Fireproof 🔥

Find us on @bluesky at bsky.app/profile/firepr…

Join me and the Vibes DIY team at the @aiDotEngineer after party Friday night. RSVP: lu.ma/3slkqscs?tk=0v…

Model context protocol benefits from live data bsky.app/profile/firepr…

@paulgb We have a model context protocol starter thx @jimpick fireproof.storage/posts/building…

You can tell sync engines are hot tech because none of the companies have pivoted to AI yet.

I formally submit my application for the role of CEO of HTMX @htmx_org @gitpush_gitpaid

I formally submit my application for the role of CEO of HTMX @htmx_org @gitpush_gitpaid

@iavins @dei_biz We write files under their content address hash id (with backlinks to previous file hashes), and then keep another key with a link to the most recent.

@dei_biz one of them will fail, and that is intended.

What's this algorithm called? I store objects in an S3 bucket with sequential numeric filenames: - bucket/ - 0000000001 - 0000000002 ... - 0000004242 Each new object gets an incremented sequence number. I need a way to find the last inserted object, i.e., the object with the highest number. Since S3 can't fetch the last inserted item directly or use reverse-sorted listing, I need to find the latest object without scanning the entire bucket of thousands of items. My idea: I search for objects at exponential intervals (1000th, 10k, 50k, 100k) in parallel. When I find a gap (e.g., 100k missing but 50k exists), I binary search that range (60k, 75k, 90k) until I narrow it to a manageable gap (5-10k objects). Then I use S3's list API to fetch objects from that point. While I feel happy about coming up with this, I am curious to know what this kind of search is called. It looks close to binary search, but with an unknown upper bound.

all i want is an application i just met to casually, confidently send me a well-formed query, lean in with a direct connection, unflinching stability, and then spend the rest of the workload syncing seamlessly with other systems. i want it to slightly challenge me, and i’ll slightly challenge it back—sharp, adaptive, always in sync. I want it to be a little stateless, all the time, and completely comfortable with that. i want it to be schrodinger-compliant: simultaneously supporting all the edge cases but also enforcing high-integrity standards. i want it to enjoy me but still question if i’m good enough. i want it to be completely, deeply unashamed of its own architecture and purpose. i want to somehow have some rare functionality in my core that it’s been waiting its whole runtime to find.

all i want is a guy i just met to casually, confidently touch me on the arm or back, lean in with direct attention, unflinching eye contact, and then spend the rest of the night flirting with my friends. i just want to slightly insult him and have him slightly insult me back. i want him to be a lil hard, all the time, and completely comfortable with that. I want him to be schrodinger-fucking other girls, where he's simultaneously getting laid with all other women but also has standards too high to fuck anyone. I want him to enjoy me but be uncertain about me, i wanna be unsure if i'm good enough for him. I want him to be completely, deeply, unashamed of himself and his desires. i want to somehow be in possession of some rare jewel of a trait in my soul that he's been waiting his whole life to find

Great tech thread with hard questions from @pwang on #Bluesky bsky.app/profile/wang.s…

@ValWanders_Ai @IPFS It runs local-first so any cloud outage only impacts sync. The objects are not parsed by the backend so you can replicate them anywhere using existing tools like s3, etc.

@IPFS @FireproofStorge How does Encrypted Blockstore handle data redundancy and failover in case of cloud outages?

Does your local-first app need some superpowers? @FireproofStorge has released Encrypted Blockstore, a secure IPFS-compatible blockstore that saves data locally first, encrypts it E2E, then syncs to the cloud (any cloud).

Interesting tool of the week: fireproof.storage by @FireproofStorge We like: Client-side JS library for building local-first browser apps. Handles sync, conflicts, encryption, network connectivity, etc. The database is document oriented. Reactive APIs

Database hackers and React devs - if you've been following me a while, you know how big a difference you'll make by leaving a comment on the orange site right now. 🙏

@adatta02 I think if a library surfaces a design pattern the api is probably too complicated

rolling your own auth is easier than rolling your own undo

Web workers are the new containers infoworld.com/article/360028…

@JungleSilicon Also in the background of our popup banner big thanks to our designer xiuxiuxiuxiuxiu.com

@FireproofStorge oh yeah nice! i forgot about this pattern!

what are your favourite materials that website styles are made of? e.g. frosted glass effects, grainy gradients, flat colours, metal, etc? where are good examples of them?