Different branches and different versions. Sometimes there are regressions between new versions. Sometimes the fixes are pushed in certain branches and not others. They are generally provided for stability and compatibilities sake. If you find a version that works for your particular game. You should be able to keep it around and run your game safe and stable unless something on the game gets updated. But outside that you should probably try to run with whatever the latest stable is.
Just to add to this a bit with some visual context, you can visit proton’s github page and look at their releases to see the improvements valve has made to each version.
https://github.com/ValveSoftware/Proton/releases
This can help put what Eldritch said into practice.
If you are having issues playing a game you can visit https://www.protondb.com/ and search for your game. Here you will be able to find what versions of proton others are using. Look for people with similar setups and match the version of proton that works for them.
Thanks for this!
But what about Experimental? Is it just a beta of some other release?
Experimental is its name for a reason. It’s for testing fixes which may or may not fix an issue that they’re investigating. If the fix doesn’t cause any immediate issues they’ll then push it to stable.
So you should really only use Experimental if you have a game or game update that just came out and isn’t running correctly in Stable.
To simplify these are the TLDR ranking:
-
Stable
-
Next (ie: Release Candidate, last bug fix check before pushing to stable)
-
Experimental (ie: Beta, latest fixes that are being tested)
-
Bleeding Edge (ie: Alpha, automated merges for the latest submitted code from devs, things can easily break)
-
Hotfix (For quick bandaid fixes for specific popular games that just released or just updated with some breaking incompatibility.)
Thanks.
If you take a look at ProtonDB, there are many, many examples of people just saying “switch to Experimental”. If Experimental is changing all the time, then this doesn’t make a lot of sense.
If a fix hasn’t made it to Stable yet, then switching to Experimental is the appropriate action to get the game functioning. Just keep in mind that if a ProtonDB review is old but mentions Experimental, then most likely the fix is in Stable by now and switching to Experimental might not be needed anymore. In those cases I’d try the latest Stable first, and then try Experimental if that doesn’t work for some reason.
Keeping note of specific Proton versions is more important if someone says that an older Proton version works better than new ones for reasons. Or if they’re using a forked version of Proton, like GE-Proton, it’s important because that fork explicitly includes things not in normal Proton, like exotic video format support that Valve can’t normally include for legal reasons.
-
It’s basically Valve’s testing center for an eventual stable release. Not everything makes it in to stable, but a lot does. Valve also works very closely with CrossOver (who develops WINE, which Proton is based on).
EDIT: CrossOver is the product that Codeweavers produce. That is essentially the proprietary upstream of WINE. That code trickles down to the open source WINE and Proton projects.
I’m making an assumption here but I think its the new release in production. Might not be as stable as they are probably building it, might be working on fixes and adding or removing things.
Thank you for this! I had JUST been wondering about the distinct differences. Really good timing.
Yeah as someone new to Linux gaming this filled in some key blanks for me.
It’s nice to see an article that’s informative and brief!
Wow, straight the point. Good article!
