My Barriers to Full-Timing Linux

By now, you've probably heard all about Microsoft's AI shenanigans, and saw the posts on social media about how people are sick of how Microsoft has been recently handling Windows 11 and how they're going to move to Linux, for realsies this time.

Now, I feel I have plenty of experience with Linux. I've used it since the early 2010s, witnessing all of its ups and downs along the way. I know my way around the Linux command line, much more so than Windows. I've installed Arch countless times, even got to the point of installing Gentoo. And KDE is probably my favourite desktop environment ever at this point. I can use Linux for a lot of casual scenarios like web browsing or listening to music via Spotify, and even for gaming at times, and if anything needs a more involved process, I don't mind doing it as much as the Linux ecosystem does not push back when I do something off-the-cuff like installing a driver for a specific piece of hardware, setting up WINE prefixes, or even setting the default web browser to the one I actually use. I know at the end of it all, the result will be "set it and forget it" while on Windows, I'm always on edge on whether the set-up will get wiped so Microsoft can shove what it wants into my workflow. It's safe to say, I know Linux pretty well as a user at this point. (fedora btw)

However, I also recognise it's not all there for me. There are some major barriers to me using Linux full time, and these should be exposed and aired out so they can be fixed. Some people in the Linux community don't like it when you complain about it or even tell you "patches welcome 👍" (essentially: do it yourself, idiot), but this is not possible for everyone, as the Linux community found out when Aral Balkan complained on the Fediverse about Fedora's screen reader being broken, essentially shutting blind and vision-impaired people out. Either the people having problems shut up to appease people's feelings, leave the Linux ecosystem with a sour taste in their mouth and allow these problems to keep plaguing the ecosystem, or they can complain and someone more knowledgable than them will notice and help fix these issues and make Linux better for them.

Virtual Reality Gaming (more specifically, the PlayStation VR 2)

I own a PlayStation VR 2, and am not in the market for replacing my headset, and right now, Sony has no plans to support the headset on Linux. The developer of iVRy, who published a beta version of their driver for the PSVR2, also has no plans to support it on Linux either (though there is a Linux category on his discussion board, possibly to check on the interest for it). I've heard plenty of stories about VR support being inconsistent or just okay on Linux, and there are changes to the Linux ecosystem, namely with Wayland, that make running VR applications on Linux much easier, so this will get better. ALVR, an open-source streaming solution for wireless headsets such as the Meta Quest, also supports Linux. But right now, I can't even try those improvements or implementations given that my headset itself isn't supporting Linux right now and isn't supported by ALVR due to its wired connection (not that I expect them to either).

Fortnite

Honestly, if you told me this time last year that I'd be running around Fortnite as a sentient banana, I wouldn't believe you, but yup. I somehow got into Fortnite. I've taken to playing Fortnite Festival (the rhythm game mode) and LEGO Fortnite (essentially a survival crafting game) and I've enjoyed both of them quite a lot.

The game, as far as I know from reading up posts about it and checking Are We Anti-Cheat Yet, employs two anti-cheats: Easy Anti Cheat, which is owned by Fortnite owners Epic Games, and BattlEye. Both of these anti-cheats support Linux, natively and via Proton, and there are major titles that use these two anti-cheats that do not block Linux players. However, Fortnite is a notable exception.

According to Sweeney himself, it's due to the headache of having to deal with custom kernel configurations just to support a small player base such as Steam Deck owners, which is frustrating but understandable. Some people in the Linux community think it's a deliberate attempt to sabotage Valve's long-term plans of making Linux a viable ecosystem for their platform. though I don't know - EAC supporting Linux flies in the face of that.

Well, hopefully Fortnite doesn't give into the kernel-level driver type of anti-cheat that EA and Riot Games have started doing. Otherwise, they're just installing malware onto millions of people's computers.

