Is Desktop Linux Becoming Fractured as Open Source Matures?
Until quite recently, the Linux world had, for the most part, only two major desktop environments: GNOME 2 and KDE. Fast forward to the present, however, and there’s an immense litany of different choices, all vying to become the new face of your open source operating system. To me, this shift signals a new paradigm in the world of free software — a turn that could have major consequences throughout the channel. Here’s why.
First, let me clarify what I mean about the choice of desktop environments available until a few years ago. By no means were GNOME 2 and KDE the only options, or the only serious ones; there have long been a huge range of interfaces for Linux beyond GNOME and KDE, many of them very stable and usable.
But until recent years virtually all mainstream distributions shipped with GNOME or KDE by default. Unless you were a power user interested in trying out obscure alternatives, GNOME or KDE was what you got when you decided to install Linux.
Now, however, the field has undeniably changed. KDE has lost the prominence it once held, and GNOME 2 has been deprecated in favor of GNOME Shell. What’s more, a variety of new projects have sprung up, many of them endorsed by major Linux distributions, for creating entirely novel desktop environments. The most prominent examples include Ubuntu’s Unity, Linux Mint’s Cinnamon and MATE, an effort to revive GNOME 2.
Compartmentalizing Open Source?
The open source ecosystem has never been short on different choices, and in that sense the rapid expansion in the number of desktop environments offered to users is unsurprising. Just as there always will be lots of Linux distributions to choose from, so always will there be plenty of desktop environments.
But given the extreme diversity of competing Linux interfaces that have popped up in only a couple of years’ time, and the complete failure of the community so far to coalesce around one or two leading ones, this may represent something more significant than the natural tendency of open source developers to fork projects. It’s a bit early to say definitively, but I wonder if we’re trending toward the fracturing of the world of desktop Linux itself into different poles that will never be as compatible as they once were.
Traditionally, one could run whichever Linux distribution one wanted and still be able to use all of the open source software out there. Most programs were installable on any Linux distribution, and KDE applications could run in GNOME, and vice versa, easily enough.
But with different distributions now clinging to their own individual desktop environments, which in some cases are being developed in-house rather than upstream, cross-distribution compatibility no longer may be such a sure thing. With Unity, GNOME Shell, Cinnamon and MATE diverging in such different directions, there may come a day when an application designed to run in one of those environments won’t work in any of the others.
Such compartmentalization of desktop Linux seems yet more likely given Ubuntu’s plans to adopt the Wayland server and the HUD interface. Don’t expect other distributions or desktop environments to rush to ensure compatibility with those changes. Nor is Canonical likely to make it any easier to install Unity on other distributions, another restriction on cross-distribution compatibility.
Would the open source channel fall apart if desktop Linux distributions grow more distant from one another, in technical as well as political terms? Certainly not, since there’s a lot more to open source than the Linux desktop. But a shift like this could be somewhat of a shock for users accustomed to the interchangeability that traditionally has been almost a given in the open source world.
Cinnamon would allow Gnome apps to run under it ..Cinnamon is just rewrite of Gnome -shell only to suit the linux mint philosophy. And if Unity support GTK+ 3 , then it should run gnome apps ..
ans for Kde, it can run gnome apps natively even in oxgyen theme ..:) but unfortuantely some another changes like systemd, wayland display server are so big, it is hard to saty in line with linux.. hopefully we will be fast to dtching out bugs 🙂
Unity is based on Gnome.
Cinnamon is based on Gnome.
MATE is based on (older) Gnome.
Where’s the issue? They are all just different user experiences running in the same Gnome ecosystem. The only one at all in danger might be MATE (sadly, as I’m a huge Gnome 2 fan).
I don’t think there will be as big a compatibility problem as you might think. As long as they all keep their Gnome base, they should all remain compatible. It would be in the best interest of all the desktops to be able to run the most popular existing software. I think compatibility with existing software will ultimately determine the compatibility of the desktops, should they stray at all away from Gnome.
Just C. Tozzi ranting against Linux, as he always does. In fact, I’m starting to believe MS (or Apple) actually pays him to criticize Linux, because every new rant is just more trivial and more “anti”. Please, stop writing, you’re not good at it.
@ Mez:
If you have a look at it from the outside you’ll probably find that Mr Tozzi is not all that off. Because it’s rather confusing.
