We missed the "year of the Linux desktop". That's not to say it happened and we didn't realize it. What I mean to say is it's not going to happen. I've been chatting with people in back rooms and on jabber about it for over a year now. Something didn't seem right to us and I think that's what it was.
The desktop, which I might as well refer to as “client” just doesn't matter as much any more. We fight for open standards and open protocols and as a result lots of providers serve content that can be viewed on several different hosts, devices, etc. We win! We also lost.
I don't have any pull with any desktop people anywhere but Luis Villa does. At GUADEC he basically told everyone that the web is the future. (Sorry if I mis-paraphrased that Luis). He's not wrong. Gnome, KDE, XFCE, whatever, none of them can deliver the content I care about to all of my devices in the same way from the same source.
That's what people want though. What they use on their client sides is mattering less and less and the incompatibilities between devices and platforms is becoming a bigger and bigger hurdle. That means as non-global features go up, so does cost to deployment. Sure if everyone ran just X and just Y it wouldn't be a problem. That might have been feasible 10 years ago but we're not anywhere near that today. Even C level technology haters love their smart phones now. These costs are getting higher as time goes forward.
I'm not talking about development costs either, I'm talking about deployment and use costs. It used to be only enterprises needed to worry about content convergence. Now though it's not that uncommon to find someone walking around with their own personal enterprise. Some sort of media center on TV, a desktop, a laptop, a phone and a tablet.
“Ok” you say, “Need has been established, we get it. Lots of incompatible clients, everyone wants them, lots of people have them. Now what?” Well, that alone isn't quite enough to cause a massive shift. We'll call the need a “pull factor” for now though. Lets look at the push factors and the disruptive technology that's going to make it all happen: HTML5.
We've got developers who are writing code and need to figure out how to get their app to users. Do they write for the tablet, the TV, the latptop, and the phone? Should it be for android, iphone, OSX, Linux and Windows? They could do this but they won't. They'll write it once, with HTML5 as the UI. It will work everywhere, it will be a very vibrant feature rich experience for the user. Most of it will be all cloud based (barf) and they won't have to worry about their 5 year old desktop at home falling apart and losing baby pictures. They won't have to worry about doing something at work and having it be different when they get home. It'll just work every time, from any device they use it from. It will look the same or close to it, it will be convenient, it will be everything our desktops aren't today.
We had our shot at the desktop that competes with Windows. We lost. Man did we lose. Canonical was our best shot and so far has proved unsustainable. Getting a little bit of money from lots of users for a free-ish desktop just isn't profitable. OSX won that battle. They did it pretty late in the game too, 5 more years and they'd have lost as well. But they won. Sure, we compete on price and freedom. Not nearly enough people care about either of those things though. Not if the functionality isn't comparable.
So where does that leave us? Well, right now we (not Fedora, I'm talking about the greater Linux universe) are still sort of fighting a war that doesn't matter anymore. The world is moving on without us. Some groups realize this. Guys like Luis saw it from a mile away. The free software world is poorly positioned for this new software as a service model. Places like Google and Salesforce are very very well positioned for it. If you've seen Google's apps lately, imagine how they'll be when HTML5 is standard in most people's browsers. We can't compete with that, not right now.
If that wasn't bad enough, look at our precious licenses. Some of the real secrets and real powers of Free Software lie in our licenses. Almost all of them are woefully ill equipped to deal with this new “someone else is running it” model.
There is still a place for us. Linux is still best positioned for servers in this cloud eat cloud world. We'll do great there. Also, fighting for open API's, open standards, free code is something in which we will continue to excel. The idea of a feature rich “Desktop” experience, however, is now just a quaint memory of the old days. Just like we now ironically think back on the thin clients and mainframes, someone will think back on our window managers with the same nostalgia.
