Steam goes Linux

I'm gonna be blunt: Virtualization just isn't gonna cut it. I want my games to run natively stable, and I want them to run natively fast. I've tried WINE several times, and have never had a positive experience with it. Always seems to be crashing or not working or can't find the right files or something. If improving WINE is Linux's best hope, they have no hope.

In addition, I think the "free" points are pretty valid. Most Linux users are in a world of their own, and even though they will point out to no end that they really don't mean "free" as in beer, they mean "free as in freedom," truth be known they'll turn away most stuff that isn't $0, whether the source code is available or not.

But seriously support is one of the strongest points of GNU/Linux.

No, it's the weakest. Forums and IRC channels don't count. I'd hate to teach my father how to use IRC. He doesn't want to deal with server names, port numbers, and nicknames. Frankly, even as simple as it is, it's still too complex. And truth be known, most people still want to get support via a telephone.

In any case, back to the subject: This is an interesting move by Valve. I wonder why they're going with Linux first, and not Mac OS X? Granted, considering how much they have in common, porting to Mac OS X will be easier after a Linux port - but still, Mac OS has always been stronger than Linux in gaming, even if they're not as strong as Windows.

I suppose they feel that porting to Linux opens a lot of possibilities, opening up the potential for moving to Mac OS X and other *nix based platforms in the future.

If this is indeed true, this would likely be the biggest change to the Source engine since the introduction of HDR. Moving to HDR required touching nearly every component, and I'm sure this will as well.

Thankfully, the Source engine was designed from the ground up to be modular, and proved that even large scale changes are possible with it. Changes of this scale are not unfamiliar to Valve, so if any company is capable of pulling off such a change, it would be them.

Still - I wonder if this is the right move for them? This is going to require a lot of resources, for sure - but will there be enough benefits? Is the Linux community going to be willing to lay down some dollars and start paying for games? Will it ultimately pay for itself?

Guys, find someone with a GP2X, take a look at it and think again wether Linux is unsuitable for games.

Well, that's the problem. Nobody has a GP2X. Linux is a perfectly capable OS, and it does have a decent number of people running it - but it's still nowhere near the behemoths that are Microsoft and Apple. Half the problem isn't getting it to work on Linux. Half the problem is justifying porting it to a platform that, despite all it has to offer, never seems to gain any sort of market share.

. . . and to be totally honest - Linux still isn't really not that user friendly. Not even Ubuntu. Defaulted to Gnome, which is confusing to Windows users, and still not intuitive on its drag and drop. I'm in a virtual machine right now, trying to drag and drop the trash onto the desktop, trying to create an icon. No can do - what gives?

. . . and don't get me started on trying to run anything as root in Ubuntu. You practically can't. At least you can run stuff as admin in Vista, even if it's super annoying. In Ubuntu, the root is pretty much sealed off.

. . . and oh, yeah, instead of shortcuts, we have "launchers" which default to a springboard icon instead of the application's icon. For crying out loud, why??

. . . and of course Gnome has two bars, one on top, and one on bottom - hello? Heard of the KISS principle?

The default theme is - well, it's decent. I dunno if it would look any better running natively rather than in a VM, but it looks a bit utilitarian. Not really anything to get excited about. Not as exciting as OS X or Windows 7. At least it's not as bad as XP's Luna, which I always turn off.

If this is the future of Linux, it's looking pretty dim. I know Linux users swear up and down it's fantastic, but I'm just not seeing it. From what I can tell, they are still as clueless as ever when it comes to user interface design.

. . . and look here, the most recent updates of Linux on my VM require a restart!! . I thought you guys were soooo proud you never had to do that??

Like, whatever. I'm just not seeing how fantastic this OS is supposed to be.

Cobra1, you basically display that you don't know GNU/Linux and you're trying to do things as if it was windows. It's not. Most of your rant is simply that, an opionated rant because you don't get it. You're hardly pointing out any objective flaw other than saying "It won't work" and "It will never happen" and "Really, it sucks, only geeks use Linux".

Half the problem is justifying porting it to a platform that, despite all it has to offer, never seems to gain any sort of market share.