However, having 2 or 3 major DE’s is not really a problem – Mike is right – running KDE/Gnome/Qt/Gtk/Pyton apps is not a problem for the various DE’s.
I am basicly repeating myself, but here I go again:
More distros should spesialize in one DE each. Ubuntu should ditch all DE’s other than Unity at earliest opportunity and concentrate 100% on Unity. I think they will – but I i fear that they will keep the other DE’s in their repos.
Debian, OpenSUSE/SLED, Red Hat/Fedora: Feel free to stick to the kitchen sink approach. It will never sell, but their work keeps on being of utmost value to the community.
Every other distro aming at “friendly” desktop:
Ditch the serverstuff – pick one DE, get rid of other DE’s and and turn the selected DE+system into a distro that work properly.
I’m still not a huge fan of Unity (far from it), but Ubuntu (12.x) is actually succeeding in some very important areas wrt consistency and harware recognition (HSDPA/Printers/Monitors).
Dual monitor setup/tools shines compared to any other Linux distro and W7.
IF distros went “one distro one DE” they will be able to optimise the experience in a similar way.
If every single distro keeps aiming at the kitchensink – none will ever succeed. Been there – done that.
Really you have missed the picture. First and most important question are Windows Managers really important to applications?
Shock to most is no. Windows Managers are almost of zero importance to Application developers. What is the key bit to application developers. Its the tool kit libraries. KDE is slowly getting rid of lots of its private KDE only toolkit. GTK and Gnome is also doing the same thing. So as long as you desktop runs QT and QT applications look fine so will the KDE applications in time. Today GTK and QT can use each others theme engine.
Mandrake tried what Ubuntu is trying now. It will end in tears.
Jack specialist in one windows manager solves nothing. We need more focus on unifying the toolkits. To make application development more and more windows manager neutral.
“Ditch the serverstuff” will come back to bite you. File sharing is server stuff, printing by network is server stuff. So a desktop needs some server stuff this is not avoidable. In fact we need more not less server stuff. Particularly server side management of clients like freeipa. More desktop focused server stuff. Also better management interfaces for server stuff the desktop needs.
Systemd brings a lot of good. Biggest thing along for the ride is latenctyd that able to fine tune resource allocations to suit the desktop.
Jack it does not matter how pretty a shell looks. If the code behind it is crap the shell performance will be crap.
Wayland and systemd are both required to deal with historic problems. That have ruined the Linux desktop experience for years.
There has been far too much focus on the shell not enough focus on what the shell is running on. Like toolkit to toolkit compatibility is far more important to consistent look of the desktop than what shell you are running.
Fragmentation of the Windows managers not causing fragmentation of the toolkits is nothing to worry about. Ubuntu instance to stay with upstart instead of migrating to systemd is a bigger issue.
Christopher Tozzi
“Traditionally, one could run whichever Linux distribution one wanted and still be able to use all of the open source software out there. Most programs were installable on any Linux distribution, and KDE applications could run in GNOME, and vice versa, easily enough.”
This is a false statement. Its only been true in recent years. I started using Linux in 1995-1996. Desktop envorniments were that far incompatible that kde applications were tagged with k gnome applications where tagged with a g. Because loading the wrong one could bring the complete house down.
Compatibility is only a fairly new thing. To give you a idea how new this compatibility is up until 2008 kde and gnome had independent and incompatible sound servers. Still 2008 completely independent toolkit theming. Traditionally in Linux with window manager to windows manager compatibility is to be a complete jerk and do stuff that is incompatible with each other.
Since its only a few years into some where near compatibility has existed some of the old jerks who don’t care about compatibility exist.
What jack is saying is let return to the world of jerks. Distributions with many windows managers have seeing the progress on quality of being able to use applications in many window managers. Yes its distrobutions running only one windows manager that has got us into the current day mess Jack. Kubuntu and Ubuntu for example. Kubuntu all kde/qt application Ubuntu all gtk applications because the gtk and qt theme engines were incompatible.
By going single windows manager you don’t have to fix the toolkit issue so when user loads up application from different toolkit it looks complete crap. Every distribution should ship with at least 1 window manager from each major toolkit base. That is a gtk and qt version. Unity matches this. It has a gtk and qt versions.
Lot of the consistency and hardware detection in Ubuntu (12.x) has nothing todo with ubuntu its just more modern versions of the shared backend.