Page Summary
arnebab : optimistic future [+5] 
smoogespace.blogspot.com : I don't understand [+1] - (Anonymous) : Content in every device [+0]
- (Anonymous) : We've not lost, we just need to retarget [+1]
- (Anonymous) : Cloud [+1]
- (Anonymous) : Oh God [+0]
- (Anonymous) : I agree [+0]
- (Anonymous) : Verizon-Google [+1]
- lordmuck : You had me 'til the end [+0]
- wakko : (no subject) [+0]
- dowdle : Winners and losers? [+13]
- dingyichen : HTML5 might be the new "UI programming language", but will people follow that standard? [+0]
- intosimple.blogspot.com : (no subject) [+0]
- (Anonymous) : Who cares? [+0]
- (Anonymous) : ugg classic boots [+0]
- (Anonymous) : ugg classic boots [+0]
- (Anonymous) : Microsoft Office 2007 [+0]
Designed by 
yoksel
Comments
And for “we lost”, please see this clean evaluation of browser stats: http://draketo.de/licht/politik/leserbr
With google translate: http://translate.google.com/translate?j
We gained about 16% desktop share against windows+OSX in 2 years, and on smartphones free software is winning big – just think Qt and qtopia.
And what’s a phone different from a low-power desktop? Though much faster than anything I owned 7 years ago.
And look at KDE’s akonadi framework, which makes it very easy to sync information without the need of a central server.
And look at the counter-movement of open diaspora and gnu social, moving the data back into the desktop, because people realize more and more that it’s a bad idea to trust a company with your personal data.
Please remember: realism is often just disguised pessimism and lack of trust in the things which work.
As far as counter movements like diaspora and gnu social. They aren't anywhere near the levels of facebook which, instead of being a counter movement, is an actual movement.
And from the „look“ of the data the random statistical error should be far lower – see the clearly visible pattern of monthly changes in the original data.
And the fact that there are counter movements with much enthusiasm (open diaspora got 100.000 startup money from private people in a few weeks) shows that more and more people become aware of the problem behind the facebook movement and want it fixed.
Sorry I couldn't let this one go. We run an AGPL application in Infrastructure. Not only is it a PITA to be compliant, but it's easy to bypass. Fedora Community is a turbogears based app. So if we were so inclined. We could present the user with non-agpl code at the turbogears level (It's MIT). We could do it at the wsgi level, we could use an apache alias, or even a rewrite at our proxy layer. Is it easy? no. Is it convenient? hell no. But if someone wanted to, they could and the end user wouldn't know. As much as I like the idea of the AGPL, it too isn't as solid as it seems like it is.
My prediction is it won't get wide adoption at all. If I'm a company looking for code to work with. The AGPL provides me with the fewest options of use. So the legal team will just suggest I not use it. It's a great license for developers to use to protect their code. It's a non friendly way to share it though.
Either it isn’t efficient, then companies don’t need to care much about its restrictions and have no reason to reject it.
Or it is efficient, then it does its job.
And if someone wants to bypass the GPL, that can be done the same way you describe for AGPL. Still most people don’t, because it’s much easier (=cheaper) to play fair. I would rather assume that this is what will happen with AGPL, too. It makes it expensive to cheat with networked programs, where it’s very cheap to cheat with plain GPL. So people will comply with it.
They have to serve the sources for the AGPL parts anyway (as I understand it, it doesn’t matter how you hide the code: it’s accessed, so you have to offer it), and sending changes upstream gives them cheaper maintenance.
I have never understood why a desktop can't be define it as the background of your iphone or google droid. It has menus and fires up applications and while some of it is on the web, for the most part you are still running games and fart-noises locally. But even that does not help define how one can 'win' without some sort of line in the sand.
So I wouldn't say we lost as much as we could never win.. [on the other hand one could say that if a desktop is just what is presented on a screen then Linux sort of did win as Droids become more ubiquitous... it just isnt the desktop that people were thinking it would be.]
And Google uses Gentoo Portage for building the Chrome OS → http://sites.google.com/a/chromium.o
And Android can simply be seen as a mobile desktop.
> apart and losing baby pictures.
Because worrying about their cloud service provider changing terms of service or discovering its "free" storage does not include backup is better?
I don't think at all Fedora has lost it however I *do* think that the Fedora Desktop Team dream of competing with Microsoft by building a Desktop stack that deliberately *ignores* all the fine server software Fedora inherited from RHL has always been a pipe dream.
While they've chased the winning formula of the 90's others constructed successful user experience with the very software they were ignoring. There's no reason a floss desktop could not be insanely successful, if it provided all the services linux-based appliances propose today (network storage, network photo repository, network video center...) instead of the stupid "if it's not in the current user session it needs to be shut down and nothing happens if there is not someone logged at the desktop".