MusicBee

MusicBee is a music player and library organiser, essentially a power user's equivalent to iTunes. I use it for its customisable playlist views, playlist folders, compressed star ratings (e.g. rating things 3.5 out of 5 and not 7 out of 10, keeping me in sync with sites that do this, like Rate Your Music and MusicBrainz) and tag hierarchy explorer. One project I chip away at every week is a tag hierarchy template for this player that clones the genre tree of Rate Your Music.

I have tried to settle into the Linux native, open-source alternatives, and struggled for one reason or another, crawling right back to MusicBee:

• Quod Libet was nice, and the query-based search system is quicker and more convenient for exploring my music library over MusicBee... until you get to huge groups of tags. I've crashed Quod Libet by requesting every genre under RYM's Electronic Dance Music umbrella. Imagine if you wanted it to give you all of the rock or metal genres in RYM's tree.
• Strawberry, which was forked from Clementine, was a no-go once I saw that split tags aren't supported.I use these a lot when a song has multiple genres, and my tag editor of choice (MusicBrainz Picard) uses these extensively (e.g. handling the MusicBrainz IDs of multiple artists).
• gmusicbrowser looked very promising and was very customisable, but it's showing its age, development is slow and a lot of distributions have dropped it as a result.
• Tauon Music Box looks beautiful, but I've had bad luck with the generators in comparison to MusicBee's tag hierarchy explorer and Quod Libet's queries, trying to make auto-playlists for groups of genres. (That being said, the developer has expressed interest in supporting this, with the main barrier being the lack of a central template to use.)

And this is without mentioning that I lose some of the features that keep me onto MusicBee, which actually runs well in WINE... well, sort of:

• You don't get any link to MPRIS - the standard used by desktop environments like KDE - for media controls at all, which means media buttons and metadata are unavailable on WINE.
  • To be fair, MusicBee doesn't link to Windows 10/11's media handler at all, but a plugin is available for this, and the issue is that WINE can't translate those calls to MPRIS.
• The functions handling the dragging of tabs are not implemented in WINE yet, which means doing that results in MusicBee throwing errors at you.
• Latin font redirection is broken in MusicBee, meaning that if you use a font that doesn't have the scripts for other languages (e.g. Chinese, Japanese, Korean, Thai), you have to contend with boxes in place of the letters you want to see.

The tab dragging and lack of support for media control I could deal with, but the other issues could kill the Linux experience for those who want to hold onto MusicBee and almost did kill my experience. That's not to say WINE isn't making progress. The fact I can run MusicBee in the state it's in at all is great, but I do feel that MusicBee has being put in the sidelines, especially as gaming has become a large focus for WINE.

Can this change?

Well, I'm just going to say "not likely" for Fortnite, right off the bat. Epic does not feel it to be economically viable to target what is basically 2% of Steam's userbase (as of May 2024) and they can look to Roblox who tried to support its Linux userbase, then recently ceased supporting WINE due to the very problem they were worried about. The only way that can change is if Linux can make a notable dent into Microsoft's domination of the PC gaming market share.

VR and MusicBee, however, I believe will change for the better in the future. There are determined hackers and tinkerers in both the VR and WINE communities who would be interested in solving these problems. Someone had even tried to push changes to fix the drag-and-drop problems that plagued MusicBee on WINE (though progress seems to have stalled?). I've even considered making a Linux native or cross-platform application with Qt for my own needs, since that is easier for me to have a crack at than reverse engineering Windows.

And I have no reason to believe that the PSVR2 will be locked to Windows forever. Hell, I'll even be really optimistic and say that if Valve pushed harder for supporting VR with Linux the way they've done non-VR games, Sony may have even considered supporting it themselves, given their past history with supporting PlayStation devices on Linux. And if not, well, someone will certainly have a go at making it work. The first PSVR is already supported in Monado, an implementation of an OpenXR runtime for Linux.

Until then, though, Windows may have to be a mainstay on my computer alongside my Fedora set-up. Maybe forever. Let's hope not.