In case you missed it, GNU/Linux market share is constantly increasing. Who even knew about it 10 years ago? Yet now even Dell is offering it and you can buy it on Walmart.

Quoting CobraA1,

reply 1

1) In addition, I think the "free" points are pretty valid. Most Linux users are in a world of their own, and even though they will point out to no end that they really don't mean "free" as in beer, they mean "free as in freedom," truth be known they'll turn away most stuff that isn't $0, whether the source code is available or not.

2)No, it's the weakest. Forums and IRC channels don't count. I'd hate to teach my father how to use IRC. He doesn't want to deal with server names, port numbers, and nicknames. Frankly, even as simple as it is, it's still too complex. And truth be known, most people still want to get support via a telephone.

Let begin with the 2) ... if you wish support via a telephone, simply buy a boxed version... sure that you have a free version who make you look like the guys from the 1)... Linux Server edition is around the 3000$... but for these version, you have unlimited support... a boxed desktop version is round 70$ and your have 10 hours phone support...

About market part for game... Direct is only for windows/Xbox ... opengl is for all the rest AND Windows/xbox... For now, stardock don't really need to use opengl but if one day, they move to the game console sector, they will need to make the move... using opengl will allow them to make game for the playstation 3 and the Xbox in place of only the Xbox...About PC with windows, OpenGL and Direct3D are both implemented in the display driver.

For info, from 1992 until 2003, Microsoft was part of the opengl project, they are one of the founding members... they have quit for strategic reason... now that directx is mature, they don't need of opengl anymore... the DX pixel shader 1.1 was nothing more than a pseudo-assembly language version of the NVIDIA-specific OpenGL extensions... the DX pixel shader 1.4 was nothing more than a pseudo-assembly language version of the ATI-specific OpenGL extensions... With the Pixel Shader 2.0 who allowed a unified code path under Direct3D, Microsoft was not more needed the OpenGl code and have quit the project...

Speed are similar in opengl and directx... last version of both have similar function... but the main difference is the API implementation... for open-gl, it is called GLX on Linux, WGL on windows, CGL on Mac OS X... these 3 have enough difference to be not fully compatible... the last DX10 implement "marshalling"... last function that DX was missing... Ultrashadow don't count since it is only available with the nvidia driver and they own gl extension...

A quote :

In the earliest days of 3D accelerated gaming, performance and reliability were key benchmarks and several 3D accelerator cards competed against each other for dominance. Software was written specifically for a particular brand of graphics card. However, over the years, OpenGL and Direct3D emerged as software layers above the hardware, mainly because of industry support for a cross-hardware graphics library. Competition between the two rose as each game developer would choose either one or the other. In the 1990s, OpenGL dominated due to a better design of the interface. In 2000s game developers began to favor Direct3D also only because of the dominance of the Windows operating system in the gaming market.

Today, the situation begin slowly to reverse... piracy is killing the PC game market... game creator turn to game console where only the Xbox is directX...the majority of game console is opengl... a lot of game are created for PC first and ported to console after... but slowly, the trend reverse... first console and if enough money is earned, maybe make a port to PC...

Congrats Steam, on getting another 1000 people of the overall gaming market... Let's face it, linux isn't a gaming platform, and has a long way to go before becoming one.

I commend them for pioneering though, it definetely needs to be done to help linux become a viable alternative to windows.

and you're trying to do things as if it was windows. It's not.

. . . which is why adoption, even though it's happening, is slow. People who try to use Linux for the first time aren't going to change their habits on the spot. Like it or not, they'll judge the platform on how well it reflects their previous experiences. If Linux seems to be stubborn about drag and drop, they'll judge it on that.

You're hardly pointing out any objective flaw other than saying "It won't work" and "It will never happen" and "Really, it sucks, only geeks use Linux".

Let's start with drag and drop:

In Windows, I can drag and drop icons anywhere, including from the quick launch toolbar to the desktop. It's considered a standard feature, and it's an expected part of the UI.
In Linux, that doesn't seem to be the case. Drag and drop works in some places, but it doesn't seem to work from the panels to the desktop, even when the icon is unlocked.
You have yet to explain why you think this behavior is desireable or at least give some sort of excuse for this behavior.