What's won is not the cloud, it's networked services. Users do not want the cloud they want stuff their other network gadgets can access (even if it's stuff running on the local laptop, or on auntie's desktop the other side of the country). Witness skype's success: its not about talking to a central cloud, it's about using remote resources, remote resources have no need to be cloud-hosted to be useful, and they can be provided by a fellow user. There are very few "cloud" services that require the interaction of lots of different users, anything that is used to share data among a dozen people only needs a user-friendly setup, an URL to share, and some peer replication to take care of availability.
The failure has not been to focus on the Desktop, the failure has been to focus on a Desktop defined by Microsoft in an age consumer networking was uncommon, slow and anecdotal.
This sounds like a job for the NetworkManager team ;-)
I think caps by cell phone companies will also keep The Cloud from happening. It did not take long for a cap with the Ipad.
You also have this thing called speed. I do not ever come close to max speed with my dsl.
Call me old fashion but I think even teens prefer apps if they know about them. My nieces and nephews are 9-15 in age. When they saw I was not chatting from FB but using Pidgin they wanted that on their computers.
So imo The Cloud for a lot people is not close. Oh and cable high speed internet in my area is a myth.
Google, netflix and companies like them are the ones making the bandwidth intensive applications. But neither of them (yet) provides you the network to use them. So they can just point fingers at the network providers while saying they're giving the customers what they want. It sucks, it's a crime but rural areas do and might continue to get the short end of the society stick on stuff like this. I just don't see rural communities as having any impact on this movement to remote hosted services at all.
I friggin hate consuming software in that manner.
Also, thanks for pointing out Luis Villa's post, missed that.
It also seems to me that if Desktop OS APIs become less of a development target, then users will be less tied to Windows.
I see the reaction of a lot of college faculty and students to Desktop Linux. Generally, I believe that attitudes about Desktop Linux have been improving.
Aka all that is old is new again, and we've been here before.
Nothing is lost, nothing is going way, technology just changes. But there is nothing fundamentally changing here: the balance between local power, remote power and network bandwidth moves around a bit, but the balance will tip back again. Meanwhile we can all get on with our lives; the desktop didn't kill the mainframe and the mainframe^Wcloud won't kill the desktop.
I don't think the Linux Desktop was ever really important in the way we thought it was, and even if it was exactly as we imagined, that time is probably a decade or more gone. I agree with you, I think we lost that battle before we even began to fight in earnest. Yet, Linux has won, and won big, in the server market. Everything from the Super Computing 500 lists to the various "Powered By..." icons in the footers of most websites says so. Linux has also won, and continues to win, in the mobile market. Android is the latest and most visible win, but Samsung, Nokia, and many others have been using Linux in their phones for years. Most people are using Linux without ever realizing it.
But that last bit has been both our goal and our curse. The height of usability is something that people can use without it ever being so in their way that they have to think about what they're using and how it works. As the mantra Robert Love and the Gnome folks started using many years ago went, "It Just Works." However, the downside to things just working is that nobody realizes and nobody concerns themselves with the branding and the marketing efforts that are required for the average person to have some concept of how valuable using Brand X over Brand Y really is.
Linux has played second fiddle to larger corporate interests exactly because we are a collective mostly of engineers, not sales people. We would have difficulty selling water to a man dying of thirst pushing his life savings into our hands for one drink because we'd spend hours discussing the merits of hydration and the process by which our water was filtered, bottled, and delivered and forget to actually hand over the bottle.
I believe that open APIs and open standards are both worth fighting for and somewhat of an inevitability. The courts are losing respect for the abuses allowed by software patenting, and it's just not in the interest of most companies to build yet another walled garden. Walled gardens work only when you've got a market share large enough to draw people in, and only as long as the practice is profitable. We can already see the cracks in the seams of Apple's walled garden and the market shifting away from them to more open platforms that offer a wider array of choices.
However, even HTML5 doesn't look like a silver bullet to me. After all, the web is really just another iteration of the old struggle between thin and thick clients. Push the computation to the client, pull it back to the big iron. The cycle has repeated a few times over the last 20-30 years. There are simply some things that native code running locally on decentralized hardware does better than centrally controlled, centrally processed applications. Also, because of miniaturization, it's easy to forget that today's commodity servers are equal or more powerful than the mainframes of 20-30 years ago. Today's mainframe is simply called "the internet", and I believe we'll continue to cycle between thin and thick clients for quite a while to come.
