Why it didn't work years, then?
Why it didn't work years, then?
Posted Apr 19, 2011 11:29 UTC (Tue) by dskoll (subscriber, #1630)In reply to: Why it didn't work years, then? by khim
Parent article: Ubuntu reaffirms Unity plan for 11.04
But it's not clear to me why do you think it's impossible for Ubuntu to get it
Apple is a massive consumer electronics company. It sells huge quantities of cool hardware and music. Mac OS X is more-or-less incidental to the hardware sales; Apple has an integrated suite of products that reinforce one another in a "virtuous circle".
Canonical has... Ubuntu, one of dozens of Linux distros. (A popular one to be sure, but nothing that special.)
Because the required ingredient of the "consumer desktop" is plenty of ISVs and plenty of ISVs need uniform platform.
That will never happen in Linux. Its users are simply too ornery and opinionated (and that's a good thing, btw.) If you want to be a Linux ISV as my company is, you have to deal with that fact. It's not the end of the world. (I don't know how old you are, but I've been following the UNIX wars since the early 1980's. You should learn from history that UNIX/Linux users tend to be adamant about what they like and unwilling to be herded along by software developers.)
I don't see the need to introduce third major flavor of Linux (in addition to KDE4 and GNOME3
KDE and GNOME are not "flavors" of Linux. They're basically overgrown window managers. I use neither KDE nor GNOME, yet KDE and GNOME applications run just fine on my desktop (XFCE4). It's not like you need One True Desktop to rule the world to run Linux.
Posted Apr 19, 2011 12:11 UTC (Tue)
by khim (subscriber, #9252)
[Link] (5 responses)
Right. And this is what they are trying to fix. But you need it to get adoption on desktop. Thankfully Microsoft splintered it's OS, so you don't need total uniformity, but as long as you must test your program with different settings to satisfy more then trivial percent of user... Linux will lose. When (and if) ISVs will start closing bugs with "I'm using XFCE4" text as "WONTFIX, please try with Unity instead" we'll be able to say that yes, first step on the road to the world domination on desktop is achieved. Linux won server when it dropped all incompatible versions of kernel, libc, c compiler (gcc 2.96 vs pgcc vs ... shudder): when it become easier to develop for Linux rather then for Unix. Sadly it only happened on the lowest level of stack. Upper levels are still in disarray. Infrastructure becomes more and more standard, but upper echelons of UI are wildly different - and this is extremely bad thing from ISVs POV. I don't really see why duplication of effort and myriads of half-backed offers are better then few solid solutions is "a good thing", but I'm not sure it'll continue forever. Till it continues most ISVs will ignore Linux. Yes, few ISVs will decide to "deal with it", but most will wait. I'm not sure how it'll be fixed: may be Ubuntu will become so popular it'll be "standard-de-facto", may be it'll adopt Dalvik and make it possible to run Android software, may be something else will happen, but I just don't see Ubuntu adopting the philosophy "users are simply too ornery and opinionated - and that's a good thing". From the day 1 when they decided to adopt GNOME and not include KDE they did things this way. Ubuntu was always about mainstream consumer desktop, not about few stray Linux users. The fact that it didn't achieve these goals yet does not mean it should or would abandon it.
Posted Apr 19, 2011 12:54 UTC (Tue)
by dskoll (subscriber, #1630)
[Link] (4 responses)
Thankfully Microsoft splintered it's OS, so you don't need total uniformity, but as long as you must test your program with different settings to satisfy more then trivial percent of user... Linux will lose.
Writing apps that work across a variety of window manager and desktops is a solved problem. It was solved decades ago. But because of the continual reinvention of the wheel by GNOME and KDE and their conscious or unconscious decisions to deprecate decades of X Window and UNIX tradition, it becomes annoying.
And completely revamping the UI serves no useful purpose as far as making it easier to write cross-desktop applications. With Unity, GNOME is diverging even further from other desktops and causing even more confusion and fragmentation.
I don't really see why duplication of effort and myriads of half-backed offers are better then few solid solutions is "a good thing"
"Half baked?" KDE users seem to think that KDE is a fine desktop. I happen to think that XFCE4 is great. And I'm sure many GNOME 2.x users are perfectly happy with it. A huge change of direction in UI is not necessary and not useful. It's simply the result of self-absorbed developers scratching their own itch. [There's nothing wrong with that as far as it goes, but don't pretend it's part of a noble effort to unify the Linux desktop and increase Linux adoption.]
