Why you dont like epic game store?

[Graf Zahl]

The download counters count downloads, they are independent of running the game.
Currently they are in a private section of the forum.

[Rachael]

The numbers from last release usually are posted after a new release comes out, though. So yes, there is public access to this data. Just check the old release topics, or search for my posts in the news section.

[Chris]

In short: The amount of technical issues in proportion to the user base is significantly higher on Linux than on Windows.

Because the amount of effort people tend to put into Linux is significantly lower than on Windows. If you don't have as much QA for a particular platform, it will of course have a higher proportion of issues compared to the size of the user base. But people aren't going to suddenly start using something on a platform if it's continually sub-par on said platform*. You can't expect a platform to grow when it's not properly nourished. See EA's treatment of the Switch (make cross-platform games, put less effort into the Switch version which causes it to be worse than it could be, causing fewer Switch users to buy it because other games/versions are better, then use that low interest to justify the low effort). But even in cases where Linux itself has higher than normal issues, this all leads back to the chicken and egg problem.

* Not saying GZDoom specifically does this. But in general, if you over-compensate for low expected returns for a platform with reduced effort to support that platform, the result of the reduced effort can be even less return proportionately.

It's probably not economical to support a platform like this (and for a company like Epic that looks to be the only relevant measurement.)

I dunno. Unlike John Carmack (who would talk from a hard economic viewpoint for official Linux support), Tim Sweeney has had some very... odd things to say about Linux, very much not economic in nature. He likens switching to Linux from Windows to be like moving to Canada from the US (implying if you have problems with Windows and think Linux is better, you should stay on Windows to help it improve). He thinks any problems Windows may have or threats it may pose can be dealt with by asking Microsoft nicely to do what you want (certainly with the kind of money Epic has, they would have some sway in what Microsoft does with Windows, but that's still limited, and no where near the kind of influence the average user has). At least from Sweeney's stand-point, there's more than purely economical factors.

It's true that Sweeney doesn't directly control what Epic does, but I imagine he influences the focus of the company. Make things take longer than they would, or require more apparent benefits before doing something that would still be beneficial. Or it could just be him being bad at PR, making the situation look worse for future Linux support than it actually is. In any case, considering both large companies (Steam, GoG) and small companies (Humble*, itch.io) support selling Linux games, I'm not sure economy is really a question for a storefront. Just don't do things to cause that platform's users to avoid buying from you.

* Don't think Humble's technically "small" anymore since they were bought out, but when starting out as a place for indies by indies, it required games to have Linux and Mac support for its initial bundles. They eventually relieved that requirement, but continued to support Linux and Mac games in the store.

Agreed, but some of the Linux people seem to operate under the impression that they can do to their system what they want and then still expect that software runs without problems.

I've seen plenty of Windows users think the same too. Mucking about with under-the-hood stuff on Windows can just as well cause things to start working improperly if you don't know what you're doing. People do this, then they go online, complain that everything borked, and be told to fdisk, format, reinstall, doo-da, doo-da (the procedure's a bit different these days, but still the same effect; get the system back to a state from before your changes). Unless you lock down the system, and even then, people will always do this.

it's its community that doesn't seem to understand what needs to be done to really displace Windows.

Because 10 different people argue 20 different things. It's hard to get a clear consensus when various people are saying different and contradictory things about what's "obviously" wrong and how to do it right; even among commercial/indie developers, one will recount their adventures in Linux support saying "Linux was hard, because XYZ support was poor", and another will say "XYZ support on Linux is good, just don't treat Linux support as an after-thought and its easy". Or one saying "Linux was hard because ABC isn't nice" and another saying "ABC is good, use it everywhere and makes support easy". How is anyone supposed to understand what needs to be done?

I've also seen plenty of arguments that Linux needs to reach some level that not even Windows can reach. Even here in this thread, like the above about people doing what they want with the system, or when you say "Visiting typical Linux user hangouts on the web is often a strange experience". You can simplify that to "I primarily use ${OS}, visiting typical ${OTHER_OS} hangouts is often a strange experience". Windows users to Mac or Linux hangouts, Mac users to Windows or Linux hangouts, Linux users to Windows or Mac hangouts. They'll almost always come away saying it's an odd experience to be in that strange alien land. How can you expect Linux to be better when it's a universal trait?

