The idea is to take boring components: PostgreSQL, Bifrost (LLM gateway), Open WebUI, LanceDB, Agentgateway (MCP and OpenAPI gateway) and deploy them in Fly.io. One Fly.io "org" per user. The closest equivalent is blaxel.ai, but it caters for AI SaaS startups, not individual customers.

The combination of the fact that Fly secrets are visible only from within the apps, distroless containers, and transparent data encryption for PostgreSQL assures that the service (Pocketdata) provider cannot access their data, only the infrastructure provider (Fly.io) theoretically can, but practically speaking, this gives an extremely high degree of privacy assurance.

The latest update on the project: https://engineeringideas.substack.com/p/tasklet-is-the-o1-mo...

It's a custom assembler built on top of the LLVM assembler (llvm-mc) that emits instrumentation code to catch ABI violations at runtime. Stuff like clobbering nonvolatile registers, misaligning the stack pointer, misusing the redzone, assuming volatile registers don't change across a function call, etc.

Hoping to finish up basic x86_64 support within the next few days. I can now reliably assemble and run unoptimized gcc output without hitting false positives, but I still have to iron out some false positives triggered by OpenSSL's handwritten assembly routines.

TODO items for the near future include porting the runtime support library into a kernel module so I can instrument Linux, and beginning ports other architectures (ideally something semi-obscure like POWER or RISC-V). I also need to figure out how to support dynamic linking, because the tool currently needs static linking to access its thread-local variables.

https://github.com/kenballus/llvm-project/tree/abisan/llvm/t...

https://github.com/addpipe/webcam-tester

Live demo @ https://addpipe.com/webcam-tester/

I've been building a Decentralized Database built on top of syncing CRDTs, and recently got it to a point I can demo. It's definitely in a "proof-of-concept" stage though, known security holes and all.

I've been focused on building out the featureset and keeping everything unstable instead of trying to finalize each piece as I build it. It's the opposite of how I normally build things but I think it's been working pretty well for this.

I've written about it a few times, most recently "Using CRDTs + Sync as a Database" - https://jackson.dev/post/crdts_as_database/

Instead of DOM scraping, it intercepts AJAX calls and figures out which API endpoint gives you the data you need. Uses visual analysis + fuzzy matching to identify the right call.

The use case: scraping product reviews, paginated listing data (products), etc. Existing AI scrapers either didn't work or were very slow and costly. A product with 1000 reviews takes 10+ minutes with Playwright, costs $10 with LLM scrapers. With Strot? 10 seconds via direct API calls.

Being used in production by a couple of clients. Would love feedback!

Blog: https://blog.vertexcover.io/strot-is-a-api-scraper GitHub: https://github.com/vertexcover-io/strot

Decided to pivot and start learning about databases and their internals more. Currently pulling down Clickhouse and reading some code along with the reading the book Database Internals by Alex Petrov.

So I'm technically not "working on" an app...I am working on myself to branch out and attempt to specialize a bit more as I progress in my career.

Any advice/papers/books to read is very welcomed!

https://play.google.com/store/apps/details?id=app.radarlove....

Just launched last month.

I’ve always wanted a typing application that’s both more than typing random words and is data-focused so I built this.

The more you type, the more the analytics system learns about your typing patterns and generates natural text to target those weakpoints (SmartPractice mode).

There’s a lot of variety as well; you can practice typing code in any programming language, or type text of various topics, use custom text, etc).

Users can create their own sub-communities, and within them, set up different categories and boards. Posts can be voted on, and board types can include regular posts, Q&A, or live chat. It's like a hybrid of Reddit and Discord but leans more towards a traditional web community. It also supports server-side rendering, making it SEO-friendly. This project is an extension of my previous Hacker News clone, dizkaz (https://news.ycombinator.com/item?id=43885998).