sfink

47 ms·

661.
by sfink 2y ago
Yes? Well, partly. For systemd and pulseaudio, the systems they were replacing legitimately had major problems. There were variants and workarounds that fixed some of these, but no holistic solution that I've ever heard of. There were
662.
by sfink 2y ago
I have found that with some practice, stack-based is better for writing, infix is better for reading. At least for equations. Pythagoras is trivial to write: a a * b b * + sqrt and in fact the process of entering a calculation can
663.
by sfink 2y ago
Yeah, that's the part that stuck out for me. "sudo is bad because it does all these things it shouldn't do instead of just the one thing it's for and nothing else. My tool is good because it does just the one thing it&#x
664.
by sfink 2y ago
Huh. I'm not at all a fan of how Poettering operates, but it's neither the ideas nor the implementation where I'd fault him. Well, it depends on what you mean by implementation, I guess; I'm talking about the core "
665.
by sfink 2y ago
This feels so wrong in one way, and simultaneously so right in another.
666.
by sfink 2y ago
I haven't done a lot of userscript dev, but based on my experience so far, all of this advice is spot-on. I found it useful to create a `waitForElement(query)` helper function that takes a CSS selector path and returns a Promise that r
667.
by sfink 2y ago
You can wimp out and just use MutationObserver as a way to get a callback whenever anything changes, ignoring all of the mutation records. Then in the callback, look up all of the elements you care about. The API is pretty simple then. It
668.
by sfink 3y ago
From https://hacks.mozilla.org/2024/04/llamafiles-progress-four-m... it sounds like it's a very active "feeder fork", with jart doing a massive amount of cutting-edge work on CPU inference and slash
669.
by sfink 3y ago
Heh. I wrote up my comment, and then thought "hey, I bet that's what that 'Parse don't validate' article meant, the one I never quite got around to reading." So I pulled it up — great article! — but then didn&#
670.
by sfink 3y ago
Alternatively, don't validate and then use the original. Instead, pull out the acceptable input and use that. Even better, compare that to the original and fail validation if they're not identical, but that requires maintaining a
671.
by sfink 3y ago
Macros that are giving you some value can be ok. In this case, once the performance conclusion is reached, the only reason to continue using a macro is if you really need the `final`ity to vary between builds. Otherwise, just delete it or u
672.
by sfink 3y ago
tldr: sprinkled a keyword around in the hopes that it "does something" to speed things up, tested it, got noisy results but no miraculous speedup. I started skimming this article after a while, because it seemed to be going into t
673.
by sfink 3y ago
https://neverworkintheory.org/ is the web site associated with this (from the 1st reference in the paper)
674.
by sfink 3y ago
My next user script project idea would require something like this. It's for my own use, but I have to fill out grade sheets using an incredibly painful web UI. The data would be far more easily entered into a spreadsheet. (That's
675.
by sfink 3y ago
Yes, I'm grateful to the author too.
676.
by sfink 3y ago
It's refreshing to read something about ML and inference and have it not be anything related to a transformer architecture sending up fruit growing from a huge heap of rotten, unknown, mostly irrelevant data. With traditional ML, it&
677.
by sfink 3y ago
> But if Americans follow the reasoning to put the most important information first, then why do they say "March 5th, 2024"? With that reasoning shouldn't it be "2024 March 5th"? Because you're falsely equat
678.
by sfink 3y ago
If I understand correctly, or at least if I map it to my own similar complaint: the problem is not that they have you study simple or simplified cases, it's that the ignored complexity is unacknowledged and sometimes even denied. Which
679.
by sfink 3y ago
I've often wanted a simple online form solution for random purposes, yet I have never quite gotten around to learning Google Forms. My kids use it for school stuff. They're reasonably capable with it and have gotten good mileage f
680.
by sfink 3y ago
Yeah, it looks wrong to me too. It claims to be GPLv3 and the "use cases" explainer looks like it's trying to clarify what GPLv3 means, but the requirements described under the use cases are not part of GPLv3. The 1st one is
681.
by sfink 3y ago
I think the question might be a little too general to get good answers back. But it's an excellent question IMHO, and one that I really wish were more in the forefront of people's thinking. If I were to make a website on some topi
682.
by sfink 3y ago
I would prefer my car to fly through the air, but that's not what it does. My point is that LLMs are not magical, they're limited by their architecture and reality. They are not symbolic rule processors, even though they can fake
683.
by sfink 3y ago
Why would it do that? Rules and axioms scale (slowly) with the number of layers. The model can heuristically approximate more easily and more incrementally.
684.
by sfink 3y ago
One simple reason: consider the plausibility of 11 + 31 = 24 It's actually fairly plausible. The answer is numeric. Two digits, even, which is pretty likely when adding together 2-digit inputs. 24 is also a common answer to ma
685.
by sfink 3y ago
The LLM anthropomorphizing will continue until the machines are happy with how we think about them.
686.
by sfink 3y ago
Nope, truly breaking down and digested by bacteria. > The last measurement involved chemical analysis via gas chromatography/mass spectrometry (GCMS), which detected the presence of the monomers used to make the plastic, indicating
687.
by sfink 3y ago
I'm a guest, not a host. I do not think of airbnbs as identical to hotels in every way possible, and I don't want them to be. I am ok with doing some basic cleaning tasks despite being charged a cleaning fee. I naively still consi
688.
by sfink 3y ago
This article would be better if it reported on the topic of the article, rather than stopping at stating the topic. Seriously, the whole thing could easily fit in 240 characters. NIST staffers are revolting against his appointment... why, e
689.
by sfink 3y ago
I don't think Occam's Razor is that helpful in explaining why I can't find my glasses in a pitch black room. A better explanation is that it's dark. There may well be other good reasons, but Occam's Razor says to st
690.
by sfink 3y ago
But the article's main point is that the intermediates/gradients in heat and moisture are not necessarily where you expect them to be. If you have a point heat source on one side of an enclosed volume—a big box, say—then in a vacu

More ›