On to launchers:

In Windows, they are called "Shortcuts" and when you try to create a shortcut, Windows will, by default, use the icon associated with the application. This is an expected behavior, and considered useful because it makes identification of the program as easy as looking at the icon.
In Linux, when I create the equivalent of a shortcut - a "launcher" - it creates a springboard icon and I have to change it manually every time I create it.
You have yet to explain why you think this behavior is desireable or at least give some sort of excuse for this behavior.

You're right - I don't "get it." But as a quote from my sowtware engineering book says: "It is a far better to adapt the technology to the user than to force the user to adapt to the technology" --Larry Marine

I don't really care how flexible Linux is - if I've got to dig around through a gazillion archaic settings - or worse, fiddle with source code - just to make basic drag and drop work, then yes, I consider it broken.

In this case, my objective is simple, really: I want to, in the most painless way possible, place a functional Trash icon on my desktop. How do I do that?

It's a basic scenario. It's not like I'm trying to do anything complex. It should be possible, right?

In case you missed it, GNU/Linux market share is constantly increasing.

Slowly. I'm not seeing any sort of revolution. I'd be surprised if Linux passes Mac OS X on the desktop, much less become a serious contendor to Windows.

You have yet to explain why you think this behavior is desireable or at least give some sort of excuse for this behavior.

Why should I have to explain anything?! GNU/Linux is not Windows. Nor will it ever be. People will not use it because it looks and acts like Microsoft decided Desktop Environments should act. It will act as the people who actually work on it have found out works best and those who don't like one method always have many other choices (Have you tried KDE? XFCE? Enlightenment? Fluxbox?). The power of the system is that it works better than Windows, not that it is similar enough.

It's not windows enough for you, fine, don't use it. That's subjective though and has nothing to do with how good the system is. People who are not stuck on windows usage have no problem getting use to it whatsoever.

In the end, I don't even understand why the rampart GNU/Linux bashing in this thread is warranted. As soon as someone posted someting GNU/Linux related, people came out of the woodwork to say how much it's not like windows and how they don't like it. This has nothing to do with the conversation.

Slowly. I'm not seeing any sort of revolution. I'd be surprised if Linux passes Mac OS X on the desktop, much less become a serious contendor to Windows.

We don't need a revolution. Slow and steady wins the race. People thinking 15 years ago that GNU/Linux might be on most servers in the world would have been laughed at. People thinking 10 years ago that it would be considered by many people (Of those who are even aware of its existence that is) a very viable alternative to MS Windows would be laughed at. People 5 years ago saying that major hardware companies would be selling GNU/Linux prepared machines would have been laughed at. All these things happened and as the coming economic problems intensify, the freedom and gratis of the OS will become more and more appealing compared to the high prices for macs and the lockdown of windows

Why should I have to explain anything?! GNU/Linux is not Windows. Nor will it ever be. People will not use it because it looks and acts like Microsoft decided Desktop Environments should act. It will act as the people who actually work on it have found out works best

This is a pretty silly way of looking at it.

"I'm going to make a game. It's going to be the best game ever. Only, you won't be able to select your units by left-clicking on them. You'll have to press Ctrl+Alt+Shift+Mouse click to select a unit. To select multiple units, you'll need to have Capslock depressed when doing Ctrl+Alt+Shift+click. No, I don't care that left click was unit select for over a decade, I think it works better this way."

Nobody's going to play that game. Like it or not, Microsoft as the market leader in OS sets the standards for the UI design. If you want people to turn to another OS, making it more difficult to use negates any possible advantage it may have over Windows.

The point Cobra is trying to make is that if Linux is ever going to come close to competing with Windows, it at least needs to be easy to get into for the average user. Average people are not going to bother with something they don't understand and find a hassle. They'll just stick with Windows which works and is easy to use.

I've tried Ubuntu a few times, 8.04 being the latest, but the overall impression i've got from them is "Why am i torturing myself with this". Back when i had Ubuntu 7.something installed, i wanted to enable DVD playback on it, and finally found the decoder necesessary, that then informed me that the software "could be considered illegal in your country" (or something like that), 10 minutes later i was re-installing XP. I later found out that decoders aren't illegal, but decrypters are. BTW i almost got Doom 3 working on it, but the OS hated my Radeon 9600 i had at the time, i wonder what it would think about a Geforce 9600?

