So, no more install.sh scripts that you have to chmod +x, no more "undefined symbol" errors from games expecting old library files, no dependency hell (though, admittedly, I don't encounter those as much anymore).
That I can definitely get behind!
That I can definitely get behind!
The gap between programmer and user is Linux's biggest fault. For example Unity, developers recognised an issue with the UI. Instead of doing user studies and tweaking it for the better they canned Gnome and reprogrammed it. Now they're in the same situation but with an entirely different codebase.
Having a large company with a product that has to work will add consistency, and hopefully force Linux to start tweaking instead of redeveloping.
I might actually install it on my next machine.
2. I do believe your view of Unity is a bit odd. Canonical recognized that the state of Linux UI/DE was NOT going to make the leap from programmer. They saw this VIA user studies. That is when they decided to create Unity. And user study after user study Unity keeps coming out ahead of all the other DE in Linux and is on par with usability of current OSX. Point being, Ubuntu is going after users the right way. Unity in 12.04 shows this and the reviews speak to it.
3. Sometimes you simply have to rewrite. This is a fact (sorry Joel) in software. The challenge is to recognize when it is necessary and avoid it when not 100% necessary to achieve your business objectives.
In 2012.
The problem here is twofold, though I'm not necessarily saying these comments apply to you, you are apparently willing to endure some inconvenience.
First, people should always check hardware support before buying a laptop if they intend to install Linux. I'm not trying to be a neck-beaded dick (although I do need to shave), I'm just pointing out that many of the problems people have with Linux stem from well-known incompatibilities that could have been avoided with different hardware choices. Which brings me nicely to the second point...
There is a communication issue in the Linux community. We (I'm including myself here) need to stop over-selling Linux. It is not a computing panacea for everyone, and so long as hardware support lags behind Windows "selling" Linux as a drop-in Windows replacement isn't a good strategy.
Getting someone to install Ubuntu/Fedora/openSUSE isn't a victory for the community if his hardware isn't compatible and things don't work right. It would be better to tell him that he should think about purchasing a Linux-compatible computer in the future and perhaps directing him to a list of compatible hardware.
Sure. But that wont cure a fundamental problem with the X11/Xorg infrastructure.
You cannot quickly and switch desktop orientation, resolution and layout (not to mention drivers) seamlessly like on Windows without killing your entire X-session. Killing your entire session. Jeez.
At the root of this problem is an architecture which is lagging on what Microsoft could deliver in Windows 98. You cannot make Windows and reboot-jokes while you have this Achilles-heel dragging your video subsystem. You seriously cannot.
And no new version of Ubuntu can overcome this underlying issue, no matter how much fluff and candy they add elsewhere, as long as they rely on the ancient relic that is X11.
As far as I know, when Google decided to make Android and ChromeOS (a complete failure as far as any metric goes, but still worth mentioning) they decided:
1. to base it on Linux, because that makes sense and
2. not to base it on X11 because they wanted it to be flexible and they wanted their users not be limited by what architects could think of in the 1984.
For end-user Linux to ever become user-friendly Xorg needs a complete architectural overhaul. I cannot see anyone arguing this.
And even if seemingly everyone agrees on this, nobody is pushing forward for it to happen. Make of that what you like.
The only thing that is really ugly in X11 is the protocol, but so is win32 api :)
On my Ubuntu system, I can change orientation, resolution, and layout from a little control panel without restarting the session. I have no idea if Xorg restarts; the superficial behavior is almost identical to Windows. From the user's point of view, it doesn't seem to matter.
> For end-user Linux to ever become user-friendly Xorg needs a complete architectural overhaul.
Every mainstream desktop environment seems to have handled the major aspects of display usability just fine. The only serious issues I've encountered have been one-off bugs related to drivers.
> I cannot see anyone arguing this.
Everyone agrees X is a mess. But why must we continue to use X and not Wayland?
#!/bin/sh
xrandr --output LVDS1 --off
xrandr --output VGA1 --mode 1920x1200 --rotate left
xrandr --output HDMI1 --mode 1920x1200 --left-of VGA1 --rotate left
In all seriousness though (and this is a question that everyone should ask themselves semi-regularly in a product-consumer situation): why do you matter? If you don't have a compelling answer to this question, it may be that you don't.
The problem with people arguing about DRM is that fundamentally most people don't seem to care. Video games are a billion dollar industry, but non-DRM'd games are not.
You can vote with your wallet, but the data seems to indicate that your wallet is too small to matter.
> Video games are a billion dollar industry, but non-DRM'd games are not.
It only should increase the incentive to avoid buying DRMed games, and specifically support those who aren't scared to release theirs DRM free.
For example look at the latest Humble Indie Bundle. Linux users pay highest average for the set. It shows that they encourage developers to produce Linux games more than users of other OSes (obviously since Linux games scene is lacking). You can argue, that Linux games market is miniscule, and if so, why even bother to send your message in such form? But you still see people showing appreciation, and developers can take note of this.
If we assume every bundle gets a total of a million dollars, shared between 5 games or so, that makes 200 000 per game, minus the donations to charity and the fees to the Humble Bundle organization. Let's say 150 000 dollars per game in the end. 15%-20% is 23 000 to 30 000 dollars, just for the Linux Port. If you are an small team of indie developpers, I'd say that is not negligible at the end of the day.