Most Linux users acknowledge that things can be improved. They'll be among the first to say "software is never finished, it can always be improved". Maybe what you say is the One True Answer. But in the end, nobody knows what it will really take to displace Windows as the main desktop OS, because nobody has done it. Certainly not in recent times, anyway. We can make (educated) guesses, and we can try various things, but given limited time and energy we can't do everything that people on the internet say. And that's assuming anything can displace Windows, as long as Microsoft has the money and desire to hold on to the desktop.

[Rachael]

Despite all the problems, I do find Linux to be a very rock-solid OS for my uses. It is lacking in some features, but it compensates with many others. And for the features it lacks, there's the option to use virtual machines or Wine, or worst case scenario, dual boot.

[Graf Zahl]

In short: The amount of technical issues in proportion to the user base is significantly higher on Linux than on Windows.

Because the amount of effort people tend to put into Linux is significantly lower than on Windows. If you don't have as much QA for a particular platform, it will of course have a higher proportion of issues compared to the size of the user base. But people aren't going to suddenly start using something on a platform if it's continually sub-par on said platform*. You can't expect a platform to grow when it's not properly nourished. See EA's treatment of the Switch (make cross-platform games, put less effort into the Switch version which causes it to be worse than it could be, causing fewer Switch users to buy it because other games/versions are better, then use that low interest to justify the low effort). But even in cases where Linux itself has higher than normal issues, this all leads back to the chicken and egg problem.

* Not saying GZDoom specifically does this. But in general, if you over-compensate for low expected returns for a platform with reduced effort to support that platform, the result of the reduced effort can be even less return proportionately.

The sad truth is that you cannot invest the same time and effort in the 1% market share compared to the remaining 99%. And this gets amplified by the vast amount of different configurations which may or may not be compatible with the chosen setup.

It's true that Sweeney doesn't directly control what Epic does, but I imagine he influences the focus of the company. Make things take longer than they would, or require more apparent benefits before doing something that would still be beneficial. Or it could just be him being bad at PR, making the situation look worse for future Linux support than it actually is. In any case, considering both large companies (Steam, GoG) and small companies (Humble*, itch.io) support selling Linux games, I'm not sure economy is really a question for a storefront. Just don't do things to cause that platform's users to avoid buying from you.

* Don't think Humble's technically "small" anymore since they were bought out, but when starting out as a place for indies by indies, it required games to have Linux and Mac support for its initial bundles. They eventually relieved that requirement, but continued to support Linux and Mac games in the store.

Well, imagine that they allow selling Linux versions but due to company policy don't produce them for their own game. Do you think that'd work better for them?

I've seen plenty of Windows users think the same too. Mucking about with under-the-hood stuff on Windows can just as well cause things to start working improperly if you don't know what you're doing. People do this, then they go online, complain that everything borked, and be told to fdisk, format, reinstall, doo-da, doo-da (the procedure's a bit different these days, but still the same effect; get the system back to a state from before your changes). Unless you lock down the system, and even then, people will always do this.

Those people always exist, but in the end Windows is a lot harder to break because the dependency of third party libraries being properly set up is considerably less.

As it just so happened, I was experimenting with FreeType the last few days to see if it can be of use for GZDoom.
And that library is a first-grade headache inducer, because too many options are not run-time but compile-time options, so with the tinkerers in the game it is virtually impossible to have a guarantee of stability, unless you statically link to it - which is frequently frowned upon by the Linux crowd.

it's its community that doesn't seem to understand what needs to be done to really displace Windows.

Because 10 different people argue 20 different things. It's hard to get a clear consensus when various people are saying different and contradictory things about what's "obviously" wrong and how to do it right; even among commercial/indie developers, one will recount their adventures in Linux support saying "Linux was hard, because XYZ support was poor", and another will say "XYZ support on Linux is good, just don't treat Linux support as an after-thought and its easy". Or one saying "Linux was hard because ABC isn't nice" and another saying "ABC is good, use it everywhere and makes support easy". How is anyone supposed to understand what needs to be done?[/quote]

The real problem is that both ABC and XYZ are probably third party dependencies and depending on the state of documentation, some people find it easier to become acquainted with ABC and others with XYZ. In the end it's the lack of consistency, both in API and documentation that causes the problems.

