ncruces

20 ms·

361.
by ncruces 1y ago
The Litestream read replicas they announced, just for my Go SQLite driver. They mostly work already, would appreciate testing from anyone who already has a larger, real-world Litestream v0.5.0 setup running. https://fly.io/b
362.
by ncruces 1y ago
My project is really healthy then, as I summarily close issues as “not planned” rather than leave them open.
363.
by ncruces 1y ago
They should. But also, the easier you make it, the more social pressure they'll feel to just do it.
364.
by ncruces 1y ago
None of my quotes should scare you. What you might call progressives, liberals, centrists, whatever, varies widely across the EU. That's the only reason I've put those terms in quotes.
365.
by ncruces 1y ago
In the past half century, “progressives” have pushed through huge societal advances ( particularly ) across Europe in areas of family rights, women's rights, gay rights, minority rights, worker's rights. Things that were anathema
366.
by ncruces 1y ago
> Or writing some data science stuff, it might be like: load data from disk or network for 6 seconds, run Numpy on it for 3 hours, write it back out for 3 seconds. > You could rewrite that in Rust and it wouldn't be any faster
367.
by ncruces 1y ago
Very likely the issue is with the test infrastructure and not with the code under test. My own fuzzing doesn't report any inconsistencies. But fuzzing is always necessarily incomplete.
368.
by ncruces 1y ago
> I am sure I could have phrased it better. I'm not a native speaker, but this sounds/reads awful: … but I can’t continue this conversation. I'd definitely swap that out for (or similar): … but I don't wanna go the
369.
by ncruces 1y ago
I merged a few PRs to SIMD optimize Wasm WASI libc, but it all got stalled in str(c)spn (which is slightly more sophisticated than the rest). There wasn't much appetite for any of it on Emscripten. https://github.com/We
370.
by ncruces 1y ago
This is not a left-right issue. You have the danish S&Ds pushing it and ALDE opposing it.
371.
by ncruces 1y ago
My version of the VFS should be ready for testing. After some testing, I expect it to be working correctly in the vast majority of cases, but the performance may disappoint. The original still needs work. https://github.com/
372.
by ncruces 1y ago
Sure. Using a cloud can make that more convenient. But obviously not so if you then keep all your data in the same region, or even “availability-zone” (which seems to be the case for the all “lost to lightening strikes” data here).
373.
by ncruces 1y ago
“The BBC understands that customers, through various backup technologies, external, were able to recover all lost data.” You backup stuff. To other regions.
374.
by ncruces 1y ago
PostgreSQL is probably smarter than SQLite. :) Anyway, at minimum you have prepared statements where a given plan may be great for one execution and terrible for the next. Maybe the database re-prepares for you, and then it's the same
375.
by ncruces 1y ago
The problem with sqlc is dynamic queries (mostly dynamic filters and ordering, which don't really affect type safety of the parameter or result-set mappings). Even after years, the solutions are unsatisfactory from a database query pla
376.
by ncruces 1y ago
Sure. But you'll be transferring the entire file everytime. The advantage of sqlite3_rsync is that, if on the other end is an SSH server with sqlite3_rsync, you only transfer the changed pages, and still get a “perfect” copy of the fil
377.
by ncruces 1y ago
The other advantage (or difference) is that you don't need a “server” on the other end, just object storage. Which may come out cheaper.
378.
by ncruces 1y ago
I… honestly don't know. I saw the announcement some time ago (the revamped post) and started following their repo. They did mention caching to hide latency. When I saw the v0.5.0 tag, I dived into just porting it. It's just over a
379.
by ncruces 1y ago
A VFS that allows you to directly open a Litestream replication target (e.g. S3) as a read-only database and run queries against it without ever having to download the entire database (e.g. to an ephemeral instance that doesn't even ha
380.
by ncruces 1y ago
I mean, that's literally their "What’s next?" from the OP: https://fly.io/blog/litestream-v050-is-here/#whats-next They already have a prototype, and... it's pretty rough on the edges. I'm
381.
by ncruces 1y ago
But would a destructor that runs automagically when a value goes out of scope, even if it's not properly initialized (and with zero regards for copies or moves) be in any way better than a defer that's explicitly called after it
382.
by ncruces 1y ago
Seems to be for weather going from current to future? IDK, TBH. https://github.com/keshikan/DSEG?tab=readme-ov-file#sample
383.
by ncruces 1y ago
There's no way to write a "destructor" defensively, if the contents of memory it is trying to "destroy" are undefined. Whereas it's perfectly possible to only defer a statement when you know the "object&
384.
by ncruces 1y ago
I encourage you to read ( at least ) this section of this blog before making simplistic suggestions: https://thephd.dev/just-put-raii-in-c-bro-please-bro-just-on... How do you mandate initialization, handle copies, move o
385.
by ncruces 1y ago
I'm trying to port the "read-replica" VFS to my Go driver. I don't think that bit is ready yet, but it may be close.
386.
by ncruces 1y ago
Rust mandates that every field in a user-defined type is initialized at once. How do you propose to retrofit that into C without "constructors"?
387.
by ncruces 1y ago
> I don't understand why people insist on simulating a poor substitute for RAII with a feature that is itself almost decent RAII. RAII doesn't make sense without initialization. Are you proposing C should add constructors, or t
388.
by ncruces 1y ago
They couldn't even resist adding a "just think of the children" in their argument.
389.
by ncruces 1y ago
I used it to pick the constants that minimized error in an approximation of inverse trigonometric functions (-π to +π). Found that it provided a good balance between minimizing either the relative or absolute error.
390.
Hyb Error: A Hybrid Metric Combining Absolute and Relative Errors (2024)(arxiv.org)
19 points by ncruces 1y ago|2 comments

More ›