self-selected out long ago, i have basically no lock-in besides very heavy reliance GH Actions (which have no viable replacement besides gitlab maybe) and i'm counting on the process of corporate death to take a few years until it starts to cause actual immediate problems
✧✦Catherine✦✧"mass producing devices combining multiple types of nanotechnology each of which individually strains against what is even possible in our universe at all, providing access to them to the general public as a boring and everpresent commodity, then having the general public store porn on them" is peak technological civilization https://mastodon.social/@danderson@hachyderm.io/115013105037015356
✧✦Catherine✦✧@alex@feed.yopp.me @david_chisnall@infosec.exchange @janl@narrativ.es if you can automate predictions for something to the point where an LLM can do it semi-reliably, in almost every case you could, and i will argue should, define an abstraction that does it deterministically
✧✦Catherine✦✧@alex@feed.yopp.me @david_chisnall@infosec.exchange @janl@narrativ.es people have been writing abstractions over HTML for almost as long as HTML existed. you can go ahead and use .jsx/.tsx in almost any environment today; the abstractions have won