Let's again use font support as an example: This has been the most aggravating and annoying issue for me recently. How the fuck am I supposed to render localized text with Linux? I have no clue. Most software I look at uses FreeType directly and forfeits support for any more complex language because FreeType is too low level. The system itself doesn't help me one bit, so the end result will probably be that I'll code proper TrueType support for Windows and Mac where I have well-defined APIs at my disposal to render some text into a graphic and leave Linux to the ugly bitmap font.

I've also seen plenty of arguments that Linux needs to reach some level that not even Windows can reach. Even here in this thread, like the above about people doing what they want with the system, or when you say "Visiting typical Linux user hangouts on the web is often a strange experience". You can simplify that to "I primarily use ${OS}, visiting typical ${OTHER_OS} hangouts is often a strange experience". Windows users to Mac or Linux hangouts, Mac users to Windows or Linux hangouts, Linux users to Windows or Mac hangouts. They'll almost always come away saying it's an odd experience to be in that strange alien land. How can you expect Linux to be better when it's a universal trait?

The important thing here is that neither for Windows nor for Mac these people are relevant. They have their weird attitudes but since they have no influence over the system they cannot do any damage.
But on Linux these attitudes very often filter down to how these people set up their system (e.g. "Never in my life will I install anything requiring systemd!") so what they think and do directly affects the target surface.

Most Linux users acknowledge that things can be improved. They'll be among the first to say "software is never finished, it can always be improved". Maybe what you say is the One True Answer. But in the end, nobody knows what it will really take to displace Windows as the main desktop OS, because nobody has done it. Certainly not in recent times, anyway. We can make (educated) guesses, and we can try various things, but given limited time and energy we can't do everything that people on the internet say. And that's assuming anything can displace Windows, as long as Microsoft has the money and desire to hold on to the desktop.

What Linux really needs - as I have said before - is a solid presentation of GUI APIs and not this horrible mishmash of different tools from different people with different coding styles and different philosophies and different quality of documentation spread out over lots of different places. Linux is a great system as long as you do not need to run a desktop on it and/or code for Linux desktops. Once you go down that rabbit hole, it becomes a nightmare.

Solve that problem and therefore making it easier to develop GUI software on it and I strongly believe that things will improve. But then we again get to the same state of "My desktop is better than yours!" and all progress will be lost because people wouldn't install that unifying component.

I don't think it's a coincidence that Android is the only Linux based system having any success in consumer space. I also don't think it's a coincidence that macOS is the only Unix-based desktop OS having any success in consumer space (hampered, of course, by Apple being Apple and not offering an adequate selection of machines to select from.) In both cases, what was done is to hide the gory details to both users and developers and offer a developer/user surface that is consistently designed and documented.

In short: With the current "spirit" of the Linux community, don't expect to see some change here. As long as all these differences are celebrated to be a good thing we won't get what we really need: A solid desktop OS free of big corporation shenanigans.

[LOZ_98]

I prefer to have my games in one place, where all my friends, achievements, screenshots...etc are, So I go with Steam. Origin is not too bad either, I have a couple of games there, some of which aren't available on Steam. my biggest issue with the Epic game store is by far the launcher, It's just so incomplete compared to Steam and Origin. There's no options to migrate game installations to different drives, or any option to detect existing game installations. 2nd biggest issue is the lack of game hubs and communities, There's no user reviews so you can't see what others think of the games in there making it even harder to decide whether you want to get said games or not. This makes it look like a perfect way for publishers of these games to hide negative reception of their games (although to be fair Origin doesn't have that either) but thankfully you can just use google and see what people think of the games.

So to sum it up, Epic game store just feels so underdeveloped in it's current state, and I would rather have the option to pick-up my favorite games in a different digital store in the meantime than be forced to deal with their launcher.

[Dancso]

Here's a fresh can of worms to add to the linux discussion...
https://rust.facepunch.com/blog/linux-plans/
-
Funnily enough on the small indie side of things I'm tempted to ignore Macs because it'd cost me more to buy a device to test with than I'd make on sales. I'd probably give linux a pessimistic try though, but it may boil down to "does it work in a virtual machine" because I can't be assed to set up dual boot.
Maybe I'll set up my current pc if I do a full upgrade several years down the line. yeah

[Talon1024]

I think the in-depth Linux discussion is better off in another thread, and the Linux discussion in this thread should be about Linux support as it relates to the Epic Games Store and its respective client.