What makes someone a winner and what makes someone a loser? Linux on the desktop has tens of millions of users. While that might not be double-digit market share it is still a significant number of users. Anyone who looks at those numbers and calls them losing must have thought the game was winnable to begin with. It wasn't. FOSS does not spend hundreds of millions on advertising and billions on under-the-table deals with hardware makers... FOSS simply does not operate on the same playing field. The free hand of the market happens to be attached to a twisted arm. :)
In reality there aren't any winners and losers because it isn't a race. While those who are in it for money have certain ways to measure success, in FOSS you just make the best software you can and hope users will appreciate it. Of course the squeaky wheels always make the most noise and there are lots of complainers out there (see discussions on KDE 3.x vs. 4.x as an example) but that doesn't mean that vast majority of us FOSS users aren't very happy.
There seem to have been quite a few "if you can't beat them, join them" type posts on Fedora Planet lately. It all depends on if joining them means I have more freedom or less... and if I can tell the difference. The mobile device world is still a proprietary nightmare full of lock-in. Are they selling well? Yes. Does that mean we should just give up and join them? Definitely not. Same goes for Facebook. Just because Facebook is the new-AOL (a members-only castle of content) with 500M users doesn't mean it is the ideal way. Does that mean the new battle ground should be in FOSS social-networking? Everyone assumes that eyeballs means money so Facebook and Twitter will eventually find a way to make money... in a way that doesn't anger their userbase and make them go somewhere else.
If the Internet has proven anything, it is that it is hard to predict the future... and predicting the future now isn't any easier than it was 10 years ago. Who thought AOL would fall? Backing up a few more years, who thought Netscape would fall? Or Word Perfect? My point is that while the Internet has proven to not to be a fad there are a lot of companies, products and services that seem to have been. Facebook might be in a completely different position 5 years from now. Google could turn into a "has-been"... as well as Microsoft... and Apple too. I know those things are hard to imagine but if the past has shown us anything, it is that the future is hard to get right.
The *ONLY WAY* we lose is to give up or to "join them". Diversity is good... and the FOSS folks have really been a bastion of innovation even if many don't think they have been. Microsoft and Apple definitely need the competition or we'd still be using something close to Windows 98 and Mac OS 9. If we decide to give in and/or join them then not only do we lose but so does everyone else.
Perhaps you think I've ignored your argument about HTML 5 being the backbone of all future app development. If that is true, FOSS can move in that direction. That is natural and to be expected. The analogy that Mr. Torvalds used to give was that Linux is like a liquid that flows with the pull of gravity... to everywhere getting into every nook and cranny and become an endless broadly deployed thing lacking any real direction or intelligible strategy. That is the way it is now with FOSS, the way it has always beenand I'm guessing he way it will always be with FOSS. That isn't a bad thing. It has worked well for the evolution of life on earth. :)
We are way behind in HTML 5, huh? How did Richard Stallman feel when he started the GNU Project? What chances did Linus Torvalds have with his private little project started with a Usenet post? How far behind were they?
If you are frustrated / unhappy in FOSS perhaps it is time for a change. That isn't directed at anyone in particular... just a general statement. Don't worry, someone will fill in your spot.
That's a lifetime in this field. The more time we spend working towards that goal, the more time is getting wasted. I should also point out that Richard Stallman and Linux Torvalds are the exception to the rule. You didn't list the countless folks that started new projects and ideas that completely failed and got forgotten about.
I'm not saying we give up and join "them" where "them" == apple or microsoft or google or whatever. I'm saying we give up and give users what they actually want. On demand, available everywhere on everything content. We can't compete on the freedom values alone. We need a higher quality product. The freedom is just our best way of getting there but not if so many of our high quality programmers are working towards a goal that is irrelevant now.
FOSS is full of lost leaders too stubborn to realize how far removed they are from reality. I can't blame them, I'd hate to spend years working on something only to have to change it because the world is changing. Music, TV, and movie industries know exactly what I'm talking about. We're next. Server software aside, we haven't lead the way on any major software products in a long time. We're just shadows of our former self. We try to copy OSX and fail, our web products aren't anywhere near Google. We've just lost our way somehow despite our growing developer numbers.
