mattashii

9 ms·

61.
by mattashii 2y ago
> Our fundamental unit of distance (the meter) is defined in terms of the speed of light, so the radius will stay exactly the same, in meters. Why so? The Schwarzschild radius formula (r(s) = 2 * G * M / c²) uses c² as dividing cons
62.
by mattashii 2y ago
Wouldn't a different speed of light impact the Schwarzschild radius of black holes of a given mass? Assuming that you can create a standard clock, and given a black hole of standard mass, you can then measure speed of light in black ho
63.
by mattashii 2y ago
> Uber is an American company with all systems running in the US The company registration with headquarters in the Netherlands begs to differ, as do all those European Uber drivers
64.
by mattashii 2y ago
In most cases those banners are not even compiant, so "malicious non-compliance" is generally speaking more accurate.
65.
by mattashii 2y ago
Article from the Dutch DPA: https://www.autoriteitpersoonsgegevens.nl/actueel/ap-legt-ub...
66.
Uber fined €290M for data transfers outside the EU(cnil.fr)
6 points by mattashii 2y ago|1 comments
67.
by mattashii 2y ago
> Vanilla SQL [...] QUALIFY is after analytic window functions Isn't that FILTER (WHERE), as in SELECT avg(...) FILTER (WHERE ...) FROM ...?
68.
by mattashii 2y ago
Not all are assigned to Kuiper or Starliner; one has also been assigned to Viasat, which is scheduled to launch sometime in the first half of 2025.
69.
by mattashii 2y ago
> A redo of this mission would cost Boeing half a billion dollars, easy. I imagine so indeed, not in the least because all Atlas V launch vehicles are already assigned to missions. The booster for another non-operational flight would thu
70.
Page Weight Matters (2012)(blog.chriszacharias.com)
1 points by mattashii 2y ago|0 comments
71.
by mattashii 2y ago
Disk redundancy won't help guarantee torn page protection if the writes across the redundant disks are not coordinated to have one start after the other finishes such that there is always have one copy of the page that is not currently
72.
by mattashii 2y ago
> EDB ... or any of the other companies that provide support and actively contribute to the PostgreSQL project, such as Cybertec, Data Egret, Crunchy Data, AWS, ... I wouldn't choose any PostgreSQL supplier for their similarity to O
73.
by mattashii 2y ago
Advertising a product with features that it doesn't have remains a scam in my book, no matter the popularity. See also: Theranos
74.
by mattashii 2y ago
> As I just learned, apparently you can make HTTP requests from postgres like so: Well, it's more "you can execute arbitrary commands on the host node, which happens to include curl if it's installed". Which is also w
75.
by mattashii 2y ago
I think it is less of a service than a disservice (I certainly didn't ask, want, or expect misinformation in my search engine results), so BSaaDS would be more appropriate.
76.
by mattashii 2y ago
> and certainly not on Stack Overflow where you agreed to have everything under the CC license Does OpenAI comply with the CC-BY-SA 4.0 license terms? AFAIK (I don't use their products, so don't know much), they don't actu
77.
by mattashii 2y ago
> Is there a particular reason PostgreSQL cannot set up column migration itself out-of-the-box? People haven't asked hard enough to the right people, I suppose. PostgreSQL is an open-source project, where wanted and needed features
78.
by mattashii 2y ago
> Confused why people do this in the first place. Because you lose a significant amount of performance if you start adding NULL and variable-length columns just because you're afraid of a table rewrite. Because the resulting table w
79.
by mattashii 3y ago
Thank you for the confidence and praise! I do believe we, Neon, try to be as open as possible while also trying to build a business around this open source product. I'm a proud PostgreSQL contributor myself and value the work of my col
80.
by mattashii 3y ago
> Calling it "branches" does seem misleading. Branching in Neon should be interpreted more like the branches seen in graph theory's trees rather than the featureset exposed by git: the whole history of a Neon project is a
81.
by mattashii 3y ago
(PG Hacker @ Neon) Stas Kelvich one of my bosses, and one of the founders of Neon.
82.
by mattashii 3y ago
(employee) Neon won't pull the rug from under an active production database by scaling to 0 when there are still active queries, instead it'll scale databases to 0 only after a longer period of inactivity. So production databases
83.
by mattashii 3y ago
(Postgres hacker @ Neon) PostgreSQL writes changes to disk using WAL for consistency. Every WAL record is a set of changes to the PostgreSQL data directory (data files, metadata, ...) that need to be persisted together. Neon indexes this WA
84.
Waiting for PostgreSQL 17 – Invent Serialize Option for Explain(depesz.com)
3 points by mattashii 3y ago|0 comments
85.
by mattashii 3y ago
> Considering that embedding vectors represent a lossy compression of the original text or images, is achieving a 100% recall necessary? For the app, maybe not. But as a database absolutist, I think you must be able to dump all rows of a
86.
by mattashii 3y ago
> SQL vector database Cool, but does it actually return the correct results for these SQL statements, especially when ORDER BY is concerned? I.e. does it somehow have a way to get a recall of 100% from its indexes?
87.
by mattashii 3y ago
It depends on what you count, as "agriculture exports" could (probably do) also include, apart from food items: - Agricultural machines - Seeds - Agricultural patents on e.g. seeds - Fertilizers - non-food plants (flowers etc.) -
88.
by mattashii 3y ago
> If SQL weren't so (needlessly) complex we would see much more competition across the database space. I think there is more competition across the database space now than back when the SQL spec was less complex (say, in 1989 with S
89.
by mattashii 3y ago
Sure, but as the linked page mentions, QuickJS is 210KiB, which is seemingly larger than "in the order of a few kilobytes" claimed by the subject's blog post.
90.
by mattashii 3y ago
> This also means that its size is tiny, in the order of a few kilobytes in contrast to the 20+MB required by its counterparts. A full, performant ES2023 runtime, in "the order of a few kilobytes"? Even when you remove all JIT

More ›