There seems to be a new Ubuntu again (8.10), and i have a hard drive with XP installed that i haven't used since Vista, it's tempting me to give it yet another go, but on the other hand, Windows 7 public beta should start tomorrow, and i've heard plenty of good about it...

Nobody's going to play that game. Like it or not, Microsoft as the market leader in OS sets the standards for the UI design. If you want people to turn to another OS, making it more difficult to use negates any possible advantage it may have over Windows.

People keep talking about the "normal user" as if you know what's in everyone's heads. Truth is that most people do not switch not because of the difficulty of doing so, but because they are not aware of it or how easy it. You spewing FUD don't help either.

And in any case, in case you didn't notice, people did make games that differed from anything else that came before them, and not only did people play them, they created a whole genre of gaming. Your example is unfortunate because it undermines your argument. Check what happened with Wolfenstein 3d, Dune 2, Masters of Orion, Elite, Civilization and many others.

People keep talking about the "normal user" as if you know what's in everyone's heads. Truth is that most people do not switch not because of the difficulty of doing so, but because they are not aware of it or how easy it. You spewing FUD don't help either.

If you're going to accuse me of generalizing, the least you could do is not make an even worse generalization in the very next sentence.

And in any case, in case you didn't notice, people did make games that differed from anything else that came before them, and not only did people play them, they created a whole genre of gaming. Your example is unfortunate because it undermines your argument. Check what happened with Wolfenstein 3d, Dune 2, Masters of Orion, Elite, Civilization and many others.

What's that supposed to show me? In Wolf 3d you used arrow keys to move around. The only thing that's changed is the usual FPS keys are now WSAD for movement and we got the mouse to look independently. In Dune 2 you selected a unit by clicking the mouse button. In any RTS today, you still select a unit by clicking the mouse button. I could go on. The games evolved, but their UIs still have the same basic elements, and for good reason - they are intuitive and easy to use. If you spend years playing Starcraft, and then try to switch to Red Alert 3 you don't need to turn everything upside down because other than a few shortcut keys and a bit different UI layout, their control schemes are remarkably similar.

Here's an example that actually works, though: Fallout series. The first 2 games and Tactics were all Isometric. Fallout 3 is first/third person. Go take a look at how much flak Bethesda got for it. People who played and loved the old Fallouts had gotten used to the way the old games played. Fallout 3 plays drastically different, and a lot of people don't like it.

I'm pointing at something and you're looking at my finger...

I was pointing out that BEFORE all these games I mentioned, this type of gameplay did not exist! By your logic, all those games should have failed.

I think Source engine games are fairly well suited to this, because they include TF2 and Counterstrike. These two games are things that people keep on their hard drive, and play fairly regularly, often for short periods. On top of that, they're played socially. Basically, someone who was doing other things on Linux would find it quite easy to just jump into a game, whereas at present they have to reboot their computer. For some big epic that you play through once in much bigger sessions, rebooting into Windows is less of an issue.

I think turn based games might be well suited to Linux as well, because people often play those by task switching to them while they're waiting for something else to finish.

As for OpenGL, games developers used to use it quite a lot about a decade ago: there was a brief period when you could choose between OpenGL, DirectX, and Glide (Voodoo cards only). OpenGL has fallen in popularity because it's failed to keep up.

The core point remains that developing major game for Linux (along side of a Windows version) is a big undertaking unless you're on OpenGL from the start, with a lot of financial unknowns. A lot of risk for an unknown reward.

Additional QA, additional support (community-provided support isn't a viable option for a retail product), additional developer time & training. There's a lot of cost wrapped up in there. You can say that you're sure there'd be tons of people waiting to buy a game that runs natively on Linux, but there aren't any hard numbers. It's all speculation, and it's hard to justify spending a lot more money to chase a guess.

(Have you tried KDE? XFCE? Enlightenment? Fluxbox?)

