My view about linux momentum.

Technology-wise things are just getting better and better for linux. Linux is now to the point that even many windows games have better performance on linux than on windows. But about the publishers my observation is that linux gaming had exponential growth in 2013-2015. But exponential gave place to linear since 2015. This is still plenty of games, it is not a matter of quantity and quality is also good (average linux quality is a lot better than average windows quality), the really bad thing is that since 2015 we have zero pleasant surprises. And since 2018 things got worse as we have some bad surprises.

I had run the numbers and they show
exponential growth 2013-2015
linear growth 2015-

growth quantitatively has not slowed down but linear growth means it is slowing down in percentages.

the surprising factor is more important for me, we need good surprises, and hopefully no more bad ones.

There's always going to be ups and downs, we've been through quite a few over the years. I'm currently working on scraping Steam to put together releases by month and yeah to see how big a difference there actually is. Will likely add it as a chart on our dedicated Steam page.

There's a few caveats in doing so though, the main one being accounting for late porting. It's basically impossible unless you know the normal release date and Linux release date for every game on Steam. It's not super common though, so it won't throw the numbers off too far, we're talking likely 1-2 a month (and not every month). I'm also being careful not to include games without a price (not released) but haven't yet account for pre-orders with a price.
Additionally, Valve have measures in place to rate limit so doing it automatically takes a while.

Here's what I have atm. Not verified fully, treat as a test run but I checked over and manually counted multiple dates myself and they seem to correctly match:


Even with the info there's a few things you need to take into account
- September is not included as it's not fair to show as it's only half way through
- Games going exclusive to Epic
- Developers going bust (quite a few!)
- Developers working on short games and then longer games, so releasing less sometimes
- Developers releasing a big game, then treating it as a live service (constantly updated) and not working on others for a long time
- SteamOS was announced in 2013, with SteamOS/Steam Machines releases in 2015 so we're finally really now starting to see the tail end effect of it fizzling out
- Considering Valve's numbers put us below 1%, the amount we get is pretty huge for such a niche even if it's not AAA

The downward trend could be market balancing the number of games to the current level of demand. Paradox said something about that I think, as the reason their sales were going down, while the total number of Linux users actually is growing. I.e. the number of releases was getting too big to be sustainable. Where it will balance out, would be interesting to see.

@Liam Dawe: You can also source data from here:

* https://www.gogdb.org
* https://www.gog.com/forum/general/gog_database_a_website_that_collects_data_on_gog_games
* https://www.gogdb.org/backups/
* https://github.com/Yepoleb/gogdb

Can't use GOG due to their heavy curation and very often long delayed releases. Doesn't really work well.

Still would be interesting to see what their trends are. There is little visibility, without analyzing the data over time.

Quoting: Liam DaweHere's what I have atm. Not verified fully, treat as a test run but I checked over and manually counted multiple dates myself and they seem to correctly match:

External Media: You need to be logged in to view this.

I am giving a like because you collect the data way more systematically than me :P You can also give a nice graph! :)

But polynomial??? This is Heresy! Technical analysis only uses straight lines. You use a bunch of them, you may even get them in random, you can also change the scale (often logarithmic for catching exponential growth) but always straight lines.

:D

Heh sorry, I tried reading up on the best one to use and got tons of conflicting information. No one seems to agree...

Quoting: chancho_zombieare you counting officially valve supported steam play games? are those 2018 spikes steam play games??

No, this is Linux native only in that chart.

I think Linux gaming has run into the same roadblock that Linux in general has always faced: there's no single business entity with deep pockets to market Linux and push it into the public consciousness. I was hoping Valve would take up that challenge, but they really haven't. Most people don't even know there's a free alternative to Windows and OSX, or they still think Linux is some archaic text-based operating system. The other problem is that even people who are aware of and familiar with Linux see no compelling reason to leave Windows for Linux.

I have a bad feeling that Linux will never be more than a niche operating system on the home desktop.

Quoting: GuestMillions are using it, it is in no way a niche. The commercial crap has millions and billions of dollars in advertising and unfair government policies to coax pepople into it. Linux only has the occasional 'weirdo' that, if lucky, shows/endorses it to his/her surroundings.

On the desktop, Windows has around 80% of the market. OSX has about 15%. Linux has around 1.5% which is barely ahead of ChromeOS at 1%. That's niche, like it or not (and I don't like it, but what can we do?).

The simple truth is that making the choice to use free OSes requires one to have a level of knowledge beyond "that's what it came with, so I'll use it". It also requires effort beyond unboxing a shiny new gadget. And if that's the way it is, then it is just nature running it's course.

That's interesting, Liam.
Thanks for doing this!

Quoting: Liam DaweThere's always going to be ups and downs, we've been through quite a few over the years. I'm currently working on scraping Steam to put together releases by month and yeah to see how big a difference there actually is. Will likely add it as a chart on our dedicated Steam page.

