Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
apetresc
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
17 ms
·
121.
▲
by
apetresc
2y ago
What is? Getting to decide how you can distribute copyrighted works by other individuals in other sovereign nations? I'm a sovereign individual, can I unilaterally decide that Disney's copyrights only last 1 year if I'm the o
122.
▲
by
apetresc
2y ago
I think you might be watching too much Succession.
123.
▲
by
apetresc
2y ago
How does this work? Asking legitimately. If an American author produces a work in America, with presumably an American copyright, don't USA laws decide when they enter the public domain? Like, could some country just decide global copy
124.
▲
Advent of Code Stats Analysis
(jvanelteren.github.io)
1 points
by
apetresc
2y ago
|
0 comments
125.
▲
by
apetresc
2y ago
Maybe, but is there a particular reason you need to use a ~20-year-abandoned IRC client when there's literally dozens of actively-maintained ones for any platform you could reasonably think of?
126.
▲
by
apetresc
2y ago
Just wanted to let you know I really thoroughly enjoyed this writeup, and I very much hope you get around to part 2 with input et al.
127.
▲
by
apetresc
2y ago
I'm not too well-versed in hardware - is it really that easy to swap out a 2G modem with a 4G modem and have it "just work" without touching the drivers? Even if the baseband/modem chips miraculously do conform perfectly
128.
▲
by
apetresc
2y ago
You don't believe hitmen exist?
129.
▲
by
apetresc
2y ago
Yeah that's somewhat of a special case - the Gemini API even has a specific CIVIC_INTEGRITY flag in its safety filters: https://ai.google.dev/gemini-api/docs/safety-settings . They literally put "election
130.
▲
by
apetresc
2y ago
Well, it's not just that - Blizzard is actively selling the non-remastered original versions (running on DOSBox exactly like the GoG ones) directly from the Battle.net launcher now too. They've essentially in-housed the exact same
131.
▲
Advent of Chess 2024
(adventofchess.com)
2 points
by
apetresc
2y ago
|
1 comments
132.
▲
by
apetresc
2y ago
Note that these aren't typical "mate in N" or tactics problems. They're all weird, interesting chess-related puzzles like "find a way to construct this position in n moves" or "find a way to mate with this
133.
▲
by
apetresc
2y ago
Very wrong. The first time most people here probably heard about OpenAI back in 2017 or so was their DotA 2 bot.
134.
▲
by
apetresc
2y ago
> 10. What Changes When You Are Free > You will not be forced to work inside something known as a “virtual environment.” Oof, this terrible advice cancels out an otherwise reasonable post. Beginners who don't know what they'
135.
▲
by
apetresc
2y ago
I'm a huge fan of Standard Ebooks, but it's pretty funny that twice in both the subject line and in the body of the e-mail they sent out announcing this, they typo-ed "1,000th" as "1,00th", given their whole ra
136.
▲
by
apetresc
2y ago
You're right that the first paragraph is bad - but it's uncharacteristically bad. Maybe the translator/editor was still warming up his workflow at the start and never went back to revise it for some reason. I think you'l
137.
▲
by
apetresc
2y ago
A private settlement of paying Nintendo $2.4M. How is that not a perfect example of "we all saw how that went"?
138.
▲
by
apetresc
2y ago
I know this isn't causing this particular incident, but what is it with the completely un-understandable speech cadence/articulation of virtually every ATC/pilot I've ever heard? Considering that they're already tal
139.
▲
by
apetresc
2y ago
Firstly, please let me know which rock you've been living under for the last ~8 years to not know what 'virtue signalling' means, it sounds awesome and I would like to join you under there. Anyway, the insinuation is that the
140.
▲
by
apetresc
3y ago
That's a valid theory, a priori, but if you actually follow up you'll find that the vast majority of these benchmark results don't end up matching anyone's subjective experience with the models. The churn at the top is n
141.
▲
by
apetresc
3y ago
Virtually every announcement of a new model release has some sort of table or graph matching it up against a bunch of other models on various benchmarks, and they're always selected in such a way that the newly-released model dominates
142.
▲
by
apetresc
3y ago
I just find it hilarious how approximately 100% of models beat all other models on benchmarks.
143.
▲
by
apetresc
3y ago
> In keeping with our commitment to open generative AI, we aim to make the model weights available for self-hosting with a Stability AI Membership in the near future. This is new, right? All previous releases as far as I can tell were ju
144.
▲
by
apetresc
3y ago
The window on being able to archive your favorite channels with yt-dlp is definitely closing fast.
145.
▲
by
apetresc
3y ago
Sorry, what's the pun? I can't find it.
146.
▲
by
apetresc
3y ago
The tl;dr seems to be that Apple pulled it because it's just a white-labeled build of Riley's (of AltStore fame) Delta/GBA4iOS, not because it's an emulator. The README on Riley's repo says: > Due to the licensin
147.
▲
by
apetresc
3y ago
I mean, the rumours at this point are deafening that Siri is getting a complete rehaul (or an outright replacement, which probably amounts to the same thing) at WWDC in just two months, so I'd reserve judgment on Apple's AI deploy
148.
▲
by
apetresc
3y ago
If Svelte templates really are a strict superset of HTML, then doesn't that mean by definition that there is "wrong HTML" that is valid Svelte? That would only not be true if Svelte was a sub -set.
149.
▲
by
apetresc
3y ago
That's very strange, I've never seen the font rendering go so wrong on Windows Firefox. My best guess is that something messed with your font library somewhere, and the browser's engine is picking up a conflicting font with t
150.
▲
by
apetresc
3y ago
Virtually every public Go server at the moment, unfortunately.
More ›