kees99

6 ms·

1.
by kees99 3d ago
> model to add new vulnerabilities to your code so it can present you with more exciting findings. Not so sure about this level of 4D chess capability just yet. The other day I asked Opus to come up with some cleverly vulnerable crypto c
2.
by kees99 5d ago
zram is a good choice when data in RAM is very repetetive/compressible. Web browsers is one such load. Android apps is another good one. When that data doesn't compress well at all, zram is no help and traditional swap-to-storage
3.
by kees99 5d ago
Not "a" bank. You (somewhat) trust your bank, I would imagine. Since you know, they have your money. GP's hypothetical here is that Anthropic or other service provider who wants to do "age verification" could part
4.
by kees99 7d ago
This is exactly the problem, unfortunately. For somebody who knows a bit how things are set up, or is willing to spend 10 minutes researching, it's a no-brainer that you can just "git clone" entire linux kernel development hi
5.
by kees99 13d ago
Having a mix of both 2LD and 3LD registrations under the same TLD is a bit of a nightmare in terms of public-suffix list [0], which is kind of important thing when enrolling your domain for some services, cloudflare among them. [0] https:&
6.
by kees99 14d ago
> Amsterdam is a uniform medium density throughout and at its border it just ends and gives way to farmland. Except Amstelveen. But then, there isn't much of a change in density there.
7.
by kees99 14d ago
If I have to guess, GP tried to translate following quote from Jane Austen into ASD-STE100: “It is a truth universally acknowledged, that a single man in possession of a good fortune, must be in want of a wife.”
8.
by kees99 16d ago
By year 2000, there were "proprietary" messengers aplenty - yahoo, aim, icq, qq, and a bunch more. All of which were simple enough to reverse engineer, so Linux user could just install Gaim (later renamed Pidgin) and connect to mo
9.
by kees99 17d ago
Or, html generation could be very, very fast in the first place. For example, serving static files is usually just as fast as a reverse-proxy cache. Interestingly enough, cgit bills itself as "a hyperfast web frontend for git repositor
10.
by kees99 18d ago
Isn't this[0] the use-case that is supposedly covered by LGPL? [0] copyleft-licensed library that could be legally linked into non-copyleft code base.
11.
by kees99 19d ago
Would love this as a browser plugin.
12.
by kees99 23d ago
I run LFM2.5 8B locally. It's not very smart, to say the least. But then, there are plenty of mindless, menial tasks out there, and it would go through those like a champ. And quickly, too.
13.
by kees99 27d ago
Few years back, Amazon had a rampant "fake USB storage" problem too. Then, one day, all the fakes were gone, just like that.
14.
by kees99 28d ago
Ornith-1.5-397B is derived from Qwen3.5-397B-A17B via post-training. That process preserves exact parameter count.
15.
by kees99 1mo ago
Re: bonsai - unsloth's quants have Q2 (UD-IQ2) variants, which are more or less same in size. ...or did Prism do something special with their "bonsai" releases? I didn't notice anything like QAT being mentioned.
16.
by kees99 1mo ago
Qwen models are slower in tokens/s, compared to similarly sized gemma4 and others, and they use more tokens per task, in part thanks to that xhigh default. On the other hand, there are some of us who are stuck with hardware that has
17.
by kees99 1mo ago
More like rack- sized GPU that takes in half-megawatt of power as 800 volts DC, and requires 10 gallons per second of liquid cooling or it'll catch fire.
18.
by kees99 1mo ago
For what it's worth, according to internet archive, this banner has been there for a while. At least since February: https://web.archive.org/web/20260211083755/https://www.dosti...
19.
by kees99 1mo ago
> multiple times, and results always same Not my experience at all. With smaller models, whenever I see a response that is going into wrong direction, I would just redo that step, and more often that not that brings improvement. This eff
20.
by kees99 1mo ago
> Cloudflare, Google, Anubis, and all these other captcha services trying to see if you're a human. Yep. IMO, this is so far the biggest AI-inflicted damage to the web. A bit of anecdata - wikipedia (and all other wikimedia sites)
21.
by kees99 2mo ago
Wasn't that abbreviation from "pipe"? As in - this is a pipe that you can use to connect terminal to a file, file to a printer... and yes, file to a file (that's cp/copy).
22.
by kees99 2mo ago
I've fed a couple of those chicken-scratch sketches to nanobanana with prompt "Treat attached as a technical drawing of a bicycle. Produce photo-realistic image of the actual bike manufactured to that spec. Try to stay close to th
23.
by kees99 2mo ago
Pretty good results overal. Even shows subnetting within e.g. digitalocean's blocks! Couple of misidentified IPs: 141.98.83.0/26 - shows as "Panama", echoing bogus whois record. Real location is probably Romania, at leas
24.
by kees99 2mo ago
MADV_SEQUENTIAL might help a bit, but not that much. Biggest problem here is throughput-vs-latency. With mmap()-ed file, for each pagefault, kernel will conservatively estimate block size to page in, so you'll have a ton of relativel
25.
by kees99 2mo ago
Qwen is quite slow in general. I'm getting ~0.5 tps from [qwen], and ~10 tps from [gemma] - roughly the same size, same quant, same hardware (8-core CPU) same software (llama.cpp). [qwen] Qwen3.6-27B-Q4_K_M.gguf [gemma] gemma-4-26B-A4
26.
by kees99 2mo ago
Yes, something like this: static inline long read(int fd, void *buf, long count) { register long rax asm("rax") = __NR_read; register long rdi asm("rdi") = (long)fd; register long rsi asm("rsi")
27.
by kees99 2mo ago
Claude has surprisingly good knowledge of X11 protocol. The other day, colleague showed me a (pretty basic) terminal emulator written in one-shot by Opus. Kicker is - that was compiled to a 30 KB static binary. That's right. No libX1
28.
by kees99 2mo ago
> same prompt 4 or 5 times LM Arena permanently flagged / sent to the captcha hell my account after trying the same prompt twice. So, YMMV.
29.
by kees99 2mo ago
I agree. This addresses one of two main failure modes of solar roadways (grippy coating disintegrating from being driven over), but keeps another one (contamination). Note that this particular trail was right next to a terminus of very ligh
30.
by kees99 2mo ago
> All it needs to do is be sent a new message. Receiving a message in a timely manner, is far more energy-intensive, for two reasons: - modern Internet, with nearly universal deployment of NAT, is such that any app (or phone OS, for th

More ›