Just downloaded CoMaps. Why does it start with a zoomed out view of the entire world, especially given that I gave it location services and it shows my current location (in North America). The very first UX experience this app gives is "I am primarily for people interested in maps and geography"
Contrast this with Apple Maps - when you open it, there are 4 big tap controls for actions like "Home" "Work", a search bar, and a map that covers a 1-mile radius around you .
I'd encourage your UX flow to go something more like: request location services > if granted, immediately start downloading their local tileset in the background > zoom to a 20-mile radius around the user
What always gets me about these red team attacks is the same thing that gets me about internal phishing test emails.
My company sent an internal phishing test last week. Several people immediately reported it to a cybersecurity engineer, posted about it in Slack, saying they were surprised that such a sophisticated phishing attack was happening.
I too was surprised - Google is usually much better about catching these kinds of things in the GMail filter before they get through. Oh well, sometimes one slips though. Reported it and moved on
Come to learn that the only reason it made it through is because we let it through _on purpose_.
By analogy to these red team attacks: _theoretically_ someone could rent a car, pose as an employee, and set up a Raspberry Pi in the network.
But who would go to all that trouble?
Theoretically, someone could craft a perfect phishing attack, but who would go to all that trouble?
Spray-and-pray, low precision, high surface area, attacks are the ones I end up reading about.
The only reason this attack vector was open is because the red team stood to gain a massive benefit from succeeding in the attack. What real-world actor would go to the trouble and stand to benefit as much?
Maybe we ought to be making little hardware passthroughs that plug into the headphone/mic jack and control them with idk the Caps Lock signal from USB HID to start recording.
It's very cyberpunk eventually...the human operator of the console needs to be able to see and hear the screen and sound, there will always be an interface that can be adapted to a machine, however low-fidelity
Been a huge fan of LibrePCB and especially the improvements in the past 3-4 releases. I have found it generally to be more performant than KiCAD and for building my own footprint and library tooling alongside, LibrePCB's library design paradigm works a lot better to me than does KiCAD's.
I wish it were easier to extend the way the KiCAD plugin system is, but I also have a lot of respect for the developer's approach. High-performance, choose features carefully, everything can be done local or with Git with no cloud dependencies.
I don't really understand what this is for... there is a lot of ML-researcher talk on the GH page about the model architecture, but how should I use it?
Is it a replacement for Kimi 2.7, Claude Haiku, Gemini Flash 3.1 lite, a conversational LLM for the situations where it's mostly tool-calling like coding and conversational AI?