Linux UI developers need to learn a few things: Linux on the desktop is niche, is likely to remain niche, and there's nothing wrong with that. They also need to learn that major changes in UI are very unlikely to attract people away from non-Linux systems, but are quite likely to annoy existing users. And they also need to learn that GNOME, KDE and XFCE alone cannot dictate to others the direction of UI development, so there is absolutely zero chance of a One True Linux UI. Rather, GNOME, KDE, XFCE et al. should work together on incremental improvements, unification of look-and-feel and updating of core desktop conventions so that ISVs can more easily write cross-desktop applications. I know there's already a lot of this work going on. Unity is a step in the wrong direction, creating divergence instead of convergence.
Posted Apr 19, 2011 13:32 UTC (Tue)
by khim (subscriber, #9252)
[Link] (3 responses)
Solved? I think not. If by solved you mean "you can run them... somehow" then yes. If by solved you mean "you can run them... and you can use the same UI in all of them" then no - it was not solved back then and it's not solved now. Even MacOS have not solved it fully, but it comes close... unless you'll run X Window System and the ported zoo - which few people do. Right. But it was never a goal. I'm perfectly fine with Ubuntu Lucid. I'm quite sure I'll be happy with Ubuntu 12.04 too (I don't yet know how it'll be called). I think they finally understood that. The solution? Stop trying to unify Linux. Create separate branch which will have unified look and feel, where all programs will work in the same way and where the whole thing will be "bigger then the sum of parts". Kinda like what Google did with Linux for the mobiles. Ubuntu does not have the resources Google has so it can not to redo everything at once - they are revamping Ubuntu desktop one piece at time instead. I'm not sure I like all their changes but I know I don't like "Linux zoo" even more. Heck, I wanted to abandon Linux for MacOS, but found that it's too difficult to live without real hardware Home/End/PgUp/PgDn keys. I still might if this "Linux on the desktop is niche, is likely to remain niche, and there's nothing wrong with that" hoopla will continue. They are trying to do something for last ten years. The end result is still a mess. Both from API point of view and from UI point of view. This is risky step, true. But if they'll find out (as they hope) that their vision is shared by most Linux users (because most Linux users will be Ubuntu users) then in few years time network effect will drive all other efforts away (they will remain a niche players but few users will care... and most ISVs will not care either) and the problem of Linux desktop UI unification will finally be solved (in practical terms). I'm not sure it'll work - but it's interesting approach nonetheless.
Posted Apr 19, 2011 13:53 UTC (Tue)
by dskoll (subscriber, #1630)
[Link] (2 responses)
Solved? I think not. If by solved you mean "you can run them... somehow" then yes.
That is what I meant.
you can use the same UI in all of them" then no - it was not solved back then and it's not solved now.
As you note, that (non-)problem is not solved on Windows or Mac OS X either.
This is risky step, true. But if they'll find out (as they hope) that their vision is shared by most Linux users (because most Linux users will be Ubuntu users) then in few years time network effect will drive all other efforts away...
I doubt that will happen for a number of reasons. First, while Ubuntu may have a lot of users, it cannot by any stretch of the imagination completely drive the direction of Linux. Secondly, with open-source software, niche systems can survive an awfully long time because there is no economic pressure for them to die. (Witness the new release of FVWM.) Third, and most significantly, do you really think Ubuntu will be happy with Unity after a while? I bet not. They'll revamp it again and again, changing directions time after time. This is classic Corel behaviour and is a recipe for disaster.
Sometimes you have to avoid change just for the sake of change and concentrate on incremental improvements.
Posted Apr 19, 2011 20:44 UTC (Tue)
by khim (subscriber, #9252)
[Link] (1 responses)
It's mostly solved. There are different degrees. I can tolerate different interface in one or two my main tools (I interact with them for a long time every day so even if they are somehow strange and unique - I'll adopt), I can tolerate it if I have no alternative (like if the program I really need is only available in a form with non-standard interface), but in general I prefer a platform which supports unified look-and-feel. I've used Linux on servers for years but only switched to it on desktop with Ubuntu Lucid (the only reason why I've switched to Ubuntu Lucid and not to MacOS is because I really feel lost without hardware Home/End/PgUp/PgDn keys... I still might if Ubuntu will screw up too bad). I think they finally understood that. That's why they decided to abandon their tries to talk with other parties and concentrated on growing the userbase: if Ubuntu will have enough users then the network effect will drive the ISVs to Ubuntu as well. At this point it'll be not important what exactly will happen in "wide, wild Linux world": they can borrow interesting libraries and components from it but otherwise they can just ignore it. Yet the exact same model works just fine for Adobe (Photoshop and Illustrator), Apple (MacOS and iPhone), Microsoft (Windows and Office) and numerous other companies. I think it's question of balance. Well... it looks to me like this is exactly what Ubuntu is doing. First they moved the buttons, then they changed the panel, then they moved menu, after some time they will change something else. All changes are small and incremental. I don't like all of them but I don't see nothing too onerous. Ubuntu is "Linux for Human Beings" - and human beings are not obsessed with minutiae details. They can switch to new layout given few years (witness Office 2003 to 2007 migration), but when every program offers it's own unique interface... they bail out. Yes, I know, there are picky and demanding *nix users - but there are so few of them they can be safely ignored.