Forget windows managers focus on toolkit to toolkit compatibility.
Embodiment of the jerks is the KDE vs Gnome war. Jack do you really want to return to the days of the KDE vs Gnome war. With massive incompatibility. I want to leave that age behind for good.
It is still struggle ahead. Distributions with many wm might not look to have enough resources to go round. Really this is the idea case. We don’t not want each windows manager reinventing the wheel for printer configuration screen configuration…. If we can avoid it. More unified the GUI simpler it will be to make training manuals for Linux.
Multi windows manager in a distribution forcing the windows manager neutral paths to be taken is the best option.
Hey guys, your DE is pure choice. I use the Awesome window manager and everything runs fine. There are things waaay more important then your DE, like your text editor
Kyle Merfeld exactly. If it a case that you favourite text edit will not run with you favourite DE because it depending on some other DE dependant library that is wrong. The issue behind the kde vs gnome war was this.
Toolkit compatibility the problem that needs fixing. Not the DE we mostly should give to rats about the DE that a person uses. Toolkit compatibility is the only thing of major importance.
Media has a bad problem covering what is in there face. Not the critical issue on the ground. We really do need someone in the media to follow and report on the progress of toolkit compatibility.
Yes the most import bit we leave in the shadows.
I think it’s great. We are spoiled for choice.
It’s like having different types choices for a category of software. For example, has anyone ever complained because there are hundreds of different types of graphics software?
Suppose, everyone had to use only Photoshop. I don’t have to go into detail what this means.
I also love the fact that there are many different distros to choose from. If we use Ubuntu as an example, if we don’t like the Unity interface, we have the option to install another one. In my case, I installed Cinnamon but am actually using Gnome classic with no effects.
The one thing I miss is the dash in Unity as it allows me to call up any program typing the first few letters without having to edit a menu as in Gnome classic. However, I am almost certain that there are installable software that has the same functions as Dash but I haven’t had time to get round to it yet. For the time being, it was only on just one occasion that I wish I had Dash handy. I could, though, log out and log back in after choosing Unity.
I am a teacher and many of my students are adopting Linux on their desktop because they love its lightness (compared to Windows), miserly disk space requirements and, for the skilled ones, the choice to play around with not only different desktop interfaces but also different distros. And, in this respect, they are even more advanced than I am – having tried other distros like Fedora while I haven’t.
The different desktops and the multitude of distros (I was told up to 4000) are the BEAUTY of Linux. Not one – not one of my students have complained about the “confusion” of different interfaces. In fact, they tell me how they are trying this or trying that. For them, it’s all fun and so it is for me too.
For the school’s library, I have now converted all the computers to the Gnome classic desktop from Cinnamon for the simple reason, most of the computers are older and the classic desktop releases about 80MB of ram while there is a noticeable increase in zip and the disappearance of a lot of lag when several windows or programs are opened.
Using system monitor, I notice too that CPU load is lower using Gnome classic. So, we just love it.
Do we have these offerings (like a menu in a fine restaurant so that we can choose our favourite dishes) in the other major operating systems? Probably some but, as far as I know, it’s not a simple task to change interfaces or try to reduce disk space usage, CPU load and ram usage. These are the very reasons why XP is still so alive and kicking after over 21 years because the old hardware is ticking along and user needs do not require the latest rocket technology to do regular tasks. Even then, video, audio and photo editing (including the newest versions of software) are still very possible in rickety-old XP. I do not use Win 7 myself but I have not heard of Win-7-only software from the major developers.
Personally, I believe the Linux desktop will thrive, though not displace Windows or the Mac OS, because of this perceived “fracturing”. In the last year or so, statistics have shown the Linux desktop presence to have grown from 1% to 1.4%. This is a 40% growth in one year – not bad. I’m pretty sure that if a graph was drawn, we would likely see a gradually steepening curve in this growth.
Great statement: “What jack is saying is let return to the world of jerks.” That’s mumbo jumbo. What I’m saying is that the kitchen sink approach is inedible for mainstream computer users because they are far more complex than needed for regular Joe desktop users. They need good desktop distros with good applications. For them Apache still are native americans and java is an exotic island in Indonesia (been there) or coffee.
Kubuntu’s status has been reduced to be in line with Xubuntu and the rest by the way. 12.04 will be the last one with support offered. Mandriva’s misery is organisational.
