spijdar

3 ms·

1.
by spijdar 5d ago
I think he has actually posted videos of the restoration before (they've popped up in my youtube feed in the past), but I've never watched them, so dunno. The LLM site design and advertisement front-and-center are certainly a... c
2.
by spijdar 5d ago
It should be back up now! And being managed by the service manager now instead of running in a shell, so it should stay up. The one and only web host [2] written in Nim [0], running on Sprite [1] ;-) [0] http://nim-lang.org/
3.
by spijdar 6d ago
I've used LLMs to get the distributed Sprite OS running again, though I've yet to release it, and I'm not sure what that release will look like, if it ever happens. One thing I really don't get, and I really hope I never
4.
by spijdar 6d ago
Having seen laptops with dedicated copy/paste/undo/redo keys, I think Apple can be forgiven for not going down the function key rabbit hole: https://pomelo.shockfox.net/pub/img/powerlite.jpg (Yes, t
5.
by spijdar 8d ago
I want to believe this, but it's hard for me to take this at face value. It's been about 4 years since I've run pmOS, so my experience IS very out of date, but I also have a hard time believing that the experience has radical
6.
by spijdar 8d ago
Do you think there would be any worthwhile gains from relaxing address-dependent load ordering, like on Alpha/AXP? Or was that just a lot of extra pain for little reward?
7.
by spijdar 8d ago
It's Unsloth's UD-IQ4_XS, and it appears to actually work pretty well, regardless of the occasional CoT amnesia. Though, I've seen the "the user didn't tell me to do anything" thoughts on OpenRouter, too, which
8.
by spijdar 8d ago
Same reason/way projects like ReactOS/Wine and friends exist in a world where source code from early 21st century Windows is literally sitting in a bunch of 6+ year old Github repositories, I guess. Or any number of other propriet
9.
by spijdar 8d ago
They seem to be doing something different with the "Qwen4" architecture as demoed in Flash-Next. I've noticed the reasoning behaves ... weirdly. Like, really weirdly compared to any model I've ever seen before. I've
10.
by spijdar 9d ago
I run Unsloth's UD-Q4_K_S on 20 GB of VRAM (RX 7900 XT) and I get ~90k tokens of context without quantizing KV cache. With 8-bit quantization, I get about a 134k token context window. That's with only one slot, but for me, it work
11.
by spijdar 9d ago
Because one of the biggest hurdles to porting the Linux kernel and device-specific software to more hardware and making existing hardware work better is one of the things that "AI" can (seemingly, caveat) do surprisingly well? Not
12.
by spijdar 11d ago
I think the premise is basically yes, you should assume there will be data-losing bugs, but: 1. The filesystem should be reasonably good at detecting an error/corruption state and informing you, and 2. You should have backups of said d
13.
by spijdar 12d ago
... You know, now there's an idea ... But in seriousness, I've gone down this rabbithole ever year or two for about as long as I've had the darn thing (including one especially masochistic attempt at making it work on ppc64le
14.
by spijdar 12d ago
I think the only way this is really compelling is if it's able to effectively function as a virtual Linux desktop with a reasonably clear floating monitor in front of you. I'm not sure that's appealing to me for a couple reas
15.
by spijdar 13d ago
At the moment, no. Beyond just fixing problems I've found, for the last year I've been doing a lot of haphazard development inside of Sprite itself, so "fixes to bugs in released Sprite" are intermingled with "featu
16.
by spijdar 13d ago
You really don't want to run Sprite on a single host, unfortunately -- the local filesystem path is buggy, and there are some major bugs in the system as preserved on the CD-ROM, e.g. vfork() is totally borked, at least for software
17.
by spijdar 13d ago
Not really relevant to keyboard symbols, and not PC symbols, and it's not really lost to time, but does anyone know where the Ethernet symbol used on Apple products (and ... Android???) comes from? You know, the <...> symbol. It&
18.
by spijdar 15d ago
Both! The Sprite kernel released by UCB doesn't run on any qemu target, but I've made ports to the PC and to sun4m, which qemu can run. I've gotten the sun3 (m68k) and sun4 (sparc) ports to run in phabric's fork/con
19.
by spijdar 15d ago
Why use swap files when we could have swap directories? ;-) Sprite had (IMO) a really interesting solution to swap. Each host had a directory, referenced by its host ID under the `/swap` directory, and inside that directory, individual
20.
by spijdar 15d ago
This strikes a chord with me, though I came to a somewhat different endpoint, at least for now. My perspective (so far) is that it was probably never a good thing for me to tie up my identity/sense of self in my ability to write modera
21.
by spijdar 17d ago
Yeah, I dunno. For me it does "talk normally" for the most part when used in an actual coding harness. One thing though, the actual prompt I used was pretty long (844 words), and ... generated by GPT-5.6 Sol (lol), with the intent
22.
by spijdar 17d ago
This is Unsloth's UD-Q4_K_S quantization (edit -- on llama.cpp, via the Vulkan backend, on an RX 7900 XT, with Unsloth's recommended sampler config), for "as replicable as LLMs can be" disclosure, done through llama.cpp&
23.
by spijdar 17d ago
It's interesting that someone else noticed this. A week or two ago, GPT-5.6 Sol starting leaking reasoning into a tool call in Pi. I don't really know what happened, but it was ... interesting: Attach. Use hub debugger. Ensure
24.
by spijdar 18d ago
There's a fuzzy gray line here. Whether formalized or not, the way I see it, pilots will de facto be penalized for making decisions that cost an airline a ton of money, especially if it turns out to not have been an actual emergency.
25.
by spijdar 25d ago
Sure, one shot generation doesn't work, you have to split into sections and generate each section independently, then usually do a "seam" pass to check continuity between two bits. Though, I'd also agree that if you&#x27
26.
by spijdar 25d ago
It's that, yeah, but coming from multiple orders of abstraction. In this case, part of the experiment was to see what "oh-my-pi", a "fat and feature rich" LLM harness, could do when coupled with modern GPT, given a
27.
by spijdar 26d ago
Claude is the one model family I've not really used. Which yeah, feels like a backwards thing to say in a world where seemingly everyone using LLMs is using Claude Code. I've used some Opus 4.5/4.6 via Antigravity and Sonnet
28.
by spijdar 26d ago
Yeah, this tracks with my experiments. About every 6 months or so for the past year-and-a-half I've tried using the "frontier" LLMs to as-near-as-possible autonomously write novel length stories, because I find it fascinating
29.
by spijdar 1mo ago
The problem isn't that people consider it boring, it's that the use of LLMs is considered essentially sacrilegious, and whatever end-result that's created is considered tainted, often regardless of the extent of LLM usage. Wh
30.
by spijdar 1mo ago
> If not, why did you even study that? I like to make computers do the things I want. Since I was a kid, the satisfaction from programming or any sort of "computer engineering" (using the word engineering VERY loosely here) is

More ›