sev_verso

Or you can use any other local model you already have.

VPNs are essential tools against government persecution. Linking identity to a VPN session under any guise (age verification or otherwise) is something out of the playbook of dictatorial states.

I've tried it out with Claude Code on my existing codebase and it seemed to hold its weight (despite being the 2-bit quant). Takes minutes on prompt processing, the actual edits are reasonably quick at above 20 tks.

The good: It succeeded with discovering, applying edits and writing a test for a small task I gave it. The bad: It could not address a small nitpick I had. The ugly: It hallucinated a conversation about "The Duck" that I had with it simultaneously while trying to solve another problem. I can only imagine it's one of examples in the initial Claude Code prompt:

--cut-- However, the user's query is "Can you track these 3 videos here?" which seems unrelated. Perhaps the user is asking if I can track the progress of three videos they are working on?

Let me re-read the user's message. The user said "Source Code" and "The Agent" and "The Duck", it could be video titles. And they are asking if I can track these 3 videos.

?? That doesn't make sense in the context. Could there be two different conversations? --cut--

Exactly my thinking. The same reason why capitalizing and putting the word NEVER in asterisks makes the model more obedient. Or repeating twice. For whatever reason, it just works.

Clean and polished design, concise Tailwind-style API, and last but not least sustained 120 FPS across complex UI.

What's better than Voxtral for locally processed voice input? More competition is always better.

I've been using the editor since the early days and have always been a fan of its visual look and feel, so I was pretty happy to see its UI library open sourced.

I wish GPUI could become the go-to Rust UI library and not just an editor backend.

For that, a couple of changes would be highly desirable: being able to switch the GPU backend from Metal to wgpu (so it could be mixed with vello, for instance), and the ability to integrate into an existing event loop like egui allows you to. If this were easy to do, I would switch from egui in a heartbeat.

Pretty neat and fun to browse. Definitely works in more than the 50 cities advertised. I am curious why it seems to cover some areas but not other? For example, Yerevan in Armenia is not covered, but a ton of smaller towns to the south of it are in.

Also, it may be just a matter of getting used to, but I found the color of the heatmap to be a bit misleading. When you switch to a specific activity - for example, night life, the high concentration areas are in red, while that color means low concentration when on the Liveability tab.

I was still using Skype to talk with my parents when they pulled the plug on it, too. Obviously (and unlucky for Microsoft), that conversation never moved to Teams. Curious how many users they lost with that move.

Interesting. Last time I dealt with currency in SQL (which was well over a decade ago admittedly), we used a fixed point format for those fields.

What do you recon Excel does, floating point math for regular entries and lossless point math for currency? Or just does not rely on floats at all?

Good work on the polished design, it does seem to blend in with the native interface. Did you need to use any private APIs to position or capture the previews?

Personally I hide my dock to reclaim extra space. One app I wish existed is an endless scrollable desktop to plop windows on, in the same manner that Vision Pro does it but without the VR. However, I think that would be hard to impossible to implement with what the native window manager lets you do.

Plays decently smooth on my M4 Max. It's probably still a long way from being a production-ready replacement for meshed environments, but I could imagine a hybrid mode where certain elements like grass and shrubbery are drawn with gaussians, perhaps with support for basic procedural animation. Great work with the playable demo!

That reminds me of split-screen BBS chats. Every keypress is sent immediately to the peer and echoed on your own screen as well. It was quite fun, kind of a collaborative editing session more so than just chatting.