oofbey

18 ms·

361.
by oofbey 1y ago
Language constantly evolves. So it’s a little unfair to say anybody is trying to change a definition. In specific professional contexts, definitions are necessarily much more precise. Because in these contexts definitions have consequences
362.
by oofbey 1y ago
25 chops per second would be EXTRAORDINARY. Possible? Probably but only with super elite training. Most competent home chefs can probably do 5 Hz and probably struggle to get to 10 Hz.
363.
by oofbey 1y ago
Google's incapable of sticking to anything good. Nobody gets promoted at Google maintaining a system. So somebody inevitably gets the idea to make a 2.0 which isn't backwards compatible, enabling them to do all sorts of bold pro
364.
by oofbey 1y ago
Sounds like letsencrypt is being quite premature by turning off OCSP. https://letsencrypt.org/2025/08/06/ocsp-service-has-reached-... Might be EOL in some theoretical sense, but by turning it off they'r
365.
by oofbey 1y ago
Not saying anything new here, but at the core there are only a few key reasons for using bitcoin: investment, hiding your finances, and the idealism of de-centralization. The intrinsic value of decentralization is the ability to operate out
366.
by oofbey 1y ago
Presence is the key problem. It scales badly - in terms of compute, bandwidth, and battery. And it's not actually useful. Lose lose. Solution: don't use XMPP.
367.
by oofbey 1y ago
Here's a specific complaint about XMPP, and possible explanation why nobody uses it any more. (I worked on a large-scale XMPP implementation back in the day.) Presence. That's the colored dots indicating "is somebody online
368.
by oofbey 1y ago
XMPP fans are continuously baffled by why XMPP isn't used for _everything_. STILL! When will the world wake up and realize its simplicity and beauty?
369.
by oofbey 1y ago
I'm sure (a) is true. But TIOBE is supposed to be some balance between "total mass of existing code" and "current interest in new code".
370.
by oofbey 1y ago
Agreed. Probably why they renamed it. I think everybody probably wishes they'd taken a new name from the start.
371.
by oofbey 1y ago
The TIOBE index says Perl is currently the #11 most popular programming language (up from #30 a year ago). ref: https://www.tiobe.com/tiobe-index/ Now, I don't actually believe this, because that puts Perl way ah
372.
by oofbey 1y ago
Part of me wants to go send Perl some money, because it was so important in my formative years. Another part of me is like "people still use perl?"
373.
by oofbey 1y ago
The Kinesis Advantage literally saved my career. I have lived in fear for a long time that the company would go out of business. So every chance I can justify I send them another $400 for their latest keyboard so I help keep them in busin
374.
by oofbey 1y ago
If you're serving static pages through nginx or something, then 10/sec is nothing. But if you're running python code to generate every page, it can add up fast.
375.
by oofbey 1y ago
Yup. Market forces at work here, folks.
376.
by oofbey 1y ago
I love uv. The one gotcha I'll warn people about is: don't touch uvx. I've lost an embarrassing number of hours or days trying to figure out why nothing works properly or makes sense when I tried to run things with uvx. I
377.
by oofbey 1y ago
At some point the world's gonna figure this out and start making tetrachrome cameras and screens and it's gonna be the next big TV upgrade after 8k.
378.
by oofbey 1y ago
Cool - thanks for explaining that. I totally see how each has its place. I imagine it's pretty complex to take the raw scan data and generate 3dgs. Are these algorithms simple & standard, or do they take a fair amount of tuning &
379.
by oofbey 1y ago
How much of the huge file size is because you need tons of splats to simulate a hard surface? Conceptually the splats seems flawed because gaussians don't have hard edges - they literally go to infinity in all directions, just at vani
380.
by oofbey 1y ago
I read a lot of AI papers on arxiv, and it's been a while since I read one where the first line of the abstract had me scoffing and done. > We introduce the FIRST method for translating text embeddings from one vector space to anoth
381.
by oofbey 2y ago
Perplexity solves this problem perfectly for me. It does the web search, reads the pages, and summarizes the content it found related to my question. Or if it didn't find it, it says that. I recently configured Chrome to only use goo
382.
by oofbey 2y ago
No - they are all competing for eyeballs right now. They need to solidify their positions in terms of user habits and behaviors before they start pissing people off.
383.
by oofbey 2y ago
Would love to see the pareto trade-off curve of "wins" vs "betrayals". Anybody drawn this up?
384.
by oofbey 2y ago
So Matrix is the new XMPP?
385.
by oofbey 2y ago
Could pick a better homepage example. The demo code doesn't render very well in MathML. Kerning on the exponents is off. In "beta^alpha" the alpha character touches the beta. Looks better switching to MathJax.
386.
by oofbey 2y ago
Why would that matter? I understood the point was to speed up matrix multiplication by doing the adds and multiplies in a different order. Shouldn't matter whether the datatype is int, float, complex, whatever.
387.
by oofbey 2y ago
They're proposing "new hardware architectures" to take advantage of this idea. Anybody with a background in GPU floating point math comment on how realistic this is?
388.
by oofbey 2y ago
In normal diffusion you train a model to take lots of tiny steps, all the same small size. e.g. "You're gonna take 20 steps, at times [1.0, 0.95, 0.90, 0.85...]" and each time the model takes that small fixed-size step to ma
389.
by oofbey 2y ago
That's true - they are quite good at OCR. But they're really bad at a bunch of tasks that seem like they should be super simple. Like "are these lines crossed" or "which letter is circled". See https:/
390.
by oofbey 2y ago
Fun! I wish the demo had the attention visualization. Would that be easy to add? Is the source code for the HF demo in the repo too?

More ›