pxc

7 ms·

61.
by pxc 6mo ago
It's pretty easy to prove lots of negatives outside of mathematics, too. It's easy to prove there's no elephant in my living room at the moment.
62.
by pxc 6mo ago
https://www.cloudflare.com/trust-hub/reporting-abuse/ Anyone can report illegal content on Cloudflare and Cloudflare will remove it. The pirate streaming sites pop up only in or just before the first few moments o
63.
by pxc 6mo ago
Is that something anyone wants? It seems like being able to plug into other editors works well. What are the experiences people are trying to get but currently can't build because ACP sucks or whatever?
64.
by pxc 6mo ago
Ah, ah, got it. So you get a kind of graceful degradation if the optional dependency is missing. Nice! And, I guess, it's headless in the sense that you can't attach to it even if you want to. I get what you meant now.
65.
by pxc 6mo ago
> No country ever won this fight short of total shutdown/disconnects. Some countries also throttle pretty effectively. So you can connect but if you're trying to do more than read Hacker News it's not very usable.
66.
by pxc 6mo ago
Cloudflare has not in fact refused to comply with any court orders! The very thing at issue is that LaLiga wants Cloudflare to do censorship on their behalf that Cloudflare, who has no contractual relationship with LaLiga, is not required t
67.
by pxc 6mo ago
LaLiga isn't Cloudflare's customer. They have no relationship. So why would Cloudflare rework their infrastructure just to instrument rapid blocking at their own expense as a favor to LaLiga? And if they don't, ISPs just brea
68.
by pxc 6mo ago
If xterm.js is slower than tmux, why don't you just use tmux for the headless sessions as well? How is tmux not headless enough?
69.
by pxc 6mo ago
What he prides himself in (in this context) is craft, which LLM use probably can enable, but definitely isn't commoditized by the kind of vibe coding that Garry Tan is doing.
70.
by pxc 6mo ago
Pijul seems great and I really want it to flourish. I love reading the blog posts when there are major updates, and its approach seems fundamentally more right than Git's (as it should— while it gets its core theory elsewhere, its UX l
71.
by pxc 6mo ago
There's a difference between a reliable hunch and really knowing something. What is obvious is not always (or even usually) easy to prove. And the process of proving the obvious sometimes turns up useful little surprises.
72.
by pxc 6mo ago
For short texts, the translation I usually want the most is fast translation, and local models are actually great for this. But for high-ish quality translations of substantive texts, you typically want a harness that's pretty differ
73.
by pxc 6mo ago
> Heh, lucky you, at least you get a message. My ISP just drops traffic to the affected IPs. No ping, no traceroute, just a spinner in the browser until it says "page not found". This is generally how the GFW works in China. In
74.
by pxc 6mo ago
We often look back on earlier stages in world history like we're somehow more advanced, or inherently smarter, than past societies. But one of the things made clear by the way this problem lines up perfectly with conflict during the in
75.
by pxc 6mo ago
I think the right way to think of it, from a self-protective perspective, is this: the real offer is the per-token pricing. Use that for a while, and iff you are consistently spending more than $20/mo, treat the subscription offering a
76.
by pxc 6mo ago
Translation generally works well with very small models compared to the frontier LLMs. You can definitely run a model on your own hardware for this.
77.
by pxc 6mo ago
It's a bit shocking to me how opaque the pricing for the subscription services by the frontier labs is. It's basically impossible for people to tell what they're actually buying, and difficult to even meaningfully report or c
78.
by pxc 6mo ago
This is a huge deal! I secretly hoped for this. :)
79.
by pxc 6mo ago
I also sometimes like to use typing test pages to kinda warm myself up before I start a project. I want to do really well and race someone else or type faster than I usually can, then ride the high of victory to get something done or break
80.
by pxc 6mo ago
Tart is really cool, impressive, and useful. Best of luck to the team!
81.
by pxc 6mo ago
Yes, please share it!
82.
by pxc 6mo ago
It's a good move. I have a case on my MBP that helps with this because it means the edges are plastic for me, and not quite so sharp. If you want to break more rules, you might consider chickenwing-ing your arms a bit. Deviate from the
83.
by pxc 6mo ago
> weird keyboard mappings for things like "start of line", "end of line", "previous word" Those are Emacs keybindings, and they're also present by default in Bash since they were copied by GNU Readline.
84.
by pxc 6mo ago
Imo that's a big deal primarily because the issue with automatically discerned vulnerabilities has long been a high volume of reports and a very bad signal-to-noise ratio. When an LLM is capable of developing PoC exploits, that means y
85.
by pxc 6mo ago
I use Linux VMs on macOS mostly headless via OrbStack. If it was convenient to run an individual window from the GUI in a way that felt native and performed well, I might run Emacs and Firefox from Linux instead of on the macOS side. Then I
86.
by pxc 6mo ago
Go to an adblock test page in Chrome and compare it to Firefox with uBlock Origin. Chrome can't block some ads, and some of the ads it can block leaves behind empty containers.
87.
by pxc 6mo ago
That's a really interesting point about Ancient Chinese and other ancient scripts. I'd love to learn more about that. I'm also more curious about tokenizers for LLMs than I've ever been before, both for Chinese and Engli
88.
by pxc 6mo ago
Does this affect editor integrations that use the SDK for ACP support via claude-agent-acp¹, like Zed's Claude integration and Emacs' agent-shell? -- https://github.com/agentclientprotocol/claude-agent-acp&#x2
89.
by pxc 6mo ago
> Are you saying Chinese is more concise than English? Yeah, definitely. It lacks case and verb conjugations, plus whole classes of filler words, and words themselves are on average substantially shorter. If you listen to or read a hyper
90.
by pxc 6mo ago
I was pretty interested until I got to this part: > Another reason that a no-exceptions policy is important: If students with disabilities are permitted to use laptops and AI, a significant percentage of other students will most likely f

More ›