navanchauhan

4 ms·

1.
by navanchauhan 3mo ago
apg?
2.
by navanchauhan 5mo ago
Technically, it is on GitHub in my swift-omnikit library. But, this library is only meant to be consumed by me for now. I plan on separating out the UI portions to its own repo and then polish it up
3.
by navanchauhan 5mo ago
I am finally getting close to my vision of `write once run everywhere with SwiftUI`. The idea is to create a drop-in replacement called OmniUI which will have different renderer backends (I currently have TUI w/ notcurses and Adwaita&#
4.
by navanchauhan 7mo ago
Not affiliated with Sesame, but this is what the realtime models are trying to solve. If you look at NVIDIA’s PersonaPlex release [0], it uses a duplex architecture. It’s based on Moshi [1], which aims to address this problem by allowing th
5.
by navanchauhan 8mo ago
I have been building a drop in replacement for SwiftUI that can render with different renderers (TUI for now and GTK/Adwaita very soon). This will be such an awesome demo use case for it. Congratulations on getting this out!
6.
by navanchauhan 8mo ago
I thought YouTube had native RSS feeds for channels?
7.
by navanchauhan 8mo ago
> What do you show to new folks when they join your team? I think this is an interesting question because we have not fully figured out the best way to onboard people to our codebases. Each person is responsible for multiple codebases (y
8.
by navanchauhan 8mo ago
The model could start building on top of things it had successfully built before instead of just straight up exponential error propagation
9.
by navanchauhan 8mo ago
Using models to go from spec to program is one use case, but it’s not the whole story. I’m not hand-writing specs; I use LLMs to iteratively develop the spec, the validation harness, and then the implementation. I’m hands-on with the agents
10.
by navanchauhan 8mo ago
This was an experiment that Justin ran: one person fresh out of college, and another with a long, traditional career. Even though all three of us have very different working styles, we all seem to be very happy with the arrangement. You def
11.
by navanchauhan 8mo ago
No. But, I do ask questions (in $CODING_AGENT to always have a good mental model of everything that I’m working on though.
12.
by navanchauhan 8mo ago
https://github.com/strongdm/attractor
13.
by navanchauhan 8mo ago
I think this comment is slightly unfair :( We’ve been working on this since July, and we shared the techniques and principles that have been working for us because we thought others might find them useful. We’ve also open-sourced the nlspec
14.
by navanchauhan 8mo ago
You're absolutely right ;) Your intuition/thinking definitely lines up with how we're thinking about this problem. If you have a good definition of done and a good validation harness, these agents can hill climb their way t
15.
by navanchauhan 8mo ago
It’s part of the “lore” that gets passed down when you join the company. Funnily enough, the marketing department even ran a campaign asking, “What does DM stand for?!”, and the answer was “Digital Metropolis,” because we did a design refre
16.
by navanchauhan 8mo ago
Domain Model ( https://strongdm.com )
17.
by navanchauhan 8mo ago
I would spend those $100 on either API tokens or donate to a charity of your choice. My interview to join this team was whether I could build something of my choosing in under an hour with any coding agent of my choice. No local model out t
18.
by navanchauhan 8mo ago
I think corporate incentives vs personal incentives are slightly different here. As a company trying to experiment in this moment, you should be betting on token cost not being the bottleneck. If the tooling proves valuable, $1k/day pe
19.
by navanchauhan 8mo ago
A few overnight “attractor” workflows serve distinct purposes: * DRYing/Refactoring if needed * Documentation compaction * Security reviews
20.
by navanchauhan 8mo ago
(I’m one of the people on this team). I joined fresh out of college, and it’s been a wild ride. I’m happy to answer any questions!
21.
by navanchauhan 9mo ago
I use a self hosted instance of [Isso]( https://isso-comments.de ) on my static site
22.
by navanchauhan 9mo ago
https://web.navan.dev (Generated from my GitHub repo navanchauhan/navanchauhan.github.io ) Web because there is also gopher://gopher.navan.dev
23.
by navanchauhan 11mo ago
I believe this is because you are logged in. You can generate a free API key (with very low limits) through Google AI Studio and use that to test it. In an ideal world, this workaround would not be needed ¯\_(ツ)_/¯
24.
by navanchauhan 11mo ago
Are the primitives not already there?[0] I don't see why we cannot build an app that when connected to an external monitor switches to a "Desktop Environment". Maybe, even a hacked version of UTM[1] that exposes a fully funct
25.
by navanchauhan 1y ago
Endpoint Security Extension and Network Extension
26.
by navanchauhan 1y ago
I used `sandbox-exec` previously before moving to a better solution (done right, sandboxing on macOS can be more powerful than Linux imo). The way `sandbox-exec` works is that all child processes inherit the same restrictions. For example,
27.
by navanchauhan 1y ago
At least on macOS, there is a third way where you can control the network connection on the PID/binary level by setting up a network system extension and then setting up a content filter so you can allow/deny requests. It is prett
28.
by navanchauhan 1y ago
Yes, but at the same time having the 1 million context enabled is nice because the model is aware that they have more context left and actually perform better. [0] [0] https://cognition.ai/blog/devin-sonnet-4-5-lessons-
29.
by navanchauhan 1y ago
You have to specify `/model sonnet[1m]` to get the 1 million context version
30.
by navanchauhan 1y ago
Ultrathink still works

More ›