emma
"should i follow?" no.
non-moot interactions are blocked
private bunny account
- wait could i not patch the pds by using bind mounts
- i could just mute and move on but like i've never had this awful experience on bluesky, yet
- i do actually have to get rid of twitter i think my brain is rotting from going there everyday
- twitter recommending me ppl dunking on bluesky thats fun
- i think if ur sweet spot is a vile place then yeah sure twitter is ur place
- seems wild to me that fly draws risque art for comic kairakuten (h manga publisher) but has never drawn actual h
- fly has never drawn anything more than this lol
- fly please u draw good u need to draw h x.com/flyco_
- you read ero manga to get off, i read ero manga because these plotlines are so silly you could never possibly get it up from reading it
- definitely wonder what will happen if leftist circles keeps moving to anti-tech instead of owning the tech
- i dont really get the mdni culture it's been a whiplash knowing that people actively put mdni on their bio like that's going to stop... someone
- transphobia (fear of trans people cuz theyre cool and its scary)
- so i thought of doing firehose right, obviously that means i'm gonna have to do a base package for doing xrpc subscriptions because then i'd want to do the same for labels
- [Not loaded yet]
- im still not doing the labeler i think
- this is you on pro copyright brainrot
- xrpc-subscription **proto-repo-subscription **proto-label-subscription
- i dont know how this would work though tbf i think event emitters scary once it's this deep
- thats TWO more packages
- i do wish i could just throw money to other people to make decentralized **proto happen but unfortunately i dont and it sucks
- clearly i should also be writing a more generic validation library
- there we go reposted on main
- python having keyword arguments is nice but it bamboozles me when there's no consistency as to what should be positional and what should be keyworded
- it's no different than js functions accepting an 'options' object because it is implemented exactly as so fn(**opts) will have opts as a dictionary
- i'm fine with options object, i'm not fine with inconsistent call structure
- holy shit he did it
- having this issue with mangadex where pages just dont load the first time its pmo
- this was a lie
- going to chill for today cause i think i've spent the last few days just rushing to get this new stuff out of the door
- come cuddle with me in bed
- whoaaa cuddles
- forwardemail.net/en/private-bus… might be something to consider if i ever have to switch away from purelymail
- yesterday i saw a post about someone talking about windows and linux and how you could've just configured windows since you're just going to do the same if you had moved to linux and like, configuration isn't what people have an issue with
- forgot what the exact post is about sigh
- i forgot that im like 22
- too good at drawing please be a hentai artist
- could've just enforced the $type by adding it into the shape huh
- what huh why
- there's no other method for refining into a type predicate
- ah it's because of type safety being broken bc refinements always get called regardless of whether the previous refinement in the chain had failed
- im like close to finished i just need to figure out how xrpc fits into it because xrpc procedures/queries don't have a validator lol they're just metadata
- most effective is like this ig
- effectively shuts down the issue wrt type expansion
- hmm
- actually this is complicated i'm just going to introduce an optional schema
- [Not loaded yet]
- recreate and delete
- [Not loaded yet]
- caught
- maybe after this months long detour ill finally get a usable schema validator
- i really do hate protocol talk because this sort of comment just seems defeatist
- [Not loaded yet]
- yeah that's not bad i think i just dislike the idea of stopping just because of laws being passed
- [Not loaded yet]
- waow
