Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ogogmad
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
9 ms
·
151.
▲
by
ogogmad
10mo ago
Are they Venezuelans living in Venezuela? I think the ones you have to worry about are the ones still living there. Additionally, might it be that every dictatorship is hated by most expatriates? I think that that was the case for the 2 (
152.
▲
by
ogogmad
10mo ago
I get why some people were neo-con the first 3 or so times (Afghanistan, Iraq, Libya) but it's criminal not to learn after failing 3 times over. I want the most severe consequences for the people who have enabled this to happen again.
153.
▲
by
ogogmad
10mo ago
I think they should have used Bastard. Here's a sample: https://museumcrush.org/wp-content/uploads/2018/05/letter-on... RETVRN
154.
▲
by
ogogmad
10mo ago
No. People would prefer the equivalent of double-click `setup.exe`. Were you being serious?
155.
▲
by
ogogmad
10mo ago
Decreasing market share in the desktop ?
156.
▲
by
ogogmad
10mo ago
Doesn't explain why Chrome beat IE. Or why MacOS has higher market share on the desktop than Linux. Wine and Proton should have levelled the playing field. But they haven't. Also, if you've only just started using Linux, I re
157.
▲
by
ogogmad
10mo ago
Because verifying something is easier than finding it in the first place. It's in some way the difference between P and NP.
158.
▲
by
ogogmad
10mo ago
This might be why OpenBSD looks attractive to some. Its kernel and all the different applications are fully integrated with each other -- no distros! It also tries to be simple, I believe, which makes it more secure and overall less buggy.
159.
▲
by
ogogmad
10mo ago
I think LLMs lower the barrier for learning certain topics. For instance, without them I wouldn't have tried to learn about RLC circuits at all.
160.
▲
by
ogogmad
10mo ago
Because you can independently check anything it tells you. You understand there can be independent sources of validation?
161.
▲
by
ogogmad
10mo ago
> It’s somewhat delusional and potentially dangerous to assume that chatting with an LLM about a specific topic is self-teaching beyond the most surface-level understanding of a topic It's delusional and very arrogant of you to conf
162.
▲
by
ogogmad
10mo ago
This might be a silly thing to point out, but where do people draw the line between an allocation happening or not happening? You still need to track vacant/occupied memory even when there's no OS or other programs around. It'
163.
▲
by
ogogmad
10mo ago
> This is sort of why I think software development might be the only real application of LLMs outside of entertainment. Wow. What about also, I don't know, self-teaching*? In general, you have to be very arrogant to say that you&#
164.
▲
by
ogogmad
10mo ago
Thanks. Corrected.
165.
▲
by
ogogmad
10mo ago
This doesn't have the distracting long "ſ"es of Oliver Byrne's editions. It was arguably a bonkers decision to have it in the first place, given that the long S was on the way out by then. [EDIT: Corrected to use his act
166.
▲
Oliver Byrne's "The Elements of Euclid", First Six Books [pdf]
(proyectodescartes.org)
9 points
by
ogogmad
10mo ago
|
4 comments
167.
▲
by
ogogmad
10mo ago
Isn't a more modest opinion that gold admits a guaranteed minimum on its price because of scarcity and demand from industrial applications? That guaranteed minimum on its price is what enables it to be used as a hedge against a pos
168.
▲
by
ogogmad
10mo ago
That's an INCREDIBLY good point about synthetic training data. During model training, AI agents could pretty much start their own coding projects, based on AI-generated wish-lists of features, and verify progress on their own. This lea
169.
▲
by
ogogmad
10mo ago
`forward_shuffle` is the (GROUP) INVERSE of `fisher_yates_shuffle`. `mirror_shuffle` is the (GROUP) CONJUGATE of `fisher_yates_shuffle` by the cyclic permutation (n-1,n-2,...,1,0). In group theory, CONJUGATEs are like changes of coordinates
170.
▲
by
ogogmad
10mo ago
There are still some pain points with Linux distros: With some, an upgrade can leave you unable to boot into a graphical login screen. This can also happen if you leave a Linux installation, like Manjaro, alone for a year and then do an upd
171.
▲
by
ogogmad
10mo ago
Maybe then the year of Linux (or OpenBSD?) on the desktop would finally arrive. Maybe anti-trust could get used. Maybe parts could get scrapped from data centres. Interesting times they would be!
172.
▲
by
ogogmad
10mo ago
That might be a bit on the paranoid side. It could just be that it's far more profitable right now for companies to sell only to data centres. That way, they don't need to spend money on advertising, or share their revenues with t
173.
▲
by
ogogmad
10mo ago
I don't think SQL is "perfect" and I'm not sure it's rational to even be saying that. For instance, why is it that the syntax for an SQL query is "select A from B" when many SQL-inspired syntaxes have swit
174.
▲
by
ogogmad
10mo ago
Actually, I've now checked more thoroughly, and RE+complement matching is O(n^2 m), which is not that good.
175.
▲
by
ogogmad
10mo ago
Just checked after reading your comment. Surprisingly to me, AFAs (Alternating Finite Automatons) do let you introduce the Complement operation into Regex while preserving the O(mn) complexity of running NFAs. That's really subtle, b
176.
▲
by
ogogmad
10mo ago
Don't you think people would have figured it out by themselves the moment programmers started writing parsers? I'm not sure his contribution was particularly needed.
177.
▲
by
ogogmad
10mo ago
:/ Damn. That needs to cost 1000x less before people can try it on their own games.
178.
▲
by
ogogmad
10mo ago
I think most PL parsers use recursive descent, which doesn't require any formal language theory. That said, it's possible that formal language theory has enabled regular expressions to enter widespread use - first via theory; th
179.
▲
by
ogogmad
10mo ago
> Are they the _only_ flavour of curve that has a nice geometric group law? For affine conics over the real numbers, the non-degenerate ones are ellipses (affine transform to complex unit circle), hyperbolas (affine transform to y=1/
180.
▲
by
ogogmad
10mo ago
The following is one of my favourite submissions to HN: https://news.ycombinator.com/item?id=44083753
More ›