Yeah, Richard Stallman and Linus Torvalds are the exceptions... but so are Microsoft, Google, Apple, Adobe and Facebook. Just remember how many commercial products and services have failed... many of them market leaders at one time.
Perhaps you are just saying that in order to continue to grow, FOSS must change. Fine. I'm certain it will. Not all of it.. it will be a splintered flow of liquid.
You say FOSS needs to focus on what users want. No one said it shouldn't. What do users want? I don't want the cloud. I don't want all of my data on other people's computers. I may or may not be in a minority... but the point is that there is a lot of diversity out there and ALL of it will be served... not just the hot new thing.
So, what needs to be done first to make you happier? A FOSS Google Apps clone? A FOSS file-syncing service? What? You really haven't stated any specifics.
The desktop is dead. Long live the desktop.
You (and I for a long time) made one major assumption.... That people care about their privacy and freedom more then they do usability. People like you and I do. Enough to fight for it. I support the EFF because they fight the good fight. But my parents and the majority of the world just don't care.
That assumption is just wrong. The only way us freedom fighters can continue is to build better stuff then what the proprietary people are building. Freedom is a feature, but only a minor one if the functionality is not up to snuff.
My parents will never run their own their own PBX or NAS at home. Ever. All we can do is help level the playing field by providing better software then what places like google or salesforce are providing (salesforce is just an example since they really have been successful).
The same themes tend to play out in the larger world when it comes to civic liberties, etc. ie. people (in western democracies at least) take their freedom for granted and it is for the few (like EFF or ACLU) to remain vigilant for the good of all.
Well, sometimes it seems like that anyway...
Anyway... he knows I'm a computer guy so he asks me a ton of questions. I use Linux. "What is Linux? I haven't heard of that." He assumes that Linux must be some third-party app for Windows.
We didn't really have a chance of winning when only a small percentage of the overall computer user pool knows what Linux is. Application availability and usability don't really matter too much if they don't even know you exist.
Again, I concede that the Linux desktop has lost but I don't necessarily think that that is all that important. I mean, nothing has changed for me. I still use my Linux desktop, the applications still work... and they are still being developed and improved rapidly.
I believe people *DO* care about freedom. It is just that it is a pretty complex notion when applied to computing and many people aren't interested in taking the time to comprehend the situation... assuming that they'll ever be given the opportunity. Just like the neighbor not knowing what Linux is, most folks have no concept of software freedom or lock-in.
Regarding various people being in their own realities and in denial... we all are in one or more ways. You are too with this concept that everything will be a web-app. Just watch the "Death of the Desktop" video. Many people are strongly opposed to putting all of their data on other people's computers from both a personal and/or a business standpoint. Web apps might offer basic functionality but they really can't compete with standalone apps with regards to features. I guess the theory is that HTML5 is going to give them more capabilities and they will be better able to compete. Ok, that may be true to a certain degree... but talking about being way behind. The point is that there are just some applications that will never translate well to web-apps... and small screens. Yes network bandwidth will (fingers crossed) start approaching hard disk bus bandwidth and application size and delivery will be less of an issue.
Let's revisit the topic in 1 year and see how things have changed.
And Android is mainly Linux. Plus Apache licensed software = FOSS.
So, where did we lose? Android is winning big. Qt is winning big (LGPL = FOSS). OpenOffice is winning bit by bit. KDE is innovating where Windows isn’t. Most of googles infrastructure is built on free software ⇒ the cloud is mostly FOSS.
And most importantly: I don’t need a bit of unfree software on my computer to do my daily work. I’m writing this from Konqueror. I listen to music with Amarok. For programming I use Python, which is FOSS, too. gpodder gets my podcasts. I design documents with Scribus. I version all my stuff with Mercurial. If someone sends me a word document, OpenOffice comes to the rescue. My mail gets downloaded with kmail. And if I had a smartphone, it would run Android. And the FOSS voicechat software mumble works far better than skype.