My thoughts on the subject are as follows:
- I don't think it will be much work to add Linux support to the Epic Games Store, since it's just a matter of allowing developers/publishers to upload builds for the Linux version of a game, and showing that the game has Linux support.
- I don't think it will be much work to add Linux support to the Epic Games Store client either, since it's an Electron app, and Electron supports Linux, unless they are using some native library or platform-specific code which doesn't support Linux, and doesn't have an alternative on Linux.

[Chris]

The sad truth is that you cannot invest the same time and effort in the 1% market share compared to the remaining 99%. And this gets amplified by the vast amount of different configurations which may or may not be compatible with the chosen setup.

I get that you can't invest the same time and effort, but it still requires some effort to see a return on investment. In some cases, it's less about the amount of cost and effort and instead about having proper forethought... investing in using OpenGL or Vulkan from the start will greatly reduce the amount of effort needed to port, for example, and have fewer overall issues when you do since most of the issues will have been handled much earlier when they were easier to deal with. If you start with Direct3D and see OpenGL/Vulkan as a separate thing you have to do after everything else for non-Windows support, of course the amount of effort to do that will be disproportionately high. You're doing more than twice the work for rendering support, adding new code after the fact that the engine wasn't originally designed to expect. A little bit extra work at the start, or more than double the workload later on... my mother always said don't procrastinate on your tasks, it's just more work in the end.

Just ask id Software. When they started focusing on having cross-platform code instead of leaving it as an after-thought (starting with cross-platform code and making platform-specific code only as needed, instead of the other way around), it helped everywhere. Linux support for them was never a question of development cost, but purely post-release support costs... and that was mostly a result of poor OpenGL drivers if you weren't using nVidia back then, as they were constantly pushing capabilities that few others touched (or was just created for them).

Well, imagine that they allow selling Linux versions but due to company policy don't produce them for their own game. Do you think that'd work better for them?

Yes. Absolutely no question, yes. Valve used to do that when Steam first came to Linux. CD Projekt still does that with GoG; the only game they've made that supported Linux is Witcher 2, which was not a particularly well done port. Witcher 3 isn't on Linux (despite initially being advertised for Linux/SteamOS, it never materialized), and there's no news of Cyberpunk 2077 coming to Linux. But GoG continues supporting other Linux games and Linux users happily use it for that.

As it is, though, Epic hasn't had too big of a problem with Linux in the past. Their Unreal Engine supports it, as does their older Unreal games, and the new Unreal Tournament they were developing supported it. It's mostly just Fortnite that's conspicuously lacking a Linux version.

And that library is a first-grade headache inducer, because too many options are not run-time but compile-time options, so with the tinkerers in the game it is virtually impossible to have a guarantee of stability, unless you statically link to it - which is frequently frowned upon by the Linux crowd.

As I've said before elsewhere, developers should first and foremost focus on what they need to do to get their game working on Linux. If that means doing something that Linux users would prefer not, so be it. For example, Feral Interactive makes Linux ports of other peoples' games by rebuilding them with winelib (not unlike how Proton works but done at compile time). That's very abnormal and not what most Linux users would prefer, but it's less overall work for Feral who don't have to completely rewrite code they didn't make, has good results, and people buy and praise them for their work. Similarly, most Linux users strongly frown upon DRM, but they'll still buy games that have it as long as it actually works well.

I don't think it's a coincidence that Android is the only Linux based system having any success in consumer space. I also don't think it's a coincidence that macOS is the only Unix-based desktop OS having any success in consumer space (hampered, of course, by Apple being Apple and not offering an adequate selection of machines to select from.)

Android can't be directly compared to desktop Linux because those are vastly different user bases with different needs and expectations. If you tried to put something like Android on a desktop system, it would fail just like Windows did trying to go the other way. Further, macOS's success could be attributed to Microsoft's bail-out (which they did as part of their anti-trust settlement with the government). Apple wouldn't be where they are now without Microsoft's financial help. And similarly, macOS targets a different audience than Windows, and has users that expect different things than Windows. You'll find very few people that freely waffle between macOS and Windows systems for regular use and not care about which they're using.

