astronautas
astronautas on Hacker News
indeed, I was thinking about the same. How do you even escape the hatch with e.g. Atlas if your migrations folder is autogenerated? What if you modify /migrations but not your entities?
Posted by
astronautas on Monday October 27, 2025 @07:04AM
from the self-post dept.
Two issues I observed with typical migration as a separate new SQL file approach:* Views don't have alter, so have to be fully redefined. Any change to a view obscures actually changed parts of a view, since you have to fully redefine it.* Duplicate change tracking with git and migration history. Can't we just use git for history?E.g. https://atlasgo.io/. Totally non affiliated with them. Just...
Posted by
astronautas on Friday March 14, 2025 @04:05AM
from the valatka.dev dept.
Posted by
astronautas on Sunday February 02, 2025 @03:31PM
from the valatka.dev dept.
Posted by
astronautas on Tuesday January 28, 2025 @03:58PM
from the valatka.dev dept.
Interesting, will check it out.
Hey, I actually made a silly mistake in my post, indeed you first activate the environment and then install stuff in it. Fixed!
I disagree though it is activated immediately, or at least to me with venv I always have to activate it explicitly.
Sorry dang, didn't know the practice + got a bit emotional haha. I agree with the remark above, my message is rather on how easy it is to run python scripts with dependencies (without mutating the state.
How about "A UV feature that intrigues me most"?
Posted by
astronautas on Sunday January 12, 2025 @03:06PM
from the valatka.dev dept.
Posted by
astronautas on Wednesday January 08, 2025 @10:27AM
from the valatka.dev dept.
call to external db, and then to llm with retrieved context.
also business rules, no?
I agree Python is hardly replacable. Btw, I only mean inference, not training. Training imho should stay pure Python, you can achieve mega throughout with it for batch processing.
can't disagree (a tradeoff).
agree, it depends, always benchmark, but my question is rather generic i.e. I am looking for a perspective.
Sure, but what about non AI / business logic pre and post? Think RAG calls, guardrails, ...? Or do they fly compared to LLM inference itself?
Good perspective. No-GIL should make things better (shared memory parallelism), but it's not bulletproof.
Indeed, my point was that Go and Rust could lead optimizing the non-AI code, which often begs to be coupled with AI code (think guardrails).
Also, what's the benefit of Python then in this case? Ergonomically, Go isn't shabby, Rust is another story though.
Posted by
astronautas on Friday January 03, 2025 @11:04AM
from the ask-hn dept.
For non-LLMs, we have a high variety of frameworks with Python interfaces for running models, often built to call C bindings. This makes it impractical to run models directly from languages like Rust—you’d need to implement an idiomatic layer for each model runtime, something already done for Python. Nvidia's Triton covers a lot, but is it even designed for embedded use? And how feasible is...
Cool! Thanks for the investigation.
Posted by
astronautas on Wednesday January 01, 2025 @05:21AM
from the astronautas.github.io dept.
Thanks, so it's connectors, nice differentiators. Seamless integrations are harder than it seems.
How does this compare to Vespa? If the key difficulty in scaling search is infra as you say, Vespa is an interesting alternative.