Live data from Hacker News

Running I3 Window Manager on Ubuntu for Windows

brianketelsen.com

91–100 of 177 posts

Re: Running I3 Window Manager on Ubuntu for Windows

#91
post #12

Earlier quoted context omitted.

X was never designed to work like that. I personally run Visual Studio Code (with a patched X lib to allow Electron to start without graphics acceleration) with the "local window manager" (i.e., Windows itself). Works great.

Okay I'll bite. Why are you running Linux vs∆ code in widnLaASaAAqaa@@@@£

(Sorry, I put my phone in my pocket and this apparently happened. Wish I could delete it, didn't even know it sent, not a fan of my tone, apologies.)

Re: Running I3 Window Manager on Ubuntu for Windows

#92

There’s something about “translating system calls” that just makes me wary. It’s the kind of thing that could be proven to work for some common cases while failing in mysterious ways for others (not to mention being hard to debug). Generally, there is a gap between what’s documented, what was thought to be true, and what is. At some point they will produce behavior that just doesn’t match what compiled code expects.…

This isn't meant for productions, so a rare failure scenario shouldn't be catastrophic. Also Windows is inherently designed for such usecase, so maybe it has some reliable process for same.

Re: Running I3 Window Manager on Ubuntu for Windows

#94
I think the real trouble with Windows for many folks is not that we can't figure out how to convert linux kernel calls into NT kernel calls, but that we don't want to convert our linux kernel calls into NT kernel calls PLUS "telemetry" data sent to Redmond. Windows is explicitly non-free (in so many ways), and that's cool and all, but no thanks.

Re: Running I3 Window Manager on Ubuntu for Windows

#95

There’s something about “translating system calls” that just makes me wary. It’s the kind of thing that could be proven to work for some common cases while failing in mysterious ways for others (not to mention being hard to debug). Generally, there is a gap between what’s documented, what was thought to be true, and what is. At some point they will produce behavior that just doesn’t match what compiled code expects.…

Bryan Cantrill has a wonderful talk about exactly this in Illumos that I recommend at every opportunity https://m.youtube.com/watch?v=hgN8pCMLI2U

Re: Running I3 Window Manager on Ubuntu for Windows

#96
post #74
post #28

Earlier quoted context omitted.

Don't write off Outlook without a second thought. It's big with loads of features, yes - but that's what's good about it. If all you need to do is plain text email, sure, Mutt is fine. If you deal with a large amount of email, calendar appointments, tasks, contacts, etc, in a professional role, nothing beats Outlook. There's a reason it's so popular.

Well, that's your opinion. Support that garbage like I do at work, and your opinion will change. Again, popular does not mean good to me. That is the point. I have seen Outlook peg processors and waste disk IO with all those features. I hate cloud services, but they keep such things minimal and some of us pray for mandates of no-fat client email, even to spite myself, because of Outlook and Thunderbird users.

> I hate cloud services, but they keep such things minimal

Have you tried Google's Inbox on an underpowered Linux box? It's horrendously slow, in my experience. Gmail is fine though. Some web based software is really good, but others can definitely be a CPU hog.

Re: Running I3 Window Manager on Ubuntu for Windows

#97
post #87

Earlier quoted context omitted.

Many people who get ransomwared are naive victims of phishing emails, some of which are fairly convincing. There are usually some red flags if you're paying attention, but it only takes one mistake.

You're right. But I'd assume anyone using Linux (and arch at that) would be smart enough to know what's up. No OS is perfect but someone using Arch is just spreading FUD. Windows is a vector not because of security problems but because of its usage ratio. It just doesn't make economic sense to target any other OS plus the fact that hackers are not god, they need time to perfect their skills and they have it on Window…

>Windows is a vector not because of security problems but because of its usage ratio.

I think the publication regarding vulnerabilities in the kernel's handling of font rendering would make you think twice about this assertion. Windows certainly has its share of design flaws.

>That loser on Arch should talk when Linux and macOS are completely impenetrable.

Maybe try again next time without the personal attack. Also, I don't think that because the alternatives are imperfect Windows should be immune from criticism. Microsoft might have come a long way, but running Windows these days is still a security liability in ways that running Linux or macOS aren't. Windows is just a big juicy target and attackers go for the big fish.

Re: Running I3 Window Manager on Ubuntu for Windows

#98

There’s something about “translating system calls” that just makes me wary. It’s the kind of thing that could be proven to work for some common cases while failing in mysterious ways for others (not to mention being hard to debug). Generally, there is a gap between what’s documented, what was thought to be true, and what is. At some point they will produce behavior that just doesn’t match what compiled code expects.…

Have you heard of testing? And have you heard of software verification? Microsoft research is far and away the leader with single person publishing multiple papers in the best conferences- POPL etc. orrying about imaginary Don't worry about imaginary future problesm. Its like me saying- my car is very complex and may crash, I will not drive it. This is how you hold technology back.

This is actually just about the exact opposite of proper testing and verification.

The main problem is the idea of running compiled programs unmodified: programs that had an entirely different OS target when they were built. If this is what Microsoft is doing then they’ve invented an environment that could not possibly have been anticipated by those who built the original programs. Anything could happen. And worse, it’s not clear who to send the bug report to when things go wrong.

Re: Running I3 Window Manager on Ubuntu for Windows

#99

For me, modern virtualization has gotten so good that it's my preferred solution. kvm, libvirt, virtio, and PCI passthrough et al have opened up a whole new world of possibilities. Currently still requires fairly specific hardware and kind of a pain to figure out how to make it work, but it's already very impressive. This is still really cool though.

Thing is you can't run Windows natively that way, it has to be a guest operating system. Then you need multiple videocards to pass through, and the guest operating systems require custom drivers at that.

My preferred setup would be Windows native, with macOS (ha) and Linux guests running with GPU acceleration—preferably in a manner that doesn't necessitate multiple videocards.

Too bad Apple is hellbent on locking macOS to Apple hardware. I'd totally license it otherwise.

Re: Running I3 Window Manager on Ubuntu for Windows

#100
When it comes to finding a decent terminal emulator for Windows: mintty is the terminal you get when you use Git Bash, right? Because in my experience it's better than ConEmu.

ConEmu makes a staggering number of off-by-one errors in putting characters on the screen when I ssh into a Linux machine and use tmux. Vertical separators end up jagged, things that are supposed to fit the width of the screen end up wrapping, my Vim status line says "IISUAL", and so on. But mintty gets tmux right.

Post reply on HN