If I'm wrong though, feel free to point out where and how.

[Graf Zahl]

I don't think it's a coincidence that Android is the only Linux based system having any success in consumer space. I also don't think it's a coincidence that macOS is the only Unix-based desktop OS having any success in consumer space (hampered, of course, by Apple being Apple and not offering an adequate selection of machines to select from.)

Android can't be directly compared to desktop Linux because those are vastly different user bases with different needs and expectations. If you tried to put something like Android on a desktop system, it would fail just like Windows did trying to go the other way. Further, macOS's success could be attributed to Microsoft's bail-out (which they did as part of their anti-trust settlement with the government). Apple wouldn't be where they are now without Microsoft's financial help. And similarly, macOS targets a different audience than Windows, and has users that expect different things than Windows. You'll find very few people that freely waffle between macOS and Windows systems for regular use and not care about which they're using.

If I'm wrong though, feel free to point out where and how.

I'm not that convinced about your macOS assessment. The target audience doesn't really expect different things it's just that many Apple users have that weird Apple-or-nothing attitude and rather waste insane amounts of money than going to the competition. Since I work in a very Apple-centric environment I see both sides of the coin and the one thing that's very obvious is that many Apple-only users often have zero knowledge about what Windows really can do, just that typical misinformation (often based on ancient Windows versions) that gets spread around in closed circles.

I work with both Windows and macOS and the constant irritation I have with Macs is not the operating system - it's that when on Windows I need some tool it's normally easy to find something good on the web for free, while on macOS nearly everything costs money - provided that it even exists -, so in addition to an overpriced computer I'd often be forced to spend a lot more money on software as well. In other words: Apple is simply way too expensive for my taste. I also frequently have to start a Windows VM when having to use some software that doesn't exist on macOS.

I think the bigger incompatibility exists between Linux users and the rest. Our server developer is one of those really hardcore Linux types - he spends most of his time on the terminal window, uses vim as his editor of choice and generally avoids the GUI as much as possible. And to be frank, looking at the tools that come out of the Linux community, this seems to be the norm rather than the exception - or at the very least these types make up a significant chunk of the 1% that use Linux.

[Chris]

I'm not that convinced about your macOS assessment. The target audience doesn't really expect different things it's just that many Apple users have that weird Apple-or-nothing attitude

Isn't that exactly the difference, though? Apple users tend to want and expect walled-garden style computing. Anything they change or install needs to either come from or be approved by Apple. In comparison, Windows and Linux users are more likely to go out and find whatever they need that does the job and use that, without requiring approval from some controlling entity. These are different audiences: people who want someone else to manage their system and everything that goes on with it, and users who want control over what they can put on and do with their system.

Our server developer is one of those really hardcore Linux types - he spends most of his time on the terminal window, uses vim as his editor of choice and generally avoids the GUI as much as possible. And to be frank, looking at the tools that come out of the Linux community, this seems to be the norm rather than the exception - or at the very least these types make up a significant chunk of the 1% that use Linux.

I would disagree with that. I don't contest that there are some people like that, but the image of a typical Linux user being a GUI-avoiding hacker nerd is so outmoded it's not even funny. Not that it was strictly true to begin with (avoiding the GUI is different from preferring functional terminal apps you're already familiar with), but these days there's a good number of popular GUI apps/editors/IDEs on Linux, from simple to complex. Some became so good that they're often or sometimes found on Windows machines too (Firefox, Thunderbird, VLC, LibreOffice/OpenOffice; Pidgin used to be, but few people really use IMs or IRC anymore). Torvalds himself said that Linux began to gain importance in 1992 after the X Window System was ported (the first beta release of Linux was late '91, and Linux 1.0 wouldn't hit until '94). Simple GUIs were there almost from the start, inherited from earlier Unix-based systems, slowly but consistently improving over time to be more user-friendly. Linux GUIs actually even had certain features before Windows did (a proper client-server model, remote desktops, hardware accelerated compositing). Not what you'd expect from users who avoid the GUI as much as possible.

[Graf Zahl]

I'm not that convinced about your macOS assessment. The target audience doesn't really expect different things it's just that many Apple users have that weird Apple-or-nothing attitude

Isn't that exactly the difference, though? Apple users tend to want and expect walled-garden style computing. Anything they change or install needs to either come from or be approved by Apple. In comparison, Windows and Linux users are more likely to go out and find whatever they need that does the job and use that, without requiring approval from some controlling entity. These are different audiences: people who want someone else to manage their system and everything that goes on with it, and users who want control over what they can put on and do with their system.