There's a few caveats in doing so though, the main one being accounting for late porting. It's basically impossible unless you know the normal release date and Linux release date for every game on Steam. It's not super common though, so it won't throw the numbers off too far, we're talking likely 1-2 a month (and not every month). I'm also being careful not to include games without a price (not released) but haven't yet account for pre-orders with a price.
Additionally, Valve have measures in place to rate limit so doing it automatically takes a while.

Here's what I have atm. Not verified fully, treat as a test run but I checked over and manually counted multiple dates myself and they seem to correctly match:

External Media: You need to be logged in to view this.

Even with the info there's a few things you need to take into account
- September is not included as it's not fair to show as it's only half way through
- Games going exclusive to Epic
- Developers going bust (quite a few!)
- Developers working on short games and then longer games, so releasing less sometimes
- Developers releasing a big game, then treating it as a live service (constantly updated) and not working on others for a long time
- SteamOS was announced in 2013, with SteamOS/Steam Machines releases in 2015 so we're finally really now starting to see the tail end effect of it fizzling out
- Considering Valve's numbers put us below 1%, the amount we get is pretty huge for such a niche even if it's not AAA

Quoting: Liam DaweHeh sorry, I tried reading up on the best one to use and got tons of conflicting information. No one seems to agree...

Actually, it's quite simple. Really. :)

Unless you have a realistic underlying model, don't fit "something".

It's actually not even needed for the given data to grasp.

Thanks again for collecting the data.

Can you somehow automatize that stuff, so that in future you/we can
track that data without much further work?

Edit:

Also, could you please (additionally) do this plot by normalizing to the total number of games released in the particular month?

This would give us the fraction of Linux games released per month.
I guess it is a reasonable assumption that the total number of released
games on Steam is still on a rising trend?

Quoting: Guest'Linux' in general needs no more knowledge than windows, heck, it requires less.

This is true only for IT enthusiast. Many users (even in the gaming world) don't really know what a live image is or why you need to partition your hard drive. With windows, you buy a computer ready to run, if it slows down or don't boot anymore you can pay someone to fix it or you just buy another one.

Quoting: Guest'Linux' in general needs no more knowledge than windows...

Assuming it is already installed and properly configured, this is arguably true. However, the average user does not have the skill set to download, install, and configure Linux whereas using Windows or OSX is as easy as buying a new computer and turning it on.

The Linux experience improved much in the recent decade. Although the user friendliness is not perfect yet, we see efforts from so many participants (Ubuntu, Red Hat, Mint Team, Manjaro Team, Valve, Endless, System76) which is the difference to the years before, where most engineering happened on the subsystems in the kernel.
Even though native gaming declined a bit (basically because of PROTON), Linux has a brighter future than ever.

Quoting: Guest

Quoting: Mountain Man

Quoting: Guest'Linux' in general needs no more knowledge than windows...

Assuming it is already installed and properly configured, this is arguably true. However, the average user does not have the skill set to download, install, and configure Linux whereas using Windows or OSX is as easy as buying a new computer and turning it on.

Assuming it is already installed and properly configured, this is arguably true. If not, you need additional operating systems just to format the drive.

Working in IT, I've done numerous fresh installs of Windows and OSX onto brand new hard drives and have never had a problem. In each case, the installers are smart enough to automatically format the drives. Granted, they don't give you advanced options like partitioning or multiple file systems, but that's not something the average consumer would ever need, or even know what to do with.

Another thing that Windows and OSX do that Linux distros typically don't is walk the user through the first-time setup with lots of helpful "Click here to make this work, you idiot" dialog boxes. It's more common for Linux distos to simply dump you onto the desktop, and if something doesn't work, it's entirely on the user to figure it out. I'm not saying that Linux distros can't be made more user friendly; it's just that most aren't.

Are we seeing the end of the most recent "golden age" of Linux gaming?

No.

Are we seeing the Linux gaming currently enters an new, uncharted territory?
Yes.
Linux gaming right now is complex situation now. Too big for niche, too small for mainstream.

You know, video game industry in it's current will be certainly collapse/crash, just we don't know when.
Many video gaming people (non-shill, not bias, not liar and certainly not ignorant) from fans to journalists to developers etc., etc., predicting video game will crash anytime.
So, this is gonna be a good test for Linux gaming durability to withstand the ups and downs of video game industry.

Linux gaming survivability will be put on test for the first time ever. Hope for the best, prepare for the worst.

Quoting: tonRYou know, video game industry in it's current will be certainly collapse/crash, just we don't know when.

I am waiting for the big publishers/developers to crash and burn for years, but right now there seem enough people content to buy the same dog food in new cans year after year. Have to admit the cans do get prettier each year, though ;-).

I don't think indie games will go anywhere as a whole, either, although I guess on an individual level only few will achieve prolonged success. And it'll be mostly indie games we'll continue to see on Linux, and likely more and more of them in the future.

So to come back to the original question: if by "golden age" the OP is mostly referring to AAA titles making their way to Linux, then I'd consider us currently past the prime. Overall, we're still golden, however, just no longer growing by leaps and bounds. I guess for that to change again it would take a couple of the massively successful multiplayer titles to become available. And I don't see that happening any time soon.