What’s Gnome Vs KDE wars got to do with it? The battleline is between conservatism and progress. Between Gnome2/KDE3++ and Unity/Gnome3/KDE4. Gnome 2 should die as KDE 3 did.
But that’s not what I’m saying.
What I’m saying is that a few pure distros (I don’t really care which or what DE each are using) will fill a huge empty spot in the Linux desktop sphere. There are piles of mediocre more or less identical distros (the difference is that they do well and suck in different areas) trying to be jack of all trades. And it doesn’t really work for anybody but enthusiasts.
Unifiying toolkits: No objections from me – but that doesn’t really prevent “pure” distros. If you want to wait for it – don’t hold your breath – it is way out there drifting with Major Tom.
Shells and appearance:
Did I mention that? Don’t think so. BUT, it’s pretty obvious that if something looks like crap it wil be perceived as crap. If the “logic” wrt UX and DE sucks it doesn’t really matter what’s beneath – nobody will use it anyway. Apart from those relentlessly trying to convince everybody else that it is fantastic.
Pure distros are easier to maintain (fewer packages, fewer bugs, lower risk of incompatibilities/conflicts and so on), thus MAYBE surplus resources could be directed towards improving the experience for the not-so-enthusiast Linux user.
Wrt ditching serverstuff I do off course mean stuff irrelevant for desktop. Want to run a server? Run a server (virtually) – or choose one of the kitchen sink distros.
Applications for administration of servers? Shure – but that’s what server distros should provide. Seen from desktop it’s just another application. A truly excellent application with truly excellent UX? Never seen one – it will be most welcome.
Wrt applications in general: Where did the workflow go? MIA or DOA?
To me it seems pretty evident that even mentioning non-kitchen-sink distros makes enthusiasts start digging trenches and rolling out razorwire. That’s nothing new.
Amongst computer users in the world perhaps 0,2-0,5% are Linux enthusiasts and 1% are regular users.
I’m after the 98,5% who don’t really care about Linux. The 0,5% who are enthusiasts will continue to do fine even if 5-10 distros deviate from their requirements by delivering pure desktop single DE to the masses. Perhaps the potential success is a threat to the enthusiast?
@Jack: That’s where you’re wrong. From the outside people don’t really care. They may know about Ubuntu, probably without even knowing what a DE is (even if Unity got a lot of media attention). And they barely know at first there’s something else out there.
They don’t know anything about the fragmentation, the choice. It’s only when they’ll start showing some interest, start digging, start actually using a Linux distro, that they’ll come to know about other DEs.
So, this fragmentation is only viewable from the inside, from people who are no more laymen about Linux (and maybe a few IT curious). And we’re not Windows, we don’t want the focus to be only on a few (2) projects and that’s it. Yes, the effort may be more spread out, but hell, that’s a good thing we have that choice, except for people like C. Tozzi who want to change Linux into something else that it is not. People who definitely show they don’t know anything about it.
Furthermore, the current situation will not last. This is just a side effect of the advent of Gnome 3. Add Unity to that and it’s the revolution. People are currently disappointed with both. Good ol’ XFCE and LXDE get a lot more attention, while MATE and Cinammon were recently launched to keep Gnome2 spirit. But Unity and Gnome3 are maturing, erasing progressively their shortcomings and getting more customization options. People are complaining a lot less either because they have migrated to another DE, or because eventually Unity and Gnome 3 have grown into them (or maybe they just decided to adapt).
So Cinammon is doomed to failure and will probably just disappear, and MATE will be supplanted by Gnome3 at somee point, the way KDE3 was by KDE4 (after people at first either switched or ranted).
The landscape will then be quite clear: Unity, Gnome3, KDE4, XFCE, LXDE and residue. We have to wait for the soufflé to fall and the situation to stabilize (as for KDE4), that’s all, and there’s no worry nor any rush.
Jack
“Pure distros are easier to maintain (fewer packages, fewer bugs, lower risk of incompatibilities/conflicts and so on)”
This is exactly where it all goes wrong.
Fewer packages right we will have just one toolkit gtk/qt… Any application that don’t run with that we will not have.
Yes a jerks define of a pure distribution only has one toolkit. So do you want to live only gtk or only qt applications. Kubuntu is a classic example of this. Even if there was a better GTK application it would not ship with Kubuntu for out the box. This is where pure is extremely bad. Causes duplication of effort.
