Sounds like a bad joke joke but the EU governing council wants to once again legislate #chatcontrol for it's 350 million citizens. According to @echo_pbreyer (EU parlamentarian) the legislation wants to force messengers to prevent users from sending images or videos unless they agree them to be scanned, AI-analyzed and potentially reported to police. They apparently want to bring it up for a council vote in two days and get it through parliament in June. German link https://www.patrick-breyer.de/lass-dich-ueberwachen-eu-rat-will-sich-auf-chatkontrolle-mit-zustimmung-der-nutzer-einigen/
@delta@delegatevoid@echo_pbreyer begs the question what happens if I have a messenger that just sends media as text messages, base64 encoded, and a client (plugin?) that detects those and decodes them back to visible media. Does that count as text then, which doesn't need to be analyzed?
I've been experimenting with using a fast approximate voronoi diagram rendering technique as a point cloud renderer. This is a voronoi diagram with a million cells :3
@ben at some point I just started to only stand at work. So I began to only sit after work and during breaks. It's not perfect for my feet (they feel a bit "thick" in the evening), but my fitness has improved and I feel more agile in general. Plus, I can't imagine working in an office anymore, because I don't want to sit the whole day
@ben I guess I would just use a normal distributed randomness and adjust the success threshold depending on the situation. Like, start with a true 50:50 rate, but increasingly shift towards 51:49 etc
Serious question - are there any#GameDev people out here using USD? What do you use it for?
So many folks pushing the universal scene description standard but I can't think of even a single use for it.
glTF 2.0 seems a much better standard to me - it's what I've moved to exclusively.
(I know they're for different use cases - but USD as "an entire scene" is better handled by your engine's native map/level format)
@Longplay_Games afaik there are a few very solid pipeline systems like omniverse that some huge companies use. Probably mostly companies that have a single team for levels and another team to manage that everything comes together (some engines aren't that good at editing, some even lack an editor completely). But I never really met a single person who uses usd, but I can ask a more experienced colleague who might have used it in the past.
@Longplay_Games I was also talking about games. Think of different engines with bad editors, because they focus on just good engines. Maybe someone like naughty dog or ubisoft
@Longplay_Games in that league, mostly their own engines. There are some public presentations with screenshots and stuff, if you want to invest the time to identify software, but you can probably just google. Bethesda is still using creationkit, afaik, and why not if they are the masters of using it...
@Longplay_Games though some companies started using UE, like cdpr. It's just too many expectations nowadays, so you have to invest decades of time or expect your own people to work for inhumane hours and weekends (talking about AAA+, not indies)
@oblomov@TomF@aeva those bindings show me that the standard for copying is not copy&past but cut&paste. Same on plan 9: copy is a cut followed by an immediate paste.
@gamingonlinux I can recommend Sable. It's a cute exploration game and the first game in years that I enjoyed playing for quite some time. It has some bad performance, sadly (at least on ps5), but it's still a great game!
@chakuari speed-wise it's like unix should be: many small programs that do their task and do it well. We have a shell that's comparatively slow, but we never hit any bottlenecks with it, although half of our boot process is managed by it.
@chakuari portability-wise: everything is cross-compilable. Our kernel is portability-oriented, our libraries are, our programs usually are just portable as is (just compile them with the compiler for the target). The build system is built around that. Our dynamic namespaces help to execute the right binaries for our platform (no need for complex $path variable hacking)
@chakuari I guess you already read a lot about it? Then my guess is, just try it. There's plenty of ways to try it and to learn more about it. But keep in mind that plan 9 is not a unix and many design choices are based on a fully networked environment (a single plan 9 system may consist of many machines)
@chakuari plan 9 is designed as a system to be used, so you will find a lot of original unix tools there. You can try to install it on your eeePC, but we don't have too much hardware support, so you might have better results on thinkpads or a VM. If it doesn't work on your eeePC now, it might work in the future. We have a few people who are capable of writing drivers and stuff, we even have some support for wifi.
@chakuari troff is one of the tools that's definitely bundled with the system. I use it a lot to write documents, and it's the man page driver, like on unix. I know there's some IME support (for CJK, at least), but I'm not one of the users, so I don't know the details. If something is missing (or you desperately need to try out or netsurf port), you can usually just get the source (with our git implementation) and mk it (we have a much simpler make alternative called mk)
@chakuari if you have any questions, it's fine to ask here, but much easier to ask on our 9fans discord server, the mailing lists or various irc channels
@chakuari for 9front the go-to channel is # cat-v at oftc, iirc. There are a few channels # 9fans, # plan9 and # 9front, but for 9front it's best to ask in cat-v
Allowing '..' (dot dot) in a path was a mistake and removing it would solve more security problems than using memory safe languages
Now that I have your attention
In security we keep hearing about memory safety and how we just need to stop using C and how it fill fix a lot of problems
This is true, but next time read about getting rid of C, I want you to think about removing .. from being supported. It's an easier problem to wrap our heads around, possibly more useful, and probably easier to do
@utterfiction@joshbressers on plan 9, programs like web servers do bind the to-sandbox directory to / within the program. Then the relative paths are the absolute paths automatically, and .. can't break out of the sandbox path. It's actually quite easy to do that if you don't need access to files outside the sandbox directory (you can still find a way around that within your program, but that depends on the program flow)