Recently, KDE, and a long time ago Enlightenment. KDE is better for the Wiindows user, but seems very incomplete, with many keyboard shortcuts not assigned.

It's not windows enough for you, fine, don't use it. That's subjective though and has nothing to do with how good the system is.

In other words when it doesn't work for me, I'm being subjective, but when it conveniently fits your (currently undisclosed) definition of "how good a system is," you're being objective.

I mentioned quite clearly a specific scenario that a user might face. How is that being subjective?

I think you are not using the word "subjective" in a rational manner. I would suggest defining your use of the term more clearly and not just throwing it around as a catch-all way of beating your own breast and pretending you have secret, superior knowledge of the internal workings of Linux.

I was pointing out that BEFORE all these games I mentioned, this type of gameplay did not exist! By your logic, all those games should have failed.

And I was not talking about gameplay but UI. In my original example, I specifically said it would be an RTS - a genre which has existed for a very long time. You're just trying to unsuccessfully twist it.

Your analogy doesn't even apply because Linux isn't a brand new thing that's never been seen before. It's an OS. It has to do much the same things as Windows or Mac OS. It may do them differently, but it still has to do the same things to be an OS.

The major difference, then, becomes the UI because it's the first thing people notice.

Your analogy doesn't even apply because Linux isn't a brand new thing that's never been seen before. It's an OS.

My analogy still works because they're all games, and they were games that were different from the other classic games that existed before them. This is actually a big problem of the current day where a game costs millions to develop. This is why you see almost no new types unless you're Will Wright or you're building on something that you know has a market (Demigod)

But fine, if you want to say that only distinctions between RTS count, then the difference between Windows and GNU/Linux is as big as the difference between C&C and Warcraft. Certanily if you're trying to build with peons and have more than 6 units selected you're going to get frustrated...so warcraft should have never been a success. it was too different than the GUI all RTS should have followed...

In other words when it doesn't work for me, I'm being subjective, but when it conveniently fits your (currently undisclosed) definition of "how good a system is," you're being objective.

Yes, because you're saying "I don't like it" and that's the definition of subjectivity.Your examples of what others would find as problematic do not apply because they are examples from your own perspective. Case in point, my (computer illiterate) girlfriend who had no problem and has now become its largest supporter just because it works so well.

Yes, because you're saying "I don't like it" and that's the definition of subjectivity.

I am saying that certain features do not work as expected, if at all. Just because one portion of my posts refers to personal tases does not automatically mean the entire post is about personal tastes. You're cherry picking invalid arguments and ignoring valid ones. Just because I have some personal opinions does not mean my entire line of reasoning is based on personal opinion.

Your examples of what others would find as problematic do not apply because they are examples from your own perspective.

There is no "perspective" in attepting to drag an icon from one location or another. Either it works or it does not work.

Case in point, my (computer illiterate) girlfriend who had no problem and has now become its largest supporter just because it works so well.

So the opinion of your girlfirend is a more objective measure than specific scenarios? I await a better explanation, as well as a definition of "subjective" that actually makes sense.

If you really want something objective, you do as Microsoft has done with Office 2007: You gather statistics. I don't know if you have looked at Jensen Harris' blog, but they took a lot of time, a lot of research, and a lot of measurements when they were trying to figure out what worked and what didn't. They didn't guess or philosophize about it - when there was more than one way of doing things, they tried both and gathered numbers about how both methods were used.

UI design is part subjective, yes, but it can be objectified, planned, and tested. We can in fact know what most people prefer, and know how people generally work with and use different configurations. You want to know why Microsoft is so popular? Because they take the time to gather the statistics and crunch the numbers, that's why. This is something Linux has been weak at doing, and it shows.

In addition, there are in fact some basically accepted principles that go with UI design: Decreasing the number of clicks and/or keystrokes, for example. Pretty obvious: The less time you take to do something, the more time you have for other things and the more productive you are. You design software in such a way that the most commonly used functions require the fewest clicks and keystrokes. And by "commonly used" I don't mean "the developer's best guess," I mean "we've done the research and figured it out."

