Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ashishb
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
19 ms
·
301.
▲
by
ashishb
2y ago
I wish we can have immutable runtime objects (e.g. final in Java or const in C++). Go has constant (`const`) type that can be evaluated at compile-time but no const type for something that can only be evaluated at Run time.
302.
▲
Gabo: GitHub Actions Boilerplate Generator
(github.com)
3 points
by
ashishb
2y ago
|
1 comments
303.
▲
by
ashishb
2y ago
GitHub Actions Boilerplate generator that generates high-quality, narrow (path filters), secure actions (with minimal permissions for GITHUB_TOKEN).
304.
▲
by
ashishb
2y ago
If a vision has not been executed despite so many people trying may be something is wrong with it?
305.
▲
by
ashishb
2y ago
How many frontend (full stack) guys even understand the difference?
306.
▲
by
ashishb
2y ago
How many startups/small companies that uses next.js do any static verification?
307.
▲
by
ashishb
2y ago
Next.js is based on a fundamentally flawed premise that one can write code that runs in the browser as well as the backend. The security posture for the code running in the browser is very different from the code running on a trusted backen
308.
▲
by
ashishb
2y ago
GitHub Actions, at least at that time, had a lot more coverage than circle CI. Anda a lot more flexibility in terms of what can you do.
309.
▲
by
ashishb
2y ago
Will it last another 2 years? Several CI systems have come and gone.
310.
▲
by
ashishb
2y ago
I used to use circle CI earlier. GitHub Actions was an upgrade for me.
311.
▲
by
ashishb
2y ago
> There are better solutions out there. And what are those?
312.
▲
by
ashishb
2y ago
> You also need to block write access, so they can’t encrypt all your files with an embedded public key. And read access so they can’t use a timing side channel to read a sensitive file and pass that info to another process with internet
313.
▲
by
ashishb
2y ago
> Can you really say you trust all of these? We need better capabilities. E.g. when I run `fd`, `rg` or similar such tool, why should it have Internet access? IMHO, just eliminating Internet access for all tools (e.g. in a power mode), m
314.
▲
by
ashishb
2y ago
It definitely will. Whether it will make a noticable difference depends on how much non-cuda work happens in model training.
315.
▲
by
ashishb
2y ago
The article did talk about maintainability issues with Python as well.
316.
▲
by
ashishb
2y ago
Why do I have to restrict myself to interpreted languages only?
317.
▲
by
ashishb
2y ago
That's my experience as well.
318.
▲
by
ashishb
2y ago
> A typical Django-based application has a lot of non Python code that is running separately and the performance of Django or Python itself are mostly irrelevant. I have seen applications where this serving cost alone becomes dominant. A
319.
▲
by
ashishb
2y ago
Python alpine images are useless for any non-trivial project. > Author started comparing a debian image's size vs an alpine one. Immediately closed the article. I did not compare Debian to Alpine. I compared Debian to a scratch imag
320.
▲
by
ashishb
2y ago
Which tools would you recommend for that?
321.
▲
by
ashishb
2y ago
Indeed. Pure data manipulation, especially, when it being done for a batch job, is a good use case for Python.
322.
▲
by
ashishb
2y ago
> But I have never really worked with pythons type extension so nowadays even this might be good enough. I have used Python type system extensively, it is unfortunately, still an honor-code based system. So, you have to catch it during
323.
▲
by
ashishb
2y ago
Yeah, you can write Python in production. I, myself, have done that a lot. I don't think it is my preferred choice in 2025 though.
324.
▲
by
ashishb
2y ago
> C is less safe than either of python or Rust and has no benefits against Rust. Indeed.
325.
▲
by
ashishb
2y ago
> Are all the AI/ML engineers just being stupid because their notebook could be faster if it was written in Rust I think Python is a great language for running inside Notebooks. Running non-AI/ML jobs in production is where Pyt
326.
▲
by
ashishb
2y ago
> and avoid making any decisions using this chart. I did not mention that chart at all in the blog post. I just used this as a nice illustration.
327.
▲
by
ashishb
2y ago
> "Oh the Go dockerfile is smaller" Yes you compiled it and is running a binary. Big woop. Funny how the author not doing the same comparison with node If I did that with Node, you would say what about Ruby. If I did that with
328.
▲
by
ashishb
2y ago
> The only thing this tells you that you should try to avoid a billon nested loop iterations in python. In case you didnt know. Where does the article talk about the billion nested loop at all? It uses that as an illustration. I mentione
329.
▲
by
ashishb
2y ago
> Depressing, since I used to like Python I used to love Python. In fact, I still maintain a fairly popular Android development tool written in Python https://github.com/ashishb/adb-enhanced Over time, my love has h
330.
▲
It is hard to recommend Python in production
(ashishb.net)
33 points
by
ashishb
2y ago
|
37 comments
More ›