So please tell me: Where did we lose? We achieved the goal of providing an alternative to everything unfree software has to provide except for mainstream entertainment titles, and these will follow, when people work more and more with creativecommons works. Creating games becomes cheaper when you can reuse creative works of others on top of the game engine, just like other software becomes cheaper when you can use FOSS libraries. Just join in and enjoy it :)
At least not in the area I work in (computer graphics, mostly realtime & interactive, that is, games).
FOSS innovates in "easier" areas, where performance considerations - the reason for a great deal of dirty hacks, second only to deadlines - doesn't matter that much.
Programmed using the FOSS versiontracking Mercurial. → http://mercurial.selenic.com
Or Irrlicht? → http://en.wikipedia.org/wiki/Irrlicht_E
Or Linux in servers, where performance is everything?
Mercurial is also a bad example... Try adding 50+ GB of files (which is easily managed by Perforce since the 1990s) to its database and see how it chokes. Even git, which is written in C, cannot handle that (at least stock version, didn't try git-large-files or how it's called). And you need to version your code & binary data together.
To sum up, FOSS projects are under heavy influence of academia, which values clean code over "blood, sweat and met deadlines" attitude typical among commercial software developers (which means a lot of hacks and manual work, but is also necessary for innovation). Since the 1980s this practice was predicted to fail, but as we can see with Microsoft & gamedev, this practice survived and currently scales well to large projects, and what's more important - 95+% of world population prefers buying "not-so-robust" products which give them something new instead of relying on (and endlessly waiting for) "cleaner" FOSS software (well, Linux itself was a "obsolete" dirty hack from academical POV because Linus didn't want to wait for GNU Hurd, mind you).
That’s not desktop, but it’s another win for free software. And on the desktop it looks quite good, too, with for example brasilian schools going for free software laptops. That’s 52 Million systems. Running KDE on GNU/Linux. And further millions sold in retail. And all german embassies around the world run KDE on GNU/Linux:
* http://home.kde.org/~akademy10/slides/R
* http://home.kde.org/~akademy10/videos/R
(oh how I love to reference that talk! :) )
And Mercurial is a pretty good example. If you want to version 50+GB files, you use the bigfiles, snap or bfiles extension:
* http://mercurial.selenic.com/wiki/Bigfi
* http://mercurial.selenic.com/wiki/SnapE
* http://mercurial.selenic.com/wiki/Bfile
You just use the right structure for the task instead of trying to shoo everything into one single structure. Look at Firefox using Mercurial.
And Linus didn’t do and continue doing the “hack”, because he didn’t want to wait for GNU Hurd. He did that partly because Tanenbaum told him in very harsh words that a monolithic kernel could never succeed.
Windows on the other hand uses a Microkernel approach… and lost massively on mobile devices. Which are just desktops in a new shape (and often much more powerful than my first desktop computers). Now there is android. Mostly free software. And it is succeeding quite well – remembering that it is spearheaded by a company with no previous experience at all in mobile devices that success is a marvel.
On the other hand, MacOSX is a success on the free software Mach Microkernel. And the free software WebKit is a huge success. Coming from the free software KDE, by the way.
Look at KDE 4 betas to see not-so-robust new stuff. Look at the 4.x.4 versions to see the new stuff being robust.
Linux was a success, because it was simple and everyone could improve it, not because it was „unclean“.
Guess when HTML5 come out, Microsoft will releases their own set of tags, Google also likely to have theirs, and they are not necessary coexistable. :-)
BTW, I am curious about your estimation of when Google chrome OS overtakes Windows. :-)
Microsoft Office 2010 (http://www.software-hotbuy.com/)
Microsoft word (http://www.software-hotbuy.com/)
Office 2007 (http://www.software-hotbuy.com/)
Microsoft Office (http://www.software-hotbuy.com/)
Microsoft Office 2007 (http://www.software-hotbuy.com/)
Office 2007 key (http://www.software-hotbuy.com/)
Office 2007 download (http://www.software-hotbuy.com/)
Office 2007 Professional (http://www.software-hotbuy.com/)
Outlook 2010 (http://www.software-hotbuy.com/)
Microsoft outlook (http://www.software-hotbuy.com/)
Microsoft outlook 2010 (http://www.software-hotbuy.com/)
Windows 7 (http://www.software-hotbuy.com/windows-7-c-2.html)