Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
orf
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
31 ms
·
451.
▲
by
orf
2y ago
> They might as well be two different types The problem is they are different types, which has huge downstream impacts. Not least of all would be subclassing. The static analysis you’ve described could work in simple and trivial case
452.
▲
by
orf
2y ago
I can’t see how it would ever be possible with Python to do this. You depend on two packages, each with a function that returns a “requests.Request” object. These packages depend on different versions of “requests”. How would you implement
453.
▲
Show HN: Pytest-scrutinize – find bottlenecks in your test suite
(github.com)
7 points
by
orf
2y ago
|
0 comments
454.
▲
by
orf
2y ago
Yes, you obviously can fit 10 TB of data onto a developer laptop with 10tb of external storage. However you’ll run out of disk before you run into Postgres issues.
455.
▲
by
orf
2y ago
Yes, you obviously can’t fit 10 TB of data onto a developer laptop. However you’ll run out of disk before you run into Postgres issues.
456.
▲
by
orf
2y ago
> This is exactly a great point. When data size goes to a billion rows, Postgres is tough. You’ve been led astray. You can handle a billion rows on a developer laptop, let alone a production grade instance.
457.
▲
by
orf
2y ago
You missed the part of your comment that explains how flatpack is anything close to what the author described, outside of basic sandboxing of often outdated applications with spotty OS integration and poor support in the ecosystem.
458.
▲
by
orf
2y ago
> I want most applications on my Mac to run within a sandbox. > I want applications to require explicit permission to access the camera and microphone, or to capture the content of my display. > I want applications to be cryptogra
459.
▲
by
orf
2y ago
Browser version guards[1] seem a bit scary: it just does basic user-agent parsing rather than actual feature detection, which is not exactly a great thing for the web as a whole. 1. https://edgeguides.rubyonrails.org/7_2_rel
460.
▲
by
orf
2y ago
Let me get this straight: you stay at a hotel, where you blindly agree that they can charge your card “to pay for damages” but you don’t sign anything that specifies what constitutes damage and what the charges may be for that damage? You’d
461.
▲
by
orf
2y ago
Seeing as you’re in the thread Dmitry, what was the Kindle bug referenced here? Sounds interesting. https://news.ycombinator.com/item?id=41207740
462.
▲
by
orf
2y ago
That’s a strawman argument: nobody seriously considers a different language because of a “fancy min/max”. However, hygienic macros that fix entire classes of issues (including this) is a more compelling argument. Not to say that it mak
463.
▲
by
orf
2y ago
This is of course one specific point out of the 10 listed points, and not the sole reason that the ban was implemented.
464.
▲
by
orf
2y ago
The thread is here: https://discuss.python.org/t/for-your-consideration-proposed... The person in question is Tim Peters. It’s weird to not name them, but not linking to the thread is appropriate.
465.
▲
by
orf
2y ago
You can likely do this without the GitHub API, with just the git protocol itself. Just ask for the commits and see what the server says.
466.
▲
by
orf
2y ago
Someone should do this for Hacker News profiles
467.
▲
by
orf
2y ago
Man attempts to write constant time algorithms using language that does not support constant time algorithms, but who’s really at fault here? Find out on next weeks episode of “lets blame compilers rather than my choice of language”!
468.
▲
by
orf
2y ago
Why do they make it so difficult to find a code sample that concisely shows what this is and why you should use it? I’ve visited about 15 links and found some go code[1] that doesn’t really make it clear. It’s like Terraform, but for Window
469.
▲
by
orf
2y ago
If we take WASM as an example, you could strictly bound the number of instructions executed for a given request in order to keep things performant. I’m not clear on what “bad things” they wheee imagining though?
470.
▲
by
orf
2y ago
I don’t get why we need only a declarative manifest format to specify blocking rules. Chrome has the ability to run code in a sandboxed environment with no external network access. This could be done in JS, or it could be WASM. So, why can’
471.
▲
by
orf
2y ago
> It correctly handles the strange array/map duality of lua tables in the most efficient way it can do safely (serializing both variants concurrently and only writing out the correct one at the end) The article doesn’t expand on thi
472.
▲
by
orf
2y ago
aws-security@amazon.com - it’s very clearly the first result when you search “AWS security report”.
473.
▲
by
orf
2y ago
I would just reach out to AWS directly: why go through hacker one? They have a direct email and are responsive. If the issue meets their criteria then you get a payout.
474.
▲
by
orf
2y ago
Yes, but generally you don’t give SQL access to the internet or completely unauthenticated users. Any user could likely cause issues for Redshift by running completely nuts queries that exhaust all the servers resources. Is “shit SQL” a bou
475.
▲
by
orf
2y ago
Why?
476.
▲
by
orf
2y ago
Which models, specifically, said we would be under water by now?
477.
▲
by
orf
2y ago
If the size of the data is small, a linear search through a contiguous array is going to be far faster than anything more complex.
478.
▲
Google abandons plan to remove cookies from Chrome browser
(ft.com)
17 points
by
orf
2y ago
|
2 comments
479.
▲
by
orf
2y ago
I think “Pin” is a good example of a technically correct but hard to understand name. “Drop” has a much more familiar meaning because it’s a common action, whereas “pinning” isn’t so common and can mean different things in different context
480.
▲
by
orf
2y ago
You’re not displaying a pattern of healthy behavior by creating numerous new accounts to try and provoke an argument on such a stupid point, without contributing anything of substance to the discussion.
More ›