Pure distrobutions fuelled the kde vs gnome war.
What we want is focused distributions not pure. Distributions that will take the best FOSS has to offer no matter the toolkit used by the application. That will push forwards sorting out the mess Pure Distrobutions caused. The distrobutions who take the path will have to face the toolkit nightmares.
Pure with limited set of packages from trying to be pure is not healthy. Its toxic rot we need killed.
Ubuntu studio is a good example of focused distribution. It does not try to be pure. It tries to have everything a audio producer could want no matter the source.
Pure is basically the completely wrong point of view.
Pure where you pick a DE and be pure to that. The complete deck of cards fall down because the next step from only 1 DE is only 1 toolkit because the DE only needs 1 toolkit then the distrobutions starts causing rot everywhere.
Focused distrobutions are fine. Where they pick an end user then match applications to there needs. If you thinking of DE or number of DE as important are you thinking about your end users.
Talking about only having 1 DE is not Focused on what the end users want. More often than not it will be 2 DE that most of your users want for different particular reasons.
@ Mez: Pure distros won’t cause or remove fragmentation. Actually, I dislike the term. Linux is diverse – not really fragmented, except the resistance against new versions.
You are promoting choice while advocating against it. Huge contradiction.
I don’t really care about Cinnamon or Unity or whatever. If it’s good it will spread – if it’s not it will die. Users moving from Ubuntu to Mint because of Unity is indeed interesting though. They move because they are conservative but always want the latest distro. That’s a contradiction. Why didn’t conservative users stick to Ubuntu LTS? If the in fact ARE conservative?
This is not really about soufflé. It’s abut serving stew that includes the starter (fish) and the dessert (ice cream) in the same pot.
@ oiaohm:
Wrong. I’m NOT saying that there should be no GTK packages in a Qt distro or vice versa. (Have a look at Chakra – it’s a pure KDE distro and delivers GTK applications).
On the other hand: Gnome, Unity and KDE has different requirements (e.g GPU driver optimization) and it’s easier to provide the tuning for ONE destop environment than for a quadzillion. Tons of Ubuntu users install Ubuntu and then put the entire Kubuntu on top of it. They believe they are testing KDE. They are not – they are testing the combo – which is not a “fantastic” experience.
You speak on behalf of “we” as in the entire community. I provide my arguments. You provide yours. Not on behalf of the community. Which I’m a part of too.
How on earth could pure distros be a “completely wrong point of view”?? If it’s anything it is ADDED choice.
Ubuntu Studio WAS a very good example. Since they turned into metapackage they have become a BAD example.
Last time I checked they were unable to deliver version 11.10 BECAUSE Ubuntu moved to Unity/Gnome 3. Ubuntu Studio was the best spinoff I’ve ever came across – and one of only very very few that did justify it’s existence.
Again: You are advocating for the enthusiast. I’m talking about what needs to be done to get distros suitable for regular Joe users. Enthusiasts already have their kitchen sink distros and those won’t disappear anywhere soon.
Users in need of 2 DE’s kind of prove my point. It shows that none are good enough. When 1 DE covers your needs – THEN it’s good enough. For you.
I am not talking about removing choice. I am advocating better alternatives to choose from. For the regular Joe computer user that is.
KDE is not fragmented, and now that it is again stable and functional most likely it will not be threathened by Razor and Trinity in the future. Trinity has not gained any traction whatsoever and will remain as a niche platform for nostalgics. Razor is good, but it fulfills a different role than KDE (not unlike Gnome and LXDE, for example).
Gnome, on the other hand, is indeed fragmented, and that’s because of the Gnome Shell fiasco. I don’t like Unity, but I see it succeeding due to Canonical’s commercial backing, while Cinnamon may work if Mint can sustain their efforts in the long haul. Gnome Shell seems destined to fail, though.
Jack
“Have a look at Chakra – it’s a pure KDE distro and delivers GTK applications.”
This would not be called a pure distrobution. In the gnome vs kde war we had what were called pure gnome and pure kde and pure….. Distributions. Problem is as a person tries to make more and more pure you end up at the case of 1 toolkit and tones of terrible basically goal less distributions came out of this. Being only kde or only gnome I don’t call a goal.