Posted Apr 20, 2011 9:12 UTC (Wed)
by spaetz (guest, #32870)
[Link]
Posted Apr 19, 2011 12:37 UTC (Tue)
by anselm (subscriber, #2796)
[Link] (1 responses)
No, but it does help if the most commonly used applications share a »look and feel«, which is where the added value of desktop environments like KDE and GNOME comes in. Microsoft and especially Apple have put very considerable effort into making their systems work as a cohesive whole, and if your plan is to get people to switch from MacOS or Windows to Linux you had better offer them a compelling user experience rather than a hodgepodge of different styles, even though applications from various backgrounds may all run fine on the same screen from a technical point of view.
In that sense it is misleading to refer to GNOME or KDE as »overgrown window managers«, when window management is only a very small part of what they actually do, and when in fact the window managers that come with these very large software packages can straightforwardly be replaced with other implementations.
Posted Apr 19, 2011 12:44 UTC (Tue)
by dskoll (subscriber, #1630)
[Link]
In that sense it is misleading to refer to GNOME or KDE as "overgrown window managers", when window management is only a very small part of what they actually do
Maybe. But as far as I'm concerned, window management is the only useful part. :) I like to get stuff done and have the "desktop environment" stay out of my way.
Well, they are working on it...
Canonical has... Ubuntu, one of dozens of Linux distros. (A popular one to be sure, but nothing that special.)
KDE and GNOME are not "flavors" of Linux. They're basically overgrown window managers. I use neither KDE nor GNOME, yet KDE and GNOME applications run just fine on my desktop (XFCE4). It's not like you need One True Desktop to rule the world to run Linux.
That will never happen in Linux. Its users are simply too ornery and opinionated (and that's a good thing, btw.) If you want to be a Linux ISV as my company is, you have to deal with that fact.
Well, they are working on it...
Well, they are working on it...
Writing apps that work across a variety of window manager and desktops is a solved problem. It was solved decades ago.
And completely revamping the UI serves no useful purpose as far as making it easier to write cross-desktop applications.
And I'm sure many GNOME 2.x users are perfectly happy with it.
And they also need to learn that GNOME, KDE and XFCE alone cannot dictate to others the direction of UI development, so there is absolutely zero chance of a One True Linux UI.
Rather, GNOME, KDE, XFCE et al. should work together on incremental improvements, unification of look-and-feel and updating of core desktop conventions so that ISVs can more easily write cross-desktop applications.
Unity is a step in the wrong direction, creating divergence instead of convergence.
Well, they are working on it...
It's all about your goals
you can use the same UI in all of them" then no - it was not solved back then and it's not solved now.
As you note, that (non-)problem is not solved on Windows or Mac OS X either.First, while Ubuntu may have a lot of users, it cannot by any stretch of the imagination completely drive the direction of Linux.
Third, and most significantly, do you really think Ubuntu will be happy with Unity after a while? I bet not. They'll revamp it again and again, changing directions time after time. This is classic Corel behaviour and is a recipe for disaster.
Sometimes you have to avoid change just for the sake of change and concentrate on incremental improvements.
It's all about your goals
> Yet the exact same model works just fine for Adobe (Photoshop and Illustrator), Apple (MacOS and iPhone), Microsoft (Windows and Office) and numerous other companies. I think it's question of balance.
Market cap and employees:
AAPL $309.92B (2 of 4683)
around 37k employees
MSFT $215.17B (5 of 4683)
> 100k employees
ADBE $16.98B (288 of 4683, 48 of 839 in Application Software)
>9k employees
Canonical (?)
320 employees
Spot the difference :)
Why it didn't work years, then?
It's not like you need One True Desktop to rule the world to run Linux.
Why it didn't work years, then?