So I ended up looking up how to add the Trash. What did it take? Well, I was told to open up the terminal to launch a GUI application, which looked a bit like a Windows registry editor, and locate a setting somewhere I could tweak. Not intuitive at all. I'd like to know the "perspective" in which this actually makes sense.

I am saying that certain features do not work as expected

Who said it's expected? I can find a lot of things that are expected in GNU/Linux and do not work in Windows.

There is no "perspective" in attepting to drag an icon from one location or another. Either it works or it does not work.

The problem is that you're assuming this is a big problem and that the things that bug you are going to bug everyone. That's the subjective part.

So the opinion of your girlfirend is a more objective measure than specific scenarios? I await a better explanation, as well as a definition of "subjective" that actually makes sense.

No, but it proves that opinions differ and what you say has very little merit as it's based only on your own preferences

If you really want something objective, you do as Microsoft has done with Office 2007:

No, because you don't have to. Microsoft tries to decide from above and judging by the way Vista was received, it does not always manage it. If someone does not like how windows works...tough.

GNU/Linux gives everyone a say on how the system should be, if you don't like how it works, you're free to make it as you think it should be, and that's what people are doing, democratically and socially. That's statistics that have relevance.

If you don't like what everyone else likes, you have again the freedom to do it your own way but you don't get to decide how everyone else should work.

UI design is part subjective, yes, but it can be objectified, planned, and tested. We can in fact know what most people prefer, and know how people generally work with and use different configurations.

And the reason why the GNU/Linux GUI is how it is is because most people who use it prefer it that way.Those who don't either use another one or make their own. In windows, microsoft decides what people should like.

I'd like to know the "perspective" in which this actually makes sense.

Most people don't care to move the trash to the desktop?

But fine, if you want to say that only distinctions between RTS count, then the difference between Windows and GNU/Linux is as big as the difference between C&C and Warcraft. Certanily if you're trying to build with peons and have more than 6 units selected you're going to get frustrated...so warcraft should have never been a success. it was too different than the GUI all RTS should have followed...

No, because both are equally intuitive You either have a unit that builds everything, or a generic build list. Either way is perfectly fine because they are both logical and easily understood and most importantly they are both easy to grasp.

The same comparison can't be made between Windows and Linux, though. It's made even more blatantly obvious by your refusal to answer Cobra's questions about the logic behind it, instead trying to argue that you don't have to answer because it just is.

No, because both are equally intuitive

GNU/Linux is just as intuitive to someone who is not used a lot to windows. For someone who is used it may get a bit to get used to the different style but once you get used to it, it works just as well.

PS: The RTS games are not intuitive. Otherwise they wouldn't need tutorial scenarios to teach you the game. Now imagine you were playing blizzard RTS for a decade and one day uou decide to play C&C and as an expert of RTS, you decide to skip the tutorial. Would you find it intuitive? Would you be able to figure it out immediately? Of course not. You'd get there eventually but it's quite probable that such a trial might get frustrating.

Quoting Zoomba,

reply 13

The core point remains that developing major game for Linux (along side of a Windows version) is a big undertaking unless you're on OpenGL from the start, with a lot of financial unknowns. A lot of risk for an unknown reward.

It was my point... Stardock is mainly a PC game company... why start with opengl... it is not financial wise to recreate the sins full code for a few opengl fan...

If one day, Stardock wish to port Sins to console... the good choice will be opengl because only one game station ( Xbox ) support directx... in these case, the financial return will be bigger that the investment...

Annatar wrote :

It has to do much the same things as Windows or Mac OS. It may do them differently, but it still has to do the same things to be an OS.

The major difference, then, becomes the UI because it's the first thing people notice.

About the Gui, Linux can work in the same way that Windows or the Mac... it is nothing new... simply people ( public ) don't know about it... usually, linux distro who look and feel like other OS are not welcome by the hard linux guy with closed mind... these hate all who seem to be a other OS... yes, in the Linux world, we have some extremist too...

A good example is the two screenshot below... they are from Linux OS who look and feel like other OS... sure that if these distro was promoted a little more by our linux community, our number of linux user will grow more fast... and it will ease a lot the transit from one OS to the other...

