bisby @ bisby @lemmy.world

Posts

0

Comments

225

Joined

2 yr. ago

Got it. Linux is not VR ready until it supports discontinued headsets. that were previously at 10% of the 2% market, but are now even less (because it's discontinued, and thus only going to continue to shrink).

The number of different branded headsets using WMR doesn't make it significant in any way. Based on Steam hardware survey, WMR headsets only account for 2.84% of VR headsets. Index, Quest 2, Quest 3 account for ~70% of VR headsets in use, and they all work on Linux. Index just naturally in SteamVR and it's my understanding that setting up ALVR for the quest ones isn't that tricky (but I've also never tried). And much of the remaining 30% other headsets work with ALVR too.

And the point of comparing things to Windows, is that if we're stating "Linux isn't ready for gaming because not every VR headset works", then by that definition Windows isn't either. Which you probably agree with, but generally speaking "people" / society view Windows as ready for gaming despite it not supporting every headset.

It's basically getting into the "Fortnite doesn't work on Linux" type of situation now. Some things are just never going to work, and it's because of the creator of those things and not Linux itself, and who cares. Even if the things that don't work are popular, that doesn't mean that on the whole, the OS isn't ready.

Also, according to steam only 1.9% of accounts have a VR headset. That alone makes VR an edge case. but 2.84% of 1.9% is 0.05% of overall steam accounts using WMR. I think Linux can be ready for gaming without WMR support.

Counter point: VR is working. It's not working for your specific hardware and use case.

My Oculus Dev Kit 1 and 2 don't work properly on Windows anymore. Does that mean Windows isn't ready for gaming because my specific VR hardware doesn't work on it? Or does it mean that "VR ready" doesn't have to include every VR headset.

I recently switched my VR PC from Windows to Bazzite. No compiling necessary.

Communication is explicitly a more than one person endeavor. If the other party isn't willing or able to use signal, then sms might be the required option. Signal removed their SMS functionality from their app.

There are perfectly valid reasons to use Google messages instead of signal.

This comes a year and a half after they resorted to disabling Wayland support

Yeah. A lot of progress has been made in the past year and a half. This is a clickbait headline. It's not like last week they were like "this is super broken... oh well shipping it anyway." It feels like pointing out their previous criticisms is almost trying to call them out as hypocritical or something.

It was previously broken. They said it was broken. And now it's fixed, and they re-enabled it as the default. There's no bigger story or drama around their previous comments.

There was a checkbox to mute the noise on the dialer. We had 2 phone lines, so I've been terminally online since windows 3.11, and I very rarely heard heard the dial up noise

EGS finally has a shopping cart now, so it's basically equal now, right?

Unless it dives with you in its mouth.

I'm surprised their website works and uses web standards and they didn't just NACK everything about HTML and go off and do their own thing.

SteamDeck + SteamOS has the same compatibility and playable games as ArchLinux/Fedora/Ubuntu/desktop linux.

PS5 and FreeBSD do not have any overlap there. If PS5 had no proprietary layers on top of BSD (or the layers were otherwise accessible to the public and could be installed on top of BSD by end users), then you would be right in a way.

But any game released for SteamDeck, I can run on a linux desktop without any special tweaking, so they are related enough. Any game released on PS5 will not run on a BSD desktop, so clearly they are not related in the same way.

SteamDeck being popular and mainstream means that more games work on SteamDeck, and more games working on SteamDeck means more games working on Linux. (Another aspect which is not true of PS5 and BSD).

So having people in the industry, who will potentially be involved with the studios, having (and enjoying) a SteamDeck means studios will hear things like "I hope I can play this on my SteamDeck" from people involved in the creation of games ... which helps push forward Linux gaming. Even if only a little.

Tory Bruno, CEO of ULA wears a cowboy hard hat. Clearly this is how you become rich.

Random broken things and weird tinkering to get some things working. And even when they work, not quite as good as windows.

Most overlays don't work because they are tied to windows specific windows capture things. On KDE wayland, the default "view desktop" from SteamVR doesnt even work.

But if youre looking for some very chill things, it's generally passable. I've been playing beat saber, which is fast paced (at least for the hand tracking) and proton handles it perfectly. From what I can tell, proton can handle VR games just fine, there's just some work to clean up the SteamVR interface in general.

I'm still delusionally hoping that the Valve Deckard is shipping soon and that when that drops, there will be a big SteamVR 3 linux update (kinda like how SteamOS 3 came out with the steam deck), and the headset will run linux itself so naturally they will have to ship all their linux VR improvements, and we'll see linux VR suddenly become mega viable.

tl;dr - working, depending on your level of tolerance for slight jank, and what games you want to play.

I have a separate PC for VR (with an old Vega64 in it) on my valve index. Just a week or two ago I got fed up with something on Windows 10 (i think it was trying to get me to upgrade to windows 11 maybe?) and installed bazzite.

