Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ncruces
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
26 ms
·
391.
▲
by
ncruces
1y ago
Yeah, in Portugal the highway operator would be liable for that accident. They'd be liable if even if it was road kill, as they're responsible for ensuring big animals don't go into roads.
392.
▲
by
ncruces
1y ago
Me too. There's only so much time in life. I'd take it for a spin if someone else did it. :)
393.
▲
by
ncruces
1y ago
I try to stick to GitHub for GitHub CI downloads. E.g. my SQLite project downloads code from the GitHub mirror rather than Fossil.
394.
▲
by
ncruces
1y ago
Sure. I don't password protect my (Android) TV like I password protect my (Android) phone, despite both of them allowing authorized access to the same Google accounts, because if someone entered my house I have bigger things to worry t
395.
▲
by
ncruces
1y ago
Redka may be interesting to you: https://github.com/nalgeon/redka
396.
▲
by
ncruces
1y ago
It's great that you did that due diligence once. It really is. If I were reviewing that merge request, and you told me that, I'd be inclined to approve. But how do we scale that to 1000 dependencies, and every one of their updates
397.
▲
by
ncruces
1y ago
Right. Allowing 500 strangers to push code to our CI infra, or developer laptops, with approximately zero review, sounds similarly ill advised. That JLR got their factories hacked, rather than customer cars , is less bad for sure. But it
398.
▲
by
ncruces
1y ago
Just the other day someone argued with me that it was reasonable for Limbo (the SQLite Rust rewrite) to have 3135 dependencies (of those, 1313 Rust dependencies). https://github.com/tursodatabase/turso/network/
399.
▲
by
ncruces
1y ago
This is a culture issue with developers who find it OK to have hundreds of (transitive) dependencies, and then follow processes that, for all intents and purposes, blindly auto update them, thereby giving hundreds of third-parties access to
400.
▲
by
ncruces
1y ago
It protects less proficient users from accidentally taking a screenshot.
401.
▲
by
ncruces
1y ago
It's going to be that it was always all about the bottom line, and only the bottom line.
402.
▲
by
ncruces
1y ago
Politicians around the world do it on purpose because they know they can more easily get away with leaving no trace. It's not an accident they don't use government email/IM and use WhatsApp/Signal instead. But then the
403.
▲
by
ncruces
1y ago
Yes. It acts exactly like WAL mode in that regard. Readers operate on a snapshot and are never blocked, writers block each other. Upgrading a reader to a writer fails immediately if there is a concurrent writer, however BEGIN IMMEDIATE tran
404.
▲
by
ncruces
1y ago
The way they got multi-writer working is a bit too adventurous for me. It involves fiddling the database header to ensure you can keep track of the read-set from the VFS layer, by convincing SQLite to drop its page cache at the start of eve
405.
▲
by
ncruces
1y ago
Because it allows streaming compression.
406.
▲
by
ncruces
1y ago
I hadn't visited Madrid in over 20 years, and found the burial of the M30 (and, increasingly, the radial highways) extremely impressive. https://en.wikipedia.org/wiki/Madrid_Río There's also the subway (impre
407.
▲
by
ncruces
1y ago
That has its own downsides, though. Both v1 packages continue work; both are maintained. They get security updates, and were both improved by implementing them on top of v2 to the extent possible without breaking their respective APIs. More
408.
▲
by
ncruces
1y ago
The latency on those requests matters, though. You'll probably benefit from using the largest possible page size; also, keep alive; etc. But even then, you'll pull at most 64 KiB per request. If you managed to have response times
409.
▲
by
ncruces
1y ago
> The evidence confirms that the emergency system installed in the flywheel located at the top of Calçada da Glória, tripping the power switch to the cabins in the event of a loss of cable tension, worked as intended, which would resu
410.
▲
by
ncruces
1y ago
Most of the recommended [1] setting are available on a per connection basis, through PRAGMAs, sqlite3_db_config, sqlite3_limit, etc; some are global settings, like sqlite3_hard_heap_limit64. A binding can expose those settings. It's no
411.
▲
by
ncruces
1y ago
Untrusted database file is not the same as untrusted SQL input. There are parts of the SQL engine that are exposed to malicious file manipulation (the schema is stored as SQL DDL text) but that's not arbitrary SQL input. If you want
412.
▲
by
ncruces
1y ago
Are their competitors more energy efficient somehow? This is a thread about them not being able to meet carbon neutrality commitments because of AI, which they feel they need to because competitors are eating their lunch. Are the competitor
413.
▲
by
ncruces
1y ago
> It's still early stages, that is why. Were we advised to check compiler output every single time "in the early days"? No, that's not the difference. A compiler from whatever high/low level language is expected
414.
▲
by
ncruces
1y ago
In Wasm you can read/write whatever to address zero of linear memory. It's still UB as far as clang is concerned so you C code can do whatever. But it won't “crash” on the spot.
415.
▲
by
ncruces
1y ago
People are willing to waste energy to avoid drinking their own (treated) sewage, before it's temporarily mixed in the ocean.
416.
▲
by
ncruces
1y ago
> Seems like the ideal thing is to spawn a goroutine and run optimize periodically. Many Go daemons I run will have uptime measured in months so running on shutdown isn't super useful there. Connections don't necessarily need
417.
▲
by
ncruces
1y ago
That doesn't really escape the gravity well, does it?
418.
▲
by
ncruces
1y ago
Now try to escape it.
419.
▲
by
ncruces
1y ago
In Go, a database/sql “connection” is actually a pool, and Go makes sure that it only calls driver methods serially for an actual driver connection from a single goroutine. So your point (which is not very clear to me, with my limited
420.
▲
by
ncruces
1y ago
To do this, it's best if the user opens two pools themselves. I've tried to create something that does this transparently (give you a pool that magically routes writes to a single connection), but it's hard. It starts simple:
More ›