Time to take a look at Windows from a Linux user point of view.

Installing Windows

Let's start with the installation. You boot from a Windows XP CD and text mode installer pops up. The installer cannot find a a usable partition. Alas, all space of the hard disk is in use by Linux partitions. The Windows installer has no functionality to reduce a partition in size. How pathetic.

So you try to reboot to Linux, but you are unable to, but because the Windows CD has overwritten the master boot record. What nonsense is this? You did just boot the Windows installation CD without giving it permission to change anything. Why does it make my computer unbootable. Shouldn't one sue Microsoft for making computers unusable?

Anyway, you don't bother because Linux vendors have long ago adressed this problem. You insert your Linux installation CD, select "boot installed system" and when your Linux installation has booted run Lilo to fix the master boot record. You shrink one of you partitions to make space for Windows and boot the Windows installation CD again.

This time the Windows installer is able to find free space to create a partition and start the installer. After a short while the installer becomes graphical, displays some nice looking screens and about an hour later the system has been installed. Obviously your master boot record is gone again. Windows appear to have a boot manager to boot multiple operating systems, but it appears it cannot boot Linux. How's that possible in the 21st century?

So you fix your master boot record again using your Linux installation CD and move on with Windows. There are quite a few problems with the system as installed. Windows has put the system in a low resultion and terrible refresh rate. You think you may need a driver, but the installer has not made your ethernet interface functional either. You have to resort to the floppy disks of the 80's to get an ethernet driver on the system.

Talking about drivers, why do I need to do all this shit? My Linux installation CD was able to deliver the system in a "ready for use" state. Why is Windows unable to do so? Where do I find drivers. You ask a Windows expert and he says you need to go to the website of the chipset manufacturer of your video card, sound card, ethernet chipset and so on. How do I know what is the chipset manufacturer of my hardware? Windows is unable to tell you.

So, you boot to Linux, execute an "lspci" and you have a list of all devices in your system. Armed with this information you boot back to Windows and are able to visit the website of each chipset manufacturer and download drivers. While you install them, some video driver tells you need to download .NET to make it work. What nonsense is that, downloading large codebases to get your hardware working?! So, you spend another and half an hour waiting for the .NET download to complete.

Next a "Windows update" round, which takes another 1,5 hours.

If you are lucky, after about half a day, you have a system that is somewhat in an usable state,.

Using Windows

The operating system has some kind of start menu that changes all the time so you never find the application you want to use, so you end up clicking on the "All programs" option all the time. Because that is too much work Windows allows you to create "icons" on the desktop so you can launch your favourite programs easily.

Unfortunately the desktop often hidden under a few Windows and your desktop quickly gets overcrowded. Windows also annoys you with some kind of "thinking menu's" where many usefull features are hidden. The user interface also suffer from a problem that is called Bobification in the Windows world, the software vendor has confused user friendliness with talking dogs, annoying text balloons etc. Some of these Bobifications can be switched off, other cannot and keep annoying you.

Windows is, even today, anno 2008 not capable of accessing Linux partitions. That means one cannot access his documents on the Linux partition. You can boot to Linux and copy the files to the Windows partition, but you quickly encounter the problem that the Windows Notepad is not capable of displaying files with Unix line-breaks. Such a bug, which is meanwhile present for two decades in this program

The web browser included isn't very comfortable. For example, the middle mouse to open a link in a new Window doesn't work, the browser does not have any popup block, spelling checker. Support for modern standards like SVG , xhtml or html 5 is missing, just as well as support for more recent CSS standards. Also it has the strange habbit to redirect incorrectly typed URL's to Microsoft's search service rather than giving a helpfull error message. It's privacy features are weak: It tries to remind passwords without asking permission for that, it offers few options to remove privacy sensitive data, offers little options to control cookies.

The included e-mail client almost forces me reply to e-mail in a nettiquette incompliant way. It's IMAP support is really slow and supports very little of the IMAP standard, you can't really use it well for my mail folders. The anti-spam options are pathetic. Must-have features like an option to show all e-mail headers, a bounce-button, the abiity to set usefull headers like Reply-To are totally absent.

