kqr

18 ms·

451.
by kqr 10mo ago
> They used the same routines they wrote for their day jobs at Softdisk in the Keen code. [...] Most of the IDLIB.C code must have come directly from the PC version of Dangerous Dave. [...] there is some extremely strong evidence showin
452.
by kqr 10mo ago
The problem of focusing on failures is that such analysis misses all the losses that occur even when everything works as designed. Analysis has to focus on all losses -- both failures (often the trivial case) and non-failures (design errors
453.
by kqr 10mo ago
It sounds like our main disagreement lies around whether to call it "design error" or "build error" but I do not believe this erases the useful distinction between "error present in the thing from day one" and
454.
by kqr 10mo ago
The first three are hardware failures, not software failures. The latter would be a design error, not a failure. The software may need to handle hardware failures, but software that doesn't do that also doesn't fail -- it's i
455.
by kqr 10mo ago
Oops! https://entropicthoughts.com/software-never-fails
456.
by kqr 10mo ago
> Publishers are now rewarded for publishing more papers, as opposed to having more readers. That's the first order effect, but you have to look beyond it. If authors have to pony up $1500, they will only do so for journals that hav
457.
by kqr 10mo ago
It doesn't help that many of the popular methodologies focus entirely on failures. They ask a bunch of questions in the style of "how likely is it that this part fails?" "what happens if it fails?" "how can we
458.
by kqr 10mo ago
Yes, 2.5 Flash is extremely cost efficient in my favourite private benchmark: playing text adventures[1]. I'm looking forward to testing 3.0 Flash later today. [1]: https://entropicthoughts.com/haiku-4-5-playing-text-ad
459.
by kqr 10mo ago
> If you post something interesting or meaningful, then people will respond. That's an oversimplification. There are things that get responses because they're flamebait rather than interesting, and then even more interesting th
460.
by kqr 11mo ago
This near miss plus that helicopter thing, I assume: https://en.wikipedia.org/wiki/2025_Potomac_River_mid-air_col...
461.
by kqr 11mo ago
I agree. I'd assume that's a holdover from a time when innovation moved slower.
462.
by kqr 11mo ago
> Most of the time when things get patented, it's strictly worse for innovation in that space until the patents expire. What happens at expiration is an important and intended feature of patents. They trade a legally guaranteed head
463.
by kqr 11mo ago
Let's just say a family member has a very important and somewhat secretive job. The most classified meetings don't happen in our home, of course, but for a state adversary even small clues can be what's needed. (So why get a
464.
by kqr 11mo ago
Our household (and I suspect many with us) bought a Roomba specifically to not give the Chinese government a roving camera in our home. Ouch!
465.
by kqr 11mo ago
> A giant baby is navel corked which if removed causes a vacuum. Very The Age of Wire and String .
466.
by kqr 11mo ago
Optimally queueing SCVs and marines and supply depots requires an APM of 11 or so in the early stages of a Brood War game on the fastest setting. Add a couple more APM for scouting, and we'ree still not talking crazy levels of multitas
467.
by kqr 11mo ago
I agree. I think people underestimate the size of the Korean Brood War scene, even relatively early on. In my country, I had seen some huge LAN parties with associated competitions, but then I got introduced to Korean Brood War competitions
468.
by kqr 11mo ago
I don't think this is true. Granted, last time I tried to get good at an RTS was toward the end of the Brood War era but the established wisdom at that time was very clear that hour-for-hour, time spent practicing resource management w
469.
by kqr 11mo ago
There's a lot of I/O hidden in CPU-bound loads related to fetching, prefetching, caching, etc.
470.
by kqr 11mo ago
Although you're correct that it would be too soon to prophesize their death, I want to clarify that Lindy ensures correlation, not causation.
471.
by kqr 11mo ago
Their annual geometric mean return is 45 %! That's some serious overbetting. In a market that didn't accidentally align with their biases, they would have lost money very quickly.
472.
by kqr 11mo ago
Extremely similar earlier submission but focused on cryptocurrencies, using real money, and in real time: https://news.ycombinator.com/item?id=45976832 I'm extremely skeptical of any attempt to prevent leakage of futur
473.
by kqr 11mo ago
The one place where radians are more convenient is when you are at the centre of the circle. Then something which is as wide (or tall) as it is far away subtends one radian in your view. (And correspondingly, if it subtends half a radian it
474.
by kqr 11mo ago
Is there a way in which this is good for a segment of consumers? When the current gen of GPUs are too old, will the market be flooded with cheap GPUs that benefit researchers and hobbyists who therwis would not afford them?
475.
by kqr 11mo ago
Wait 'til you hear about the PDP-11 emulator of a CPU it is running on.
476.
by kqr 11mo ago
To people who find this stuff useful in practise today (and not merely fascinating or useful 50 years ago): what is your line of work? I have needed to know the values of a few integrals in my job, but I have always ended up with a close en
477.
by kqr 11mo ago
I want to pull two quotes from the article. > “Our findings emphasise that exercise remains beneficial even in polluted environments,” lead researcher Professor Po-Wen Ku said in a statement. [...] “We don’t want to discourage people fro
478.
by kqr 11mo ago
I agree bathroom doors are sensible when traveling with company. When staying alone, though, I prefer no doors. If there is a door, I tend to use a small piece of gaffers' tape to hold it open. I don't spend a lot of time in the h
479.
by kqr 11mo ago
You are claiming humans copy-and-paste copyright headers without copying the corresponding code. To prove you're correct, you only need to show one (or a few) examples of it happening. To prove you incorrect, someone would have to go t
480.
by kqr 11mo ago
Huh, I didn't know that! Are there efforts to automatically reduce the number of parameters once the model is trained? Or do the relationships between parameters end up too complicated to do that? I would assume such a reduction would

More ›