Notices by Alexey Yerin (yyp@fosstodon.org)

@lanodan From my understanding, most RISC-V specifications (including SBI) don't expect any kind of input devices or special shortcuts.

So any kind of SysRQ-magic is just using the default input driver from the kernel. This is what Linux does at least.

It's possible that a device has an interrupt source for this use-case - check the device tree. Though I'm not very familiar with RV hardware, no idea if there exists a board with such a feature.

Had some nightmares doing bit manipulation in Python. That was fun, for some definition of fun.

I accidentally read two bytes as signed instead of unsigned, then bitshifted them to get a relative offset for RLE decoding. This meant that the result was the correct length but completely wrong contents.

Also, for some reason it seemed to gravitate towards the octet 0xBA, filling lots of places with it for no apparent reason.

That was the entire morning of staring at the code and vbindiff.

It is a bit of a stupid situation when I misclick on a fedi poll and then realize that you can't retract your vote here :/

"it's not very stable but that's what you get for pc support"

- Me, 2024, apparently

Oh, looks like there isn't an SQLite library for Java that fits my requirements

Fine, I'll do it myself

(this already happened with gettext)
(it feels like every time I try to work on this project some technical limitation always bites me)

Tiers of C boolean types:
1. bool
2. _Bool
3. gboolean
4. BOOLEAN_T

Not sure what "export control regulations" is this referring to. As far as I understand those only cover physical goods or encryption tech (in some jurisdictions). This is neither of those!

In fact, serving that error message would have probably been a violation of those regulations. Sending an error page is no different from sending the plugin (which is probably just a .jar with metadata).

In fact, Oracle had a similar geoblock a while ago but they quickly reverted it, for some reason

I love international politics

why do text editors need discord rich presence all of a sudden

stop

(/s, probably)

Summer over!

No longer overheating

@lanodan I think Rust's stdlib only uses POSIX APIs and not general C things like stdio.h, so this doesn't really apply here. Hare also does the same thing on OpenBSD

@lanodan Even without libc, the language itself has a lot of really nasty footguns especially relating to integer promotion and the infamous signed overflow

"C is safe if you just program well" and other fairy tales

Q: What do astronomers and Haskell programmers have in common?

A: They both use parsecs

My opinion on Hare multithreading meta is slowly going from "mildly excited" to "can we not", TBH

@vyivel ext-hyper-extended-screen-transfer-v1 am i right

This was certainly a day.

Found two compiler bugs with two silly typos, somehow

For quite some time I've been thinking about making a "better Java" so a language that would target JVM and be fully compatible with existing Java code but "better". I envisioned null safety, more consistent pattern matching, maybe more expression-oriented syntax, inverted access flags (everything is private final by default) but now that I think about it, is it even useful? Everything other than null safety is just pointless fluff that's "good to have" but that doesn't warrant a new language

Why is it always that functional or functional-adjacent programming languages have the weirdest function names ever

like Haskell has nub and unsnoc (so un is an inversion prefix and snoc is reverse of cons????)

Scheme has car, cdr, current-jiffy and jiffies-per-second

Hardest problems in computer science:
1. Caching
2. Uh, off-by-one errors
3. Getting slicing to work in harec's eval are you kidding me it's still broken I need to rewrite more stuff help