The Windows command-line is terrible. Copy-pasting for example cannot be done with mouse selection, it must first be requested from the context menu. Essential commands, like equivalents for test, for, find, grep, tr and so on do not exist. Some of the missing functionality like finding files can be done with the GUI, but doing anything with it is not possible. It is also very hard to find (perhaps because it is such piece of crap?).

While Windows does allow one to access files on a Windows server, an SSH client to access files on a Unix server is absent, just well as an NFS client. A command-line ftp client is included with the OS is present, but it is so pathetic that one would pray not to have to use it. Unfortunately, ftp is often the most common sense to transfer the files to/from the Windows machine.

Windows is so unusable that you start to put lots of work into installing lots of third party utilities for its shortcoming. You install tools like WinSCP, Total Commander, Winrar, Irfanview, Putty to make your life somewhat bearable. Unfortunately this is again a lot of work, and also, many of these tools do not come for free. Many Windows users do not care anyway, because they pirate everything.

Maintaining Windows

Windows does not come with a package management system. This means that installers cannot automatically requisites for a program automatically, it is up to the user to find them. This also means the system cannot clean up unused middleware automatically.

Some custom installers do allow testing the integrity of the installed files, many do not. Even if it is possible, this functionality is often not automatable to system administrators cannot automatically check for errors in the installation of programs.

Scripting common maintanance tasks is very hard, because most tasks need to be performed using the GUI, causing a lot of handwork in administrating the system. Many settings can be configured using the GUI in either the control panel or "system console", but many usefull setting are inaccessible through the GUI. In such case one much make the setting through the incomprehensible registry. The registry is so uncomprehensible that it is very hard to find incorrect settings in it, and also keep track what various programs change in it. One must hope that the programs do their job well, because once there is an error in the registry it is very hard to find and solve the error.

The verdict

Windows is unable to give any convincing alternative as a desktop OS, wether looking at installation, usability, or maintenance. Many essential functionality is missing, otherwise functionality is present in a very primitive way that no professional user wants to use. There is even functionaility that is outright annoying. The security issues haven't even been discussed yet, as a Windows user depends on external tools to keep his computer clean from malware like virusses, trojans, adware, and spam botnets.

For a professional computer user, Windows can only be regarded a toy OS. Because Windows is good for playing games, and other forms of entertainment on the computer.

What does any of this have to do with the topic of this thread? The entire point was to talk about the Source engine being ported to Linux and the implementations there of. What does the faults and advantages of various OS' have to do with this?

If you won't play a game because you need Windows to play it, then you clearly do not desire to play that game that much. You cannot complain about how developers should switch to a Linux capable method of development when the incentive doesn't exist for them. If developers could stand to gain from developing for Linux, they would already be doing so. It boils down to simple economics and has nothing to do with how easy an operating system is perceved to be. People have their tastes and using poor arguments such as "people from our room expect things this way" or "people in this room think that the way this room is is superior" are meaningless. This arguement has gone on long enough and has gotten no where at all.

~No Name McGee

Quoting No Name McGee,

reply 23

If you won't play a game because you need Windows to play it, then you clearly do not desire to play that game that much.

And a real Linux nerd ( like me ) will not complain too but seek a solution for make run the game within Linux... Sins is not from Linux but with a little work, it run perfectly... remaining problem and impulse have people working on it...

It boils down to simple economics and has nothing to do with how easy an operating system is perceved to be.

Again true... Big game business can allow some risk but for little one like Stardock, it can mean the dead if it don't work... and seriously, today, almost no game are ported to Linux... having them using open-gl make thing more easy but it don't resolve the main problem with windows game running on linux, who is online game not working ( for now )

The original post speak about DRM... these can be implemented on Linux... maybe with a strong encryption layer over... old securom security system are already implemented in Wine but the windows crack for these game don't work anymore... So, windows or linux version of Steam will certainly not change anything about DRM...

It was my point... Stardock is mainly a PC game company... why start with opengl... it is not financial wise to recreate the sins full code for a few opengl fan...

Stardock is mainly a Windows applications and customization company, not a games company.

The Forums Are Now Closed!

The content will remain as a historical reference, thank you.