HackerTrans
TopNewTrendsCommentsPastAskShowJobs

astronautas

no profile record

Submissions

[untitled]

1 points·by astronautas·8 mesi fa·0 comments

Declarative database schema migrations – yay or nay?

1 points·by astronautas·9 mesi fa·4 comments

[untitled]

1 points·by astronautas·9 mesi fa·0 comments

[untitled]

1 points·by astronautas·9 mesi fa·0 comments

[untitled]

1 points·by astronautas·11 mesi fa·0 comments

[untitled]

1 points·by astronautas·11 mesi fa·0 comments

[untitled]

1 points·by astronautas·11 mesi fa·0 comments

Should we replace Spark with DuckDB?

valatka.dev
6 points·by astronautas·anno scorso·0 comments

FOSDEM 2025 has exceeded my expectations

valatka.dev
20 points·by astronautas·anno scorso·0 comments

Wheel - a surprisingly good model package format

valatka.dev
2 points·by astronautas·anno scorso·0 comments

Uv's killer feature is making ad-hoc environments easy

valatka.dev
502 points·by astronautas·2 anni fa·417 comments

Where async-io cripples, and the promise of No-GIL

valatka.dev
2 points·by astronautas·2 anni fa·0 comments

Ask HN: Do you think Python will disappear for LLM inference?

2 points·by astronautas·2 anni fa·19 comments

Why I am excited about No-GIL (and skeptical about asyncio)

astronautas.github.io
3 points·by astronautas·2 anni fa·0 comments

Let's build a Bloom filter. Part 1

astronautas.github.io
1 points·by astronautas·3 anni fa·0 comments

comments

astronautas
·9 mesi fa·discuss
cool!
astronautas
·9 mesi fa·discuss
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?
astronautas
·anno scorso·discuss
Interesting, will check it out.
astronautas
·2 anni fa·discuss
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.
astronautas
·2 anni fa·discuss
see my reply above
astronautas
·2 anni fa·discuss
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.
astronautas
·2 anni fa·discuss
How about "A UV feature that intrigues me most"?
astronautas
·2 anni fa·discuss
uh, thanks I guess.
astronautas
·2 anni fa·discuss
Indeed, you're right ;).
astronautas
·2 anni fa·discuss
Neat!
astronautas
·2 anni fa·discuss
Fair points!
astronautas
·2 anni fa·discuss
call to external db, and then to llm with retrieved context.

also business rules, no?
astronautas
·2 anni fa·discuss
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.
astronautas
·2 anni fa·discuss
can't disagree (a tradeoff).
astronautas
·2 anni fa·discuss
agree, it depends, always benchmark, but my question is rather generic i.e. I am looking for a perspective.
astronautas
·2 anni fa·discuss
Sure, but what about non AI / business logic pre and post? Think RAG calls, guardrails, ...? Or do they fly compared to LLM inference itself?
astronautas
·2 anni fa·discuss
Good perspective. No-GIL should make things better (shared memory parallelism), but it's not bulletproof.
astronautas
·2 anni fa·discuss
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.
astronautas
·2 anni fa·discuss
Cool! Thanks for the investigation.
astronautas
·2 anni fa·discuss
Thanks, so it's connectors, nice differentiators. Seamless integrations are harder than it seems.