Pure substance and other usage of pure will come out from people will think of. Pure in the sense of only 1 item. Pure is a dangerous word. Basically where ever you would put pure put the word only Jack if it still sounds good then pure is safe to use there. Anyone with a chemistry background will be thinking pure in the sense of only. Same with many other science backgrounds.
Read Chakra description. “created by a bunch of people who like the KISS principles”. This is before DE choice. Focused distribution is what it is. Group of users with a particular goal how they want there distribution to be.
2 windows managers there are particular cases were I have seen and its valid. Its really hard to make a really light windows manager and a feature rich one. KDE itself is 2 Windows managers configurations. Netbook and standard.
Lets say one day Chakra found something that might suit there users better than kde or kde has had another run of insanity. If they stick to there core ideal they can change.
11.10 Ubuntustudio did make it.
Basically if you want to target “regular Joe users” you need to research them and basically write a goal or goals of some form to meet those users needs. These goals should not list software. List functionality. Then build a distribution to match this working with upstream to help them match what functionality that group of users want.
This goal stays with the distribution even if the chosen DE with time has to be changed. When you change between DE for some reason you need 2. To allow the current DE users to migrate with least disruption. 2 is always a number of DE that has to be allowed for that case that the DE you are using chooses to go a insane path.
Like a KDE/Gnome distribution. KDE 4.0 lands and its hell Gnome is still good you can migrate to gnome. KDE comes good Gnome goes to hell. This is the exact pattern it followed. If you give your self no wriggle room this will hurt you. Same mistake Ubuntustudio had no backup windows manager plan if Ubuntu went a direction that did not suit there users.
Upstream a distribution don’t have full control of. 1 only normally leaves you open to a world of hurt.
Pure KDE Pure Gnome…. all are very bad goals. Also are very bad picture to have what the distribution is because developers who turn up might start believing that is what the distribution is and start pushing the distribution to be that only at the expense of everything else.
Jack sorry you are newish. KDE vs Gnome wars make people like me very worried when people start using pure. Seen the results.
All the people who started the pure distrobutions that corrupted where like you Jack who did not know that pure is a bad word. It only seams like a good idea until a few years down the track and the people I call jerks have messed with it.
Really pure something describing a distribution should be almost as bad a swearing. We know swearing is bad. We need to think of using pure that way as equally bad.
X distribution Focused on *end users group* with the following goals and requirements. Result of something like that is distrobutions that last the test of time more often than not.
Tiny Core Linux and dam small linux are two with really good descriptions about objective. We want more distributions will clear objectives. Focus is the key.
If you want what I call really bad Vector Linux. One of the key goals “let user decide”. Ok basically lets skip out on deciding anything.
Also you need to watch the FOSDEM Wayland 2012 Qamp;A. Wayland is cat in the pidgins. Windows Manager equal under Wayland does not draw any part of the application window. This includes the title bars. Yes the applications only toolkit is responsible to draw the title bar just like everything else in the window.
Really this is the best way forwards. No Windows Manager drawing anything. This way toolkits will have to learn to play nice with each other. Common theme or die.
Focus on end user does not come from if distribution has 1 DE or 100 DE. Focus comes from the distributions goals. Jack you are thinking pure will make focus it will but not where it should be. Pure kde or Pure Gnome or Pure any other DE risks putting the focus on achieving that instead of the most important person the end user. End Users should be first goals of distributions.
Of course different distrobutions might target different End users this is fine.
Seems like you assume that i want EVERY distro to be “pure”. Not so.
I don’t really care about the words “pure” or “clean” that much, but swearing? Please… Those words makes it easy for people to grasp the essence of the topic and that’s why i use them.
Ubuntu “downgrading” and moving KDE is actually a very good thing. For Ubuntu and for Kubuntu.
But I give you that: Focus is required. Focus on the not-enthusiast-linux-user that is. To get that focus in a distro (or a DE) it is generally a good idea to reduce priority for the stuff that’s irrelevant for that usergroup. Thus, I would very much like to see clean and tidy distros.
FYI: KDE’s objective is a different UX for desktop, netbooks and pads/embedded – using and reusing the same code everywhere (common TOOLKIT). I am waiting for one more – cellphones. One size will never fit all.
Jack form my point of view a pure distributions have mostly been a waste of space. Because at some point the pure point as only causes them to fail to serve their user group.
One size will never fit all. That is the problem with the word pure to a particular groups of people it will always suggest to attempt to be 1 size to fit all. Leading to disaster.
