91–100 of 104 posts
I feel like this kind of tool has been completely obsoleted by LLMs (even local models). I've used similar tools such as tiny for Emacs, which is a DSL for generating text based on numeric ranges. Now, it's simply more efficient to ask AI.
Earlier quoted context omitted.
also made ripgrep
So this is who I get to "blame" for LLMs constantly saying "I want to use `rg` but I see that it's not installed on this system...".
The thing Biff gets right that gnu `date` and most stdlib datetime APIs get wrong: it treats "civil time" and "absolute instants" as different types. You cannot answer "what's 30 days from 2024-03-08 in America/New_York" without picking a side — DST means that's either 29d23h or 30d0h of elapsed time, and most APIs silently pick one without telling you. Jiff (the underlying Rust crate) gets this from Temporal in TC39…
The thing Biff gets right that gnu `date` and most stdlib datetime APIs get wrong: it treats "civil time" and "absolute instants" as different types. You cannot answer "what's 30 days from 2024-03-08 in America/New_York" without picking a side — DST means that's either 29d23h or 30d0h of elapsed time, and most APIs silently pick one without telling you. Jiff (the underlying Rust crate) gets this from Temporal in TC39…
That's an interesting perspective because if I use America/New_York I would assume that it's subject to the laws of New York City. If I wanted a different DST regime I'd just use +4, +5, or a different city.
Using +4 or +5 explicitly instead of a time zone when DST is relevant is bad juju. Then your arithmetic can become silently wrong in subtle ways when it cross a DST boundary.
Earlier quoted context omitted.
That's an interesting perspective because if I use America/New_York I would assume that it's subject to the laws of New York City. If I wanted a different DST regime I'd just use +4, +5, or a different city.
That's a complaint against how time zones are themselves named and organized. See https://ftp.iana.org/tz/tzdb-2022b/theory.html for all the gory details. Using +4 or +5 explicitly instead of a time zone when DST is relevant is bad juju. Then your arithmetic can become silently wrong in subtle ways when it cross a DST boundary.
same with macports
so it is pretty much not existing, cause I am not getting some binaries and drop them into some folder for some command line tool
Earlier quoted context omitted.
I believe AI-for-everything will become unsustainable financially for many and I’m genuinely curious to see how people deal with it. When to use it? When is it wasteful? My big hypothesis is that tokens are going to get much more expensive. Either that or OpenAI/anthropic are going bankrupt. I’m almost excited to find out, I have to admit. Your remark just reminded me of this, I went a bit off topic, I admit.
Have you tried DeepSeek V4 Flash? It's very competent and extremely cheap. I think Gemma 4 is also a good example of a capable small model. I mention these not only because they're cheap but because they can run on consumer devices. The "every year bigger and more capable SOTA model" trend is mirrored by "the every year smaller and more capable open source model" trend.
I need to find stats on average RAM of personal devices, but I expect it will be so low, we are light years away from running a frontier model (from today) locally on a smartphone, let’s stop dreaming (and I really would love having it).
I do agree local models are progressing and I am to this day in awe at what a 50GB file can do – it still feels like black magic to me.
Also granted, something like Gemma 2 2B seems to have similar performance to ChatGP 3.5 and only require 2GB of RAM. But I think the RAM/performance ratio curve over time is logarithmic and not linear, it’s moving slower and slower.