It's a bit more complicated. The main issue here is that there's this basic assumption that someone willing to pay double price for half performance is also willing to pay for the software they need. The ones who like walled gardens are a minority. I don't know any of this type.

Our server developer is one of those really hardcore Linux types - he spends most of his time on the terminal window, uses vim as his editor of choice and generally avoids the GUI as much as possible. And to be frank, looking at the tools that come out of the Linux community, this seems to be the norm rather than the exception - or at the very least these types make up a significant chunk of the 1% that use Linux.

I would disagree with that. I don't contest that there are some people like that, but the image of a typical Linux user being a GUI-avoiding hacker nerd is so outmoded it's not even funny.

The problem is, all the Linux users I know are this type - the one I was talking about a perfect textbook example. Maybe that's because they are all professional Linux users - I do not know any person who uses Linux and is not a computer expert in some form.

[dpJudas]

I'm not sure you can call Firefox and Thunderbird Linux apps ported to Windows. I'm sure you will now say that the first build of Firefox was for Linux, but that'd be ignoring the render engine it was built on was first and foremost always developed for Windows and Mac. How do I know that? Because early versions had so horrible Linux DNS lookup code (it would hang the main thread for over 30 seconds) it could only really be explained by the main devs not using it as their primary platform.

OpenOffice is even more silly to mention as the commercial version it originates from was clearly developed for Windows first (it is quite evident if you look at its codebase and how its drowned in silly Win32 junk from the 90's).

As for good IDEs on Linux, well, I can only say I disagree on that as its the main blocker for me to use it as a platform.

In general, I can only echo what Graf said about Linux users. Almost all I've met are developer types that have a preference to a console life. Yes, the console isn't ugly DOS based 80x25 text mode, but they still clearly see the terminal as an acceptable solution to a lot of problems. That doesn't mean its wrong to use your computer that way (hey, if it works for them). It does however mean that they don't spend time coding for people that wants a different way.

[Chris]

I'm not sure you can call Firefox and Thunderbird Linux apps ported to Windows. I'm sure you will now say that the first build of Firefox was for Linux, but that'd be ignoring the render engine it was built on was first and foremost always developed for Windows and Mac. How do I know that? Because early versions had so horrible Linux DNS lookup code (it would hang the main thread for over 30 seconds) it could only really be explained by the main devs not using it as their primary platform.

Not Linux-apps-ported-to-Windows per-se. I mean, the way most people develop software on Linux generally makes them inherently cross-platform with no real "porting" going on. But as far as I can tell, those things I listed all found a place among Linux users before becoming a common find on Windows.

OpenOffice is even more silly to mention as the commercial version it originates from was clearly developed for Windows first (it is quite evident if you look at its codebase and how its drowned in silly Win32 junk from the 90's).

According to the information I can find, StarOffice didn't get much market share. After Sun acquired it in 1999, they open-sourced it in 2000 as OpenOffice to compete with Microsoft Office, and was primarily developed for Linux, Windows, and Solaris. It quickly became the standard office suite for Linux, and followed to gain ground on Windows.

As for good IDEs on Linux, well, I can only say I disagree on that as its the main blocker for me to use it as a platform.

Even most Linux users will agree that Visual Studio is probably the best IDE out there. But to be fair, there are decent IDEs on Linux. I regularly use KDevelop. It has its issues (I wouldn't doubt some of which is just a result of my ancient hardware), but it's more than usable for me. I would say it's good, but it's a subjective call.

In general, I can only echo what Graf said about Linux users. Almost all I've met are developer types that have a preference to a console life.

I can say I don't have a preference to console life. I couldn't tell you the first thing about vim or emacs other than there used to be some religious wars over it. I will use the console where it makes sense for me to, but I'm using GUIs more often than not. The number of desktop managers and GUI tools constantly being made and worked on for Linux would suggest I'm not an outlier.

[NeuralStunner]

Apparently Epic is telling indie devs that they can't publish on EGS at all if they're not exclusive. (And check out Sweeny's evasive response near the end of the article.)