I started with HORRIBLE performance issues. Like could barely run beat saber smoothly issues. And then I changed something minor around (Disabling the VR Home was I think the biggest thing, it's like it was constantly running in the background or something), and ran some script i found online (https://gist.github.com/galister/a85135f4a3aca5208ba4091069ab2222 - i think it was this one, but disclaimer, i have not looked deeply at what this does, I was running this on a fresh gaming only distro so I had nothing to lose), and suddenly performance was just fine. I'm sure this isn't motion smoothing, but going from stuttery to smooth made me think of this. And a Vega64 is pretty old, and pre-dates any modern "rdna" AMD improvements. But it is GCN at least. I might

Audio switching on Bazzite does work. In fact it works more reliably than it did on windows for me. I feel like "using a gaming dedicated distro" can go a long way in making gaming things work, and this is a dedicated gaming PC. YMMV

Base station auto sleep mode does not work, but https://github.com/ShayBox/Lighthouse this CLI script can solve that. Just set something to run lighthouse --state on and lighthouse --state standby and you're good.

Performance is generally worse than windows, and some things won't work (OVR toolkit requires some windows specific things, so naturally doesnt work). But on windows, the first time i launch steamvr for any session (its not just per boot, its just more like "if the headset has been off for more than an hour"), the headset screen wouldnt turn on. Put the headset on, i can see the tracking is working via the mirroring on the display, but the headset doesnt light up. "Restart headset" and then it works. Every time. Doesn't happen on linux. And with bazzite, the power button does a quick sleep just like on a steam deck, and the index still works reliably after the computer wakes from sleep.

I don't do a lot of VR, i just regularly play beat saber for exercise. And it works well for that. I'm perfectly happy sticking with bazzite for VR workouts. I havent really tried any other games, but would be willing to test drive anything for compatibility if anyone cares about something specific.

https://en.wikipedia.org/wiki/Fast_inverse_square_root

even if you can figure out specifically WHAT a function does, it's not always clear WHY a function does, and honestly, if this function wasnt labeled in the code, no way in hell would I know what it does.

It has an entire wiki page dedicated to explaining it, and it involves enough math that most people wouldn't be able to follow along.

Nothing this atrocious lives in any current codebases I work on... but if you work at an old enough company, some of the load-bearing code will be tricky to figure out what is calling it, but also it was written in a time where little hacks were needed to eke out performance.

You only have to experience it once for it to be a memorable enough thing that you will cite it for the rest of your days.

Or more realistically, it IS comprehensible, but the level of effort necessary to comprehend it is not worth it. So you leave it as "undecipherable" and move on.

The 4 freedoms of open source per RMS apparently dont include calling your project whatever you want to call it. You have to run that past him.

The first few years of self hosting tend to have a lot of experimentation, so the overlap is natural.

I'm hitting my grumpy old man phase of self-hosting where I want my Minecraft server and Jellyfin to to be stable so I don't have to hear about it from my family. So ironically, my setup is starting to look more like an overkill setup because I want to self host with stability instead of tinkering around to see if I can run a different server distro, etc. My home lab years got me to find a real nice base, but now I just add things to that base and I don't mess with the formula I have.

IMO the distinction is that if you are doing it for fun (or education) and could afford to lose any service you run for an extended period, you're home labbing. If you are doing it for cost savings, privacy, anti-capitalist, or control reasons and the services are critical and need to stay up, you're self-hosting.

tl;dr - experimentation vs utility

XMPP has been an option for decades, if your contacts aren't using it by now, they arent going to. And with communications tools, both parties have to agree on a tool. Even if one party doesn't care about privacy or security.

Raw brute force security isn't the point most of the time, and ease of use and simplicity of setup are going to be major factors in adoption. Signal is much easier to get started with for most people than XMPP.

The point of a terminal like this isn't necessarily to have more features. I have the tabs turned off (I also just use tmux). The point is to render smoothly and look/feel nice.

Some people would rather spend a lot of money on a nice pen. It still is just a pen that writes. No additional features over a 25 cent Bic pen. But the smoothness of the writing, the hand feel, consistency of line thickness, etc... to some people that matters. No extra features, it just looks and feels a bit better... But if all you are doing is writing a grocery list, you may not care. And if you don't care, you aren't wrong. This just doesn't apply to you. If you don't have a reason, you don't need to find one. It's just not applicable.

But some people do care. They do have a reason. And they are also not wrong to care. Their reasons just may not apply to you because you have different workloads or priorities (or maybe they do, and you just haven't realized that it's a thing you care about)

I think this just happens to fall under the category of "some people care about milliseconds of rendering time, and some people don't." I don't know if the GPU acceleration has anything to do with it, but this terminal emulator also has really good font rendering.

If you are happy with your current terminal emulator, continue using it. If you heavily use your terminal emulator for a lot of things and in some things you've found that it stutters a bit, and you wished it was a bit smoother, get a GPU accelerated terminal emulator.

And secret bonus option: Even if you are happy with your current terminal emulator, give it a try anyway. Ghostty has a "zero configuration" policy where their goal is for most people to never need to configure anything. Sane defaults. It's a good out of the box experience. Give it a few test drives, and if you're still perplexed about why you should care, then maybe it's just not for you and you can switch back. If you go "that was pretty smooth, i dont have a reason to switch back" then maybe you'll think about it differently.