Remember the Dalai Lama has used every single swear word on public tv in the correct context and so was not censored. Just because something is a swear world does not mean it always offensive. It the fact you can expect it to be taken a bad context. Pure used with distribution description fall into this camp. Clean and Tidy distribution don’t cause the only option to be placed in particular groups minds.
Jack using the right words get your point across. Using the wrong words someone will take what you say the wrong way. If you are meaning pure in the sense of clean. Use word Clean.
KDE was also the first to attempt common theming with GTK to GTK applications would blend in. If you look at what KDE does they try to develop common back-end parts everyone can use. Even if you are not using there toolkit. These days. KDE has basically grown up.
In the time of the Gnome vs KDE war. They would serous-ally go out of there way to break each other.
There is also a major issue. We don’t really need more distributions. What we need is a a better way of providing binaries.
There is only so many maintainers of binary packages to go round. If each distribution does there own. There is simply not enough maintainers. Supporting the 4 major blood lines if Linux distrobutions now is stressing limits. Debian, Ubuntu, Redhat and slackware are all unable to provide binaries for every open source Linux application out there.
Does forking off resources from the main really help. Most no. The exception is when some something has a really well targeted user group and can feed advantages back up stream.
If there is anything we need to fight over to be possibly pure its package management. As this would free up more resources. Would allow more in-depth quality control if one binary can be installed everywhere dependably.
You should probably get over the GTK/KDE war and leave Dalai Llama out of it.
Terminology is not the point – it’s just a distortion. Semantics.
Who said ANYTHING about more distros?
Who’s talking about anything that requires more maintainters?
Who wants more forks?
You are looking for problems that doesn’t really exist while disregarding any advantage clean distros would have.
Clean distros face the same problems all distrobutions do. How to build all the applications end users need and want.
As you community around a distribution gets larger the number of applications you need will grow.
Jack this is the hard point. Clean in theory has advantages. The big picture we have hell due to limited resources. There is not enough resources to go around now. So the solution cannot create more fully independent distrobutions. Solution has to reduce the number of fully independent distributions somehow.
Forks off common base does not really work either. I just gave you the reality of the mess we are in. Does working form the existing common bases help us. Not really they are part of the problem. gcc is built like by all of them. Repeat this many times over. This explains where so much of the resources that exists to build applications is going. Time is a item we cannot magically make more of. So a developer maintain a package could be adding new features.
Solution to the problem at hand is not simple. Jack. Its very simple to think you have a solution until you try apply it and find that in the big picture you have disaster that is not getting better.
If you have infinity amount of maintainers we could go out and create many new clean distributions. Not worry about fall out and wait for the the old distrobutions to die. Time moves on that is how the 90’s and the 00’s were.
The biggest issue that is threat to Linux future is distribution maintainer issue. How to make the process more effective and a clean process so the same libraries and applications are not being built many times without need access all distributions.
What good is a desktop if you don’t have the application you need to run because the distribution you are using don’t package it. Jack.
Yes I know the window dressing is pretty and attention grabbing. The stuff supporting the windows dressing is many times more important than the window dressing. Since without the supports it would not be there to dress the environment.
The problems I am referring to do really exist. A distro new starting out don’t notice it at first. As the number of wanted applications grow thing trouble of getting maintainers appear.
Jack there is something important. If you don’t learn from history you are doomed to repeat it. Part of digging our way out of this mess is understanding exactly what went wrong to create it in the first place.
Gnome vs KDE war is the most visible form of what created todays disaster. There is also distrobutions creating there own packaging instead of working on a common standard. Some of it traces to a logic that the strongest and best would win out. This logic that the strongest and best will win does not always happen in nature.
“Who wants more forks?”
Exactly I don’t some how we need to undo the mess the forking has caused.
You have your mindset locked on problems – not solutions.
Meanwhile status quo remains: Linux will never become a viable alternative for oems sell to consumers. For that clean desktop distros with single DE’s is a must.
This discussion leads nowhere.
I’m guessing quite a few people have already pointed this out already. But there still are only two major GUI platforms for Linux. One is QT and the other is GTK+. LXDE and KDE are QT based. Gnome Shell, Unity and Cinnamon are GTK+ based.
More to the point. I’ve never had any major issue getting any Gnome application to run in KDE or vice versa. There is no fragmentation. Only choice.