gf000

3 ms·

61.
by gf000 18d ago
Could you sketch a spring-loaded button? I just have a hard time imagining how could this time dimension be added properly to a button - not that I would disagree with you in general. Also, is this really different from a button that refle
62.
by gf000 24d ago
There is an immich proxy project: https://github.com/alangrainger/immich-public-proxy I wouldn't make the instance itself public, but I think this is a decent tradeoff. Of course still requires one to be a sysadmi
63.
by gf000 24d ago
What do you need it for? Which version of the JVM?
64.
by gf000 24d ago
It absolutely doesn't need any memory tuning whatsoever. People just read 2 decades old blog posts and apply whatever random flag they see. For 90% of workflows you don't add any command line arguments, for the remaining 10% you m
65.
by gf000 25d ago
It doesn't use more memory (at least not due to GC itself), it wastes some memory that it won't touch again unless absolutely necessary.
66.
by gf000 26d ago
AFAIK everyone's brain has a sort of hard coded number of "registers", around 3-4. I think the visual "instant counting" also goes up to that number, so in some way we are all "bottlenecked" and require ac
67.
by gf000 26d ago
Well, that's just one implementation (the easiest). You may have the harness automatically inject stuff into the ongoing conversation, in which case most of the work is not done by the working agent (though you may still end up using
68.
by gf000 26d ago
I mean, it's all just tool calls and context-management in the end. Like it is pretty self-evident to have a "heartbeat" and to do specific stuff on a schedule, etc.
69.
by gf000 1mo ago
In my experience some of these models may have learnt some censoring during distillation of Western models, but these are mostly just there as a probable response. So if they first happen to respond "I ain't doing this because leg
70.
by gf000 1mo ago
> silent sound fingerprinting mechanisms I hope you also block the other half of the internet because sure as hell google and facebook are way worse than that.
71.
by gf000 1mo ago
That's a TOS thing though. I don't think that works just as trivially in e.g. the EU, you can't just single-handedly stop a service someone is paying for.
72.
by gf000 1mo ago
Chinese government has Chinese censorship, Western one has western ones. You are not concluding what you think you do here.
73.
by gf000 1mo ago
(assuming English text) 512 tokens ≈ 350–400 words ≈ a long paragraph or two. 1024 tokens ≈ 700–800 words ≈ about a page and a half to two pages. I would be very surprised if that amount of text is not enough to encode a general topic int
74.
by gf000 1mo ago
Absolutely a novice in this topic, but I would imagine that by simply having sufficiently big chunks it's simply not a problem? You surely have enough information in like a couple of paragraphs to denote in vector space roughly what it
75.
by gf000 1mo ago
While way too restrictive, I believe android actually has the correct idea here, having separate data folders per app. The problem is that you can't really touch this per app data folders - even backup is a special operation only doabl
76.
by gf000 1mo ago
Another common format is `jsonl`, basically a json object per line containing a message
77.
by gf000 1mo ago
Not really solving all your cases, but I found tmux (or herdr) on a home server works pretty well. I just ssh into my home server and continue where I left off with the same claude/opencode open. On a longer term, I think "assista
78.
by gf000 1mo ago
Yep, many successful people greatly underappreciate the effect of simple dumb luck in their lives. And often they just make up complex reasoning chains, even wholly believing them, that more of it was in their control/talent, etc. None
79.
by gf000 1mo ago
All those tools still get into context. Also, it's not hard to filter a rest api doc - like with a special tool for that the harness itself could pre-filter it and add only the relevant ones as tools to be even more leaner than what an
80.
by gf000 1mo ago
Absolutely not an expert on this, but I would imagine that they would train the pipeline on a raw-ish image data to have better consistency? I would think that some kind of algorithm "bettering" the image suddenly might cause a di
81.
by gf000 1mo ago
If google allows that for pixels, I don't see why Motorola wouldn't. Besides, GrapheneOS would probably don't want to get into a situation where their keys are burnt into hardware non-modifiable. They had to revoke such a key
82.
by gf000 1mo ago
well, apparently not even Microsoft had enough capital. It's almost like the free market only works when we have well-defined and regulated markets . This has been known by Adam Smith and quite logical, yet people expect Google and Ap
83.
by gf000 1mo ago
Well, it's kinda hard to scan a QR code displayed on your screen when the camera is attached to its back (semi-joking)
84.
by gf000 1mo ago
Because it's a mobile platform and "GNU+Linux" is laughably terrible in this space. It's almost like Android has put millions of expert dev hours into making it the most used OS in the world. Like GNU+linux on laptops on
85.
by gf000 1mo ago
Well, you either spend energy freeing stuff or you waste some memory. It's a tradeoff, that you can trivially set with a single flag in Java. In pretty much every other runtime's case you are stuck with whatever their GC uses, a
86.
by gf000 1mo ago
Then apparently no one less than 40 works at a faang or any other company from the top 100? Because java is still one of the top 3 languages by any ranking worth its salt (not you, tiobe).
87.
by gf000 1mo ago
JVM has real and virtual threads and arguably just as good if not better on all these points. (I actually have/am writing a harness in Java fwiw, but mostly as a hobby/experimentation)
88.
by gf000 1mo ago
But why? Why wouldn't I want to enjoy stuff like this: https://ciechanow.ski/mechanical-watch/ Or watch a video. Or do some basic editing of some kind of medium that I will only do once and would make no sense to
89.
by gf000 1mo ago
> Facebook is the ad network and serves its own ads. Well, it certainly replaces valuable data like how long have you looked at a given ad with either nothing, or random noise (e.g. if you draw a rectangle over them). We are no longer li
90.
by gf000 2mo ago
Ironically, LLMs are pretty good at finding stuff like that from vague descriptions. With that, I think the parent commenter meant Art Graesser's AutoTutor work based on the terminology used (hint/pump). TLDR: Actual human tutorin

More ›