vunderba

10 ms·

181.
by vunderba 3mo ago
From the article: > Steve suddenly got more intense. "Rectangles with rounded corners are everywhere! Just look around this room! Do you want to know why there are a lot of rounded rectangles on things in the physical world? They’
182.
by vunderba 3mo ago
Neat! Maybe consider doing a longer postmortem/blog post on this. Small bit of feedback: it might help to add an explicit small (i) icon for areas which contain tooltips especially for interfaces (phones) that don't support hover
183.
by vunderba 3mo ago
Relevant Bojack Horseman clip: https://imgur.com/a/hdTA32S
184.
by vunderba 3mo ago
Thanks! I completely forgot about RHIDE~ I remember seeing a friend of mine using it and mistakenly thought it was Borland's Turbo C++ since they look so similar.
185.
by vunderba 3mo ago
Hmmm... sorry about that, I'll try to put a more descriptive error in place. The 2D/3D CRT effects use shaders and require WebGL; that's probably what's causing the issue. Can you try hitting this and seeing if you get
186.
by vunderba 3mo ago
Yeah, good point. To expand on that: this actually started out as an animated typist for classic literary works in the public domain (some of which are in the Library), like The Machine Stops or The War of the Worlds . That’s why, if you
187.
Show HN: Watch random code typed out on an MS-DOS IDE(hackerman.specr.net)
17 points by vunderba 3mo ago|8 comments
188.
by vunderba 3mo ago
The example linked in the post has a lot of issues. It seems to be case‑sensitive (it errors on `Screen` instead of `SCREEN`), a simple `GOTO` loop can lock up the VM (in real QB you can break out with Ctrl+Break), doesn't attempt any
189.
by vunderba 3mo ago
Yeah I hear you. I’ve done some experimentation with infinite image generation, and it’s very difficult to pull off without some level of human supervision. Even if a model can technically generate infinite content, you’ll find that it even
190.
by vunderba 3mo ago
Nice job. How are you handling outpainting? Outside of using an open-weight model, I imagine it's going to get pretty expensive quickly with something like NB or GPT-Image. Feedback: Add an option to hide user tags and only show the di
191.
by vunderba 3mo ago
None of that is necessary. It’s like people who install Obsidian and then spend weeks obsessing over the “top ten must-have plugins” for the ecosystem instead of just using it as it’s intended: as a tool to take notes. The purpose of a codi
192.
by vunderba 3mo ago
> To get the cells to stretch dynamically based on their velocity, I use a technique called Domain Warping. Basically in the fragment shader, the fragmentPosition is adjusted inwards/outwards depending on the current velocity, whic
193.
by vunderba 3mo ago
There are basically two answers to this. One option is to orchestrate more agents across different repositories or projects at the same time, so there’s essentially less downtime while you’re waiting between tasks. But I really dislike that
194.
by vunderba 3mo ago
Ideogram4 is great! Just make sure that you either use some kind of preprocessor to convert your natural language prompt to the strict structured JSON format otherwise the results can suffer. I personally use Qwen3 8b LLM for this purpose,
195.
by vunderba 3mo ago
Most of the disappointment around the previous open‑weight BFL release (at least for Flux.2 [dev]) came down to two main issues: • It initially required significantly more VRAM and was much slower than alternatives released around the same
196.
by vunderba 3mo ago
I run a fairly high-traffic site for generative image models focusing on complex prompt adherence. Flux.2 doesn’t score anywhere near SOTA proprietary models. Klein 9b is decent for image-to-image, but when used for pure generative purpo
197.
by vunderba 3mo ago
Whenever I think of a true musical multitasker, I picture Dick Van Dyke’s character in Mary Poppins . There's an element of multitasking that is (while not exclusively) inherent to piano definitely is an important focus. For instanc
198.
by vunderba 3mo ago
Honestly not sure. I think Fugu tries to leverage multiple models simultaneously where something like NotDiamond is more about picking the most optimal singular model. This space is so crowded it feels like I see a new "model router&
199.
by vunderba 3mo ago
It feels a bit more like Fugu to me, which acts as a multi-LLM orchestrator (though I think Fugu combines open- and closed-weight models), but without being able to see the “secret sauce” behind how any of them decides the number of "p
200.
by vunderba 3mo ago
The two major image generation models are both by Alibaba - Z-Image and Qwen-Image-2512/Qwen-Edit-2511. They are both decent but have different use-cases. Unfortunately with the recent Qwen-Image 2 and now 3 it doesn't seem likely
201.
by vunderba 3mo ago
This. The idea of having some kind of cryptographic proof of ownership to a JPG copy of an original piece of artwork was just about the silliest thing I’d ever heard when it first started going viral. I (and most others) support artists by
202.
Show HN: Infinite Programmer ASMR(hackerman.specr.net)
3 points by vunderba 3mo ago|0 comments
203.
by vunderba 3mo ago
I called it out a while ago but the largest market to get destroyed by gen AI was never the professional market - it was the gig economy. The quality of generative AI may not be top-notch, but it presents an easy drop-in replacement for one
204.
by vunderba 3mo ago
There’s a related benchmark along these lines that uses an LLM with a discrete “instructive graphic language” to get models to generate Minecraft voxel structures with varying degrees of success. https://github.com/Ammaar-Al
205.
by vunderba 3mo ago
I’d probably add “Save Early, Save Often, Save Multiple.” because there are so many ways to accidentally softlock yourself making the game totally unwinnable without realizing it until hours later. There’s a reason most of us as kids star
206.
by vunderba 3mo ago
Flux.2 [dev] was a pretty decent model when it released back in Nov 2025 but was an absolute GPU memory hog to run on local hardware (even with the Turbo LoRA). That and it's frustrating licensing were the driving reasons why it got he
207.
by vunderba 3mo ago
I have a couple of enthusiast friends who use CompactFlash-to-IDE adapters, and that’s one of the upgrades that almost makes it worth it. You still get the full retro experience (well except for the actuator arm seeking sounds), but on the
208.
by vunderba 3mo ago
Totally agreed. Related, I actually wrote a satirical blog post on the brilliant but equally frustrating experiences of playing King's Quest IV: The Perils of Rosella when I was a kid. Just call me Captain Ahab because I hated that d
209.
by vunderba 3mo ago
It's a proprietary model as a service - it doesn't require anything outside of a browser. But even when they released open-weight (the original Qwen-Image) it's a diffusion model and can't use custom font files.
210.
by vunderba 3mo ago
Well from the images I've seen they seem to be pushing text-heavy "infograph" capabilities pretty hard. Outside of that, I have serious doubts that it'll be better than other proprietary models like NB Pro and gpt-image-

More ›