torginus

11 ms·

181.
by torginus 2mo ago
Yeah another kind of a*hole who I've encountered on Reddit were the ones who would keep mouthing off to you (carefully keeping within 'letter of the rules' of moderation), trying to goad you into snapping at them, and they&#x
182.
by torginus 2mo ago
Yeah, one extreme case of this was on Reddit, where people DM'd 'kill yourself' messages to others. When mods started banning people for this, attackers switched to abusing Reddit's mental health features, reporting peop
183.
by torginus 2mo ago
A bit of clarification: Thread is a layer 2 protocol, for the user, it looks like standard IPv6. Matter is application level, and just expect standard TCP/IP underneath. There's nothing binding Matter to Thread, and in fact a lot
184.
by torginus 2mo ago
Dunno, personally I have found HA + MoT to be strictly worse than Z2M. It's the classic case of the Douglas Adams quote: “The major difference between a thing that might go wrong and a thing that cannot possibly go wrong is that when a
185.
by torginus 2mo ago
This is both good and bad, and no surprise its an American standard. Lower frequencies have higher range, which is good if you live in a house (because range), and bad if you live in an apartment (because interference and crowded channels).
186.
by torginus 2mo ago
The whole idea of SPAs is just unfortunate most of the time. If you look at the list of the most popular websites: https://en.wikipedia.org/wiki/List_of_most-visited_websites Basically all of them can be decomposed to:
187.
by torginus 2mo ago
I know, that's why I said, that when you hit things like that (state that is too big to pull onto the client at once, and/or displayed at once), React's (and I guess a lot of other immutablity-based frameworks') dataflo
188.
by torginus 2mo ago
I am talking about virtualized lists. And it should be a framework feature. I used to use WPF on desktop, and it had pretty good virtualization support (though the framework in general was more like Angular) - most containers had virtualiza
189.
by torginus 2mo ago
> This makes me wonder if AI companies even have a MOAT in the first place. Generally speaking they do, at least from my experience when switching from one model to the other - their performance decreases, and they often do large refacto
190.
by torginus 2mo ago
And I'm not a python expert, so I might be wrong here, but: My understanding is that this decorator generates some class in the background that wraps the function into some remotely executable container thing, and handles the networki
191.
by torginus 2mo ago
Imo the biggest issue with this functional model (at least in React), is that it handles things like virtualization, async, etc. poorly. Which is kinda ironic, because in a true functional language, it'd be feasible to provide 'a
192.
by torginus 2mo ago
This is the kind of shit why they probably didn't want generics in the language. This is like building a very crude general-ish DSL inside the language. Because the tools are intentionally limited (as to limit the scope of the feature)
193.
by torginus 2mo ago
Dunno, a 128bit integer add is essentially 2x as expensive as a 64 bit one, a multiply is 4x. We've already had SSE in the early 2000s that could do computations like this in a single cycle (tho on multiple 32 bit numbers, not 128-bit
194.
by torginus 2mo ago
Not a finance person, but my understanding is that the US govt is indebted via bonds which pay according to their issued yields. So the interest rates are the yield rates of said bonds, and if the dollar were undergo hyperinflation, then sa
195.
by torginus 2mo ago
Why is this a meaningful figure? It's not debt that matters its how much it costs to finance it. It's Finance 101 that if you manage to borrow below inflation rate, and you have the luck that what you paid for appreciates, then yo
196.
by torginus 2mo ago
Generally a >10x is easily achievable if we use a chemical reaction that uses oxygen - which we do not have to carry around, that's why things like hydrogen fuel cells have a theoretical energy/mass that we can pretty much assu
197.
by torginus 2mo ago
Using VR for archviz has been a thing for quite a while, but my friend who works as a (building) architect told me that clients prefer 3d printed scale models, now that 3d printing got cheap and good.
198.
by torginus 2mo ago
Which has been a general point about AI - it's expensive to build and expensive to run, and requires a lot of infrastructure. Which is increasingly making it look like software might turn into a limited margin business, like making car
199.
by torginus 2mo ago
I'm not a security person, but how realistic is it to assume that you can carry on trying to exploit a company for days with a fairly large volume of activity, and not get detected?
200.
by torginus 2mo ago
Yes, imagine if 50 burglars showed up at your house and they tried to disassemble every camera, pick every lock and and force open every window for hours until they got in eventually.
201.
by torginus 2mo ago
Also these deals are not circular - NVIDIA has to pay a lot of money for chips and foundry capacity just as an example, so money leaves the circle and participants have to raise capital to replace it, which means if this were just a financ
202.
by torginus 2mo ago
There is undoubtedly a bubble in the sense that AI is crazy overfinanced, and there's a semiconductor shortage - chip makers usually have like mid to low two digit margins for manufacturers of complex chips like NVIDIA, and single digi
203.
by torginus 2mo ago
Interesting and its noteworthy to observe that the new GC actually increases the rate of L3 misses slightly. For reference, L1 misses are quite fast, and often free since the CPU can cover up the empty slots with useful work, however L3 mis
204.
by torginus 2mo ago
On my own time I still like the 'chat' format of me writing a short prompt and having the model turn it into a few dozen lines of code, or make a simple refactor (like turn a field into a method parameter. I wasn't impressed
205.
by torginus 2mo ago
This is a general theme with technology and the 'S-curve'. Let's not even get into whether the improvement for AI reasoning ability has slowed - for practical purposes of writing a React frontend, it has. But other tech is li
206.
by torginus 2mo ago
Yeah and these latest and greatest models universally suck hard for any use case outside of the few 'blessed' ones. Like I'm sure their ability to write prose and generally sound like a human being has regressed quite a bit,
207.
by torginus 2mo ago
Imo superglue is a shitty glue - it's far too strong and weak at the same time. It's strong in the sense that if you glue two thing together, you're unlikely to get them apart without damaging them. At the same time, superglu
208.
by torginus 2mo ago
My personal opinion is they're not hard to detect at all. Latency is a huge giveaway, If the client can't respond in a time you'd expect a client to be able to based on its geographic location, that's a pretty big tell t
209.
by torginus 3mo ago
Would it not be more efficient to create the decoder in-process, then you could feed it the bitstream via a compute buffer, and the video would be rendered to a texture.
210.
by torginus 3mo ago
Forgive me for my ignorance - but my understanding is that VLC is a frontend for ffmpeg, and even ffmpeg is 'just' a wrapper for hardware video decoders such as NVDEC when viewing supported codecs, like h264/5.

More ›