Maybe someone here has tackled this before. I’m trying to connect Antigravity or Cursor with GLM/Qwen coding models, but haven’t had any luck so far. I can easily run Open-WebUI + LLaMA on my 5090 Ubuntu box without issues. However, when I try to point Antigravity or Cursor to those models, they don’t seem to recognize or access them. Has anyone successfully set this up?
I don't believe Antigravity or Cursor work well with pluggable models. It seems to be impossible with Antigravity and with Cursor while you can change the OAI compatible API endpoint to one of your choice, not all features may work as expected.
My recommendation would be to use other tools built to support pluggable model backends better. If you're looking for a Claude Code alternative, I've been liking OpenCode so far lately, and if you're looking for a Cursor alternative, I've heard great things about Roo/Cline/KiloCode although I personally still just use Continue out of habit.
https://learn.cisecurity.org/benchmarks - this seems broken at least right now. Are these benchmarks on github so that I can download and run it on a linux box?
It’s misleading to cite two government-funded supercomputers as evidence that NVIDIA lacks monopoly power in HPC and AI:
- Government-funded outliers don’t disprove monopoly behavior. The two AMD-powered systems on the TOP500 list—both U.S. government funded—are exceptions driven by procurement constraints, not market dynamics. NVIDIA’s pricing is often prohibitive, and its dominance gives it the power to walk away from bids that don’t meet its margins. That’s not competition—it’s monopoly leverage.
- Market power isn't disproven by isolated wins. Monopoly status isn’t defined by having every win, but by the lack of viable alternatives in most of the market. In commercial AI, research, and enterprise HPC workloads, NVIDIA owns an overwhelming share—often >90%. That kind of dominance is monopoly-level control.
- AMD’s affordability is a symptom, not a sign of strength. AMD's lower pricing reflects its underdog status in a market it struggles to compete in—largely because NVIDIA has cornered not just the hardware but the entire CUDA software stack, developer ecosystem, and AI model compatibility. You don't need 100% market share to be a monopoly—you need control. NVIDIA has it.
In short: pointing to a couple of symbolic exceptions doesn’t change the fact that NVIDIA’s grip on the GPU compute stack—from software to hardware to developer mindshare—is monopolistic in practice.
The problem I see with investing in such technologies:
1. Satellites are too far to fight anything on ground (Power per unit area (i.e., intensity) decreases as the square of the distance)
2. If the Satellites are relaying to things on the ground, they are also relaying their location (easy adversarial targets)
3. In a war (they mention Ukraine in the article), first thing that is toast is these satellites.
I don't think this is the right replacement for GPS. Perhaps someone here can correct me if I am wrong?
It's probably worth considering the military and non-military uses separately.
The USG military uses is attractive not as a replacement of GPS, but as a supplement/complement. If they could truly manage to use the same receivers, then this provides an extra layer of redundancy. There are 32 GPS satellites in the current constellation. Being in MEO means you need pretty beefy ASAT to take it take them down, but we could assume that China could pull it off. Xona's constellation would add redundancy (splashing 258 is just a lot more targets).
For non USG uses, I imagine Xona is making two different pitches.
a) You can achieve GPS+RTK level accuracy without needing RTK base stations.
b) Increased jamming/spoofing resiliency, intended for short of war (aka hybrid war/grey zone) situations. For example, I imagine Xona will attempt to setup a private encrypted signal which they'll sell to friendly/allied nation airliners and similar industries.
Looks like openrouter api can be self-hosted, which means you should be able to run this locally. If anyone is able to run this with ollama, please do post how you did that? :)
The openrouter api is the same as the openai api, so you should be able to use the openai api compatibility built into ollama after updating the url in /src/acmsg/constants.py
The lack of a readily available, installable package (pip install pygraph - has no relation to this paper as far as i can tell) makes it difficult to fully assess the reproducibility and practical applicability of the work.
I think that might be just a feature of the catalyzex platform for papers with no linked code yet that might internally add a +1 to code requested on their db and thats it
some times papers come out a few weeks before code when its bleeding edge
I actually agree with the article, but this piece feels like a conflict of interest — a library conducting research on how libraries improve everything.
Maybe it's hard to justify studying the positive affects of libraries (and third spaces generally) within a community to a non-library inclined audience.
Anecdote: While chronically underfunded to a overwhelming degree, local libraries and their staff in my area have consistently stepped up to the plate by trying to support the communities. Most notably when schools around here stopped offering any sort of afterschool/enrichment programing internally for kids, the libraries basically just added to their model.
I believe there's a balance to be struck between writing your own code and relying on dependencies. If Armin (the author of this post) is the one writing the code, I would highly recommend using that dependency!
I don't know of one. However, Python's socket API is mostly a thin wrapper around the C socket API. Assuming you know Python, it should be easy enough to learn from a tutorial like this one without a Python-specific guide. Note the matching function and constant names:
Bug fix request: On the webpage (https://shapecatcher.com/unicodefonts.html) - the download links for the fonts is broken. Perhaps fix? Also, does it make sense to combine the OTF files into a single file that shows the match you make instead of having it broken down into multiple files?
