ryanhecht
ryanhecht on Hacker News
Kid Pix 3 and Kid Pix 4 were foundational computing experiences for me.
I remember wasting way too much time with the rudimentary speech-to-text
Notably, the limit per employee != actual average usage per employee!
Something I learned the hard way: it's a bad time if you are a Google Fi customer and your Android phone with an eSim becomes inoperable while traveling :)
I can't wait to run Stemdeck on my Steam Deck and trigger it with a button on my Stream Deck!
This just feels like good educational practice too? I remember really enjoying TA'ing as an undergrad for courses I took the previous year, because THAT'S when I felt I _really_ learned the material.
Likewise, when I was a lab consultant, I felt like advising others was when the concepts of Linux/systems administration _really_ sunk in.
I always crammed the bare minimum for any kind of assessment (by the end of college I was actively avoiding any classes with final exams), but when I need to be able to teach someone else or explain my understanding in a live environment? That's when I _actually_ learn.
I truly think you folks did the account transition in the most user-friendly, considerate way you possibly could. There was SO MUCH TIME to migrate and SO MANY WARNINGS about it needing to be done. I'm sorry you haven't been appreciated for it. Well done!
I knew if I scrolled through the comment section I'd find geerlingguy!
Hi Alex! I'm Ryan, I run a Minecraft theme park recreation server. While our gamemode is very far from vanilla Minecraft (it plays more like a MMORPG), I've been interested in standing up a model like this to help automate agentic testing of new software features in my monolith Bungee/Paper codebase.
In its current state, do you think it can handle non-vanilla Minecraft tasks like interacting with custom UI's, running commands, or doing other unusual tasks in the world?
That's what we've spent the last five months doing! Have you tried the Copilot CLI recently? We've onboarded loads of feedback from Microsoft devs who were switching from Claude Code -- I'm proud of how far the team has come!
This announcement comes at a time where Copilot CLI usage has been greater than Claude Code usage at Microsoft for several weeks; we've been winning hearts and minds!
> Microsoft internally is actively hostile to cross-org collaboration
The Copilot CLI has ushered in the beginning of a change in this dogma -- I've helped dozens of Microsoft engineers get access to GitHub source code so they can contribute to Copilot CLI! It's fun to subvert expectations when a Microsoft IC pitches an improvement and I can respond with "submit a PR!"
> The developers voted with their feet and didn’t use Copilot.
This was true in January -- since then, the Copilot CLI team has spent countless hours with engineering leaders and the biggest Claude Code users at the company to understand Copilot's shortcomings, define evals to properly test them head-to-head, and close the gap between the products.
The result? Claude Code usage was organically decreasing and Copilot CLI usage was organically increasing -- when this announcement was made, internal Copilot CLI usage had been greater than Claude Code usage for weeks!
Something cool that I've always liked about working at GitHub is how much of the company _runs on GitHub_ -- A lot of teams, even non-technical teams, have their own repos just to organize docs/SOP's/designs/etc like a traditional knowledge work company might use a Sharepoint
In my case, it helps improve my ADHD-affected executive function. It lowers the barrier to entry to ingesting information into a system I know I can extract it from later. It gives me peace of mind to know I can ramble semi-coherently at the speed of conversation and know that the salient points are being captured.
Local models will continue to improve, if your concern is privacy -- already they do a decent enough job at interacting with a well-schema'd PKM
Copilot CLI PM here: I'm super proud of our team's velocity and vision as we took this from our admittedly bare-bones initial public preview back in September to what I believe is an industry-leading agentic harness.
Hey! I'm a PM on the Copilot CLI team. This sounds like a bug, we should follow the same premium request scheme as the VSCode extension! If you still have the session logs kicking around, can you email them to me? It's my hn username @github.com
The Copilot CLI team has been making great strides towards improving our agentic harness! I'm curious, what have you found are the biggest shortcomings with it these days?
Give Copilot CLI a try if you haven't in a while! The team's been working really hard to improve the harness, and we're taking as much community feedback as we can get! Let me know if you run into any problems :)
https://github.com/github/copilot-cli
Maybe someone will finally build my dream: a WSL distro that I can also dual-boot natively. I'd love to switch between bare-metal Windows with WSL and bare-metal Linux with virtualized Windows at my leisure!
Maybe someone will finally build my dream: a WSL distro that I can also dual-boot natively. I'd love to switch between bare-metal Windows with WSL and bare-metal Linux with virtualized Windows at my leisure!
Definitely makes sense!
> In fact, even if you ungzip them manually, as of today nerdlog doesn't support more than 2 files in a logstream
Ah, interesting! I read the limitation as "we don't support zipped files," not "we only support two files!"
Best of luck, this is neat!
I definitely intend on playing around with this later! I see that [gzipped log archives aren't supported](
https://dmitryfrank.com/projects/nerdlog/article#depends_on_...), minimizing the use case for me personally. You've at least thought enough about that to bring it up as a limitation you think people will call attention to -- any plans to eventually support it?
I feel ashamed to admit this in public, but...me too. The barrier to entry is so much lower than therapy, I feel less anxiety about explaining my situation correctly, and I can quickly start over with a "new therapist" whenever I want.
Is it a replacement? No, of course not. But boy if it isn't a big help.
John Hench (widely considered the coiner of the "go away" colors as it relates to WDI) brings up the concept of "go away" colors in the color theory chapter of his book "Designing Disney" when discussing how colors establish the identity of a space (this is partially quoted in the Wikipedia article):
> One particular challenge in designing the interior of the [American Adventure pavilion at EPCOT Center] was the metal safety railing we had to include by law, which, being out of place in the middle of a Colonial staircase, undermined the identity of the main entrance room. We chose a neutral gray-brown for the railing, a 'go away' color that did not call attention to itself, even though it was entirely unrelated to the Colonial color scheme. In this way ,we got a functional safety railing that satisfies legal requirements without conflicting visually with the Colonial theme
https://archive.org/details/designingdisneyi0000henc/page/11...
Given how he talks about color in the rest of the book, I think Hench and other early Imagineers were coming at it from film/scenic design perspectives rather than scientific "here's how camouflage and blocking patterns work" perspective. I agree though, I'd love to have someone connect this artist-minded explanation with a scientific one!
> about 50% of its marketed capacity
> they will offer a 50% discount
I feel like I would need more than a 50% discount in order to go along with this...needing to charge/swap a battery more frequently is added cost! :p
Reminds me of a joke in the musical "How to Succeed in Business Without Really Trying" (written in 1961):
PETERSON
Oh say, Tackaberry, did you get my memo?
TACKABERRY
What memo?
PETERSON
My memo about memos. We're sending out too many memos and it's got to stop!
TACKABERRY
All right. I'll send out a memo.
Ugh, and some of the rows of that table are "sets of models" while some are singular models...there's the "Flagship models" section at the top only for "GPT models" to be heralded as "Our fast, versatile, high intelligence flagship models" in the NEXT section...
...I like "DALL·E" and "Whisper" as names a lot, though, FWIW :p
But what would we call that model?