In my opinion Kickstarter (or grant) model is infinitely better, because it doesn't insist on treating a game like a physical object. The "selling software" model is horribly flawed. Writing software is like research - a big cost and effort once, then very little cost to benefit from it. It is fair to have effort of research/software development proportional to reward the researchers/devs are getting. With "selling software/drugs" it's not proportional.
http://www.forbes.com/sites/danielnyegriffiths/2012/05/18/th...
Hell, one of the biggest reasons I patronize the Humble Indie Bundles is because I can get the games into my Steam account.
There is a positive side of Valve's interest in Linux though. It'll encourage others to pay more attention to Linux, and eventually more DRM free games will be available too.
I don't think that follows.
Steam takes that to the next level -- because they can add fines of real money (losing your account), and actual permanent bans (of credit cards and real identities) as the cost of cheating, the gaming community embraced it and cheers for it.
So, we are in a situation where the vast majority of the player base considers DRM not just acceptable, or good, but a strictly necessary part of the product. The things that are commonly considered the evils of DRM (ability to take away ability to use product) are instead considered a primary selling point. How's that for food for thought.
And to single player games this issue doesn't apply at all.
GMs do not have the power to that job, unless they have something that's approximately equal to DRM. Catching a subtle cheater is a much, much bigger job than cheating. If someone cheats every day, you'd be lucky to catch him once a month. To actually stop cheating from happening, you need penalties that are high enough to make that once a month count -- so permanent global (so they won't just move on to the next server) bans (and ip bans don't cut it), or failing that, monetary punishment.
Tell me how your third party arbiters achieve that without essentially implementing DRM?
> And to single player games this issue doesn't apply at all.
Except when there is publicly visible scoring/achievements system that you want to keep pure. >90% of new games on the market are either mp or do that.
From the player perspective, I'd choose a multiplayer game based on the community. Sometimes community can identify cheating just by observing something unnatural, and those users will quickly be boycotted. So GMs are needed to intervene when human decision is necessary. They not necessarily have to detect the issue themselves first.
(Correct my if I'm wrong, that's what I remember from reading about it before, but something could have changed recently.)
If they have publicly promised it, and people have taken action upon it (for instance buying games that they wouldn't have bought otherwise), then if they fail carry through on it you have grounds to sue for promissory estoppel. See http://en.wikipedia.org/wiki/Estoppel#Promissory_estoppel for more on that.
Winning said lawsuit is not a sure thing. But the pledge may wind up being more binding than it would appear on the surface.
"Unless there was some situation I don't understand, we would presumably disable authentication before any event that would preclude the authentication servers from being available."
So, that's pretty clearly not a guarantee of anything. And furthermore, as another poster there said, the thing Valve employee doesn't understand is bankruptcy. Unless there is a legally binding agreement, they won't be allowed to just give away such a valuable asset. Bankruptcy doesn't mean they fire everyone, shut down everything, and burn down the buildings. It means they sell off the profitable parts and try to keep them running. And the ability to control tons of DRMed games people have installed is very valuable.
A real, legally binding promise would be something like KDE had with Trolltech. Trolltech explicitly (real contract, real lawyers, etc) guaranteed, even in bankruptcy or change of ownership, that if they stopped making open source releases, then Qt would become available under the BSD license. (This was more important back before Qt was already GPLed, as it is now)
I personally love and trust Valve, but I understand that others may have a different risk tolerance.
Please point out where I said it wasn't.
But if you're going to be snarky and pedantic about it, please point out where I said you said Steam isn't DRM.
For example requiring one to connect to their server prior to playing (implicit requirement of having an Internet connection) is a ridiculous restriction.
What overhead? You mean the 100MB RAM and 0.1% CPU the Steam client takes? Wow, that really is a high price to pay for the ability to re-download any game I own on any PC, and the automatic updates.
requiring one to connected to their server prior to playing is a ridiculous restriction.
Steam offers offline mode. You only need connect to their server when you download the game. Now, some games are more restrictive, but those are not made by Valve- only offered on their platform. That was someone else's choice.
It is a good model, IMO, and similar to what Amazon does when an e-book is priced at $14.99 ("This price was set by the publisher"). Let other companies do what they like, and let the market decide for itself.
Clearly, if you hate DRM for philosophical reasons, you will hate Steam. If you're pragmatic, you may love it.
I do use steam. A lot. Unfortunately, everything sometimes doesn't just work. For example, I've been left without access to my games a number of times due to internet problems (either in general or just with my gaming desktop) and while offline mode could get around that, I need to be connected to the internet to put it into offline mode and its rare, if ever, that I'll know about internet problems in advance.
So not everything just works for me.
Besides that, though, Steam is the first store in which I've ever impulse bought games - and I've done so multiple times now and probably the majority of all games I own are through Steam. They just make it so easy to find games I want, check the forums for criticism, buy the game, download it and then connect to friends. That experience is one they've done very well and why I still use it despite having had problems. Luckily my account has never been locked - thats a horror story I don't want to experience...
Dependencies is still a problem, however. There are still a few cases where I had to spend significant time to make a game run correctly because of dependencies. And the online resources on this matter are scarce.
But even on a Linux backend, it's possible to introduce specific formats encompassing all needed libraries. That's what being done on the Open Pandora, where a single PND file contains, in a zipped format, the executable, all the needed libraries, and the data for the game. That way, you avoid the problem altogether.
I'm tempted to install another partition of Windows with nothing on it except steam and the drivers necessary for me to play games, to see if it'll make starting steam any faster.