There is difference between truck and train, no? Fucking yes. Different transport ways call for different paradigms.
On modern hardware, the network transparent Xorg is inefficient, because no one has supermachine with 1000 thin graphical terminals at home anymore. These "features" are not used, but are required for graphical throughput, resulting in excessive complexity, loss of performance and degraded coding/debugging transparency.
Xorg is in no freaking way is going to be abadoned or broken.
The road map is either to finish Wayland and then implement Wayland over network or rewrite Xorg from scratch specifically to match needs of datacenters with multiple thin clients to "as needed" degree.
They got criticized only because they have listed many Wayland shortcommings that are false statements because they actually do not know Wayland. That was cleared by them.
Which resulted in another criticism vector, they rewrite simply because they want full control over it even though they have not been hindered to change and upstream anything they change, which is always seen as stupid thing to do.
Are we clear now?
That was the year of Linux desktop for me, because I booted into it and never booted back into windows since then. Never.
I also were not talking about "something completely different". So suggestion - how about you shut up?
No, I haven't heard of Linux dependency hell, but I know about DLL hell and newer version WinSUCKS (WinSxS) which grows beyond 100GiB very rapidly due to each application being installed with own dynamic libraries in seperate containers.
Also, you once more point out you should shut the fuck up because amount of updates one gets depends solely on software choice and distribution policy - whether its rapidly updating or not; whether its bleeding edge or stable.
If you join microshaft developer channel, you winblows beta is updated no less frequent as for example, debian sid. But one difference again - one can PIN the packages and prevent any updates without breaking anything.
I am a turd cleaner and you are wiped once again.
http://doc.qt.digia.com/qtcreator/cr...r-engines.html For the more tech-savvy, you always have Emacs that with the CEDET extension turns into a full-fledged IDE.
he fragmented the linux desktop.
He tried to create a copy of windows 95
he tried to emulate microsoft in every way. Including 'people are stupid'.
his ximian posse almost destroyed Suse.
And now, thankfully, he is gone.
Good riddance. Hopefully he just vanishes.
> Icaza gone
No more sabotage. No more U-Boot-Patents. One troll less.
*opens a bottle*
http://benchmarksgame.alioth.debian.org/ and that's actually something I've been curious about, how do Mono and .NET compare speed wise? On the Java issue I do wish that Microsoft would finally finish putting everything under Apache, because that would finally give the killing blow to Java... Oracle throttling it to death is taking far too long.
As far as a GUI builder goes take a look at Qt Creator, and as far as the optimization issue that is true for pretty much any language, and even moreso for a managed language like C# or Java, because part of the whole argument for JITing is that it can optimize for the architecture it's being run on at runtime and improve the optimization over the long term. In terms of debugging Clang gives nice compile time debugging and then Qt Creator hooks into Valgrind for it's debugging stuff, I haven't played enough yet with Visual studio to say whether one debugger is better than the other in that regard, however it is integrated.
It's becoming clear to me I'm discussing with a child. Come back once you've grown up, moved out from your nerdcave and had some real world experience with computers will you?
The Linux desktop have always been and still is a trainwreck. Even the devs are realizing it and they are either abandoning ship or trying to fix things by creating more fragmentation as usual. Too bad UnitedLinux failed, they were on the right path until SCO went insane.
Instead what you need is a company large enough to design an OE from end-to-end and re-use existing pieces where possible (Linux kernel, some common GNU tools, etc.). And, like Microsoft, Apple, and Google, this company has to depend on consumer sales to drive profits. In other words, make a turd and you get $0. (This is the opposite of the GNOME mentality... where they can force their "vision" on the product, no matter how batshit insane it is, because they don't have to answer to customers at all.)
Now unfortunately there aren't too many companies that have the resources to do this. It's kind of like trying to make a car company to go up against Ford and GM -- the latter have years and years of accumulated experience. But what we do know is that the current distros are kind of wallowing around in the same nerdiocrity year after year after year without really making any inroads in terms of marketshare, and what's baffling is that people in these efforts still can't figure out why.
So I completely reject the idea of a "Linux desktop", just as a "Linux mobile" would likewise be a failure. We need a desktop OE that has Linux at its core and that's about it. Whether it be Ubuntu or ChromeOS or Android or whatever, this focus on commonality and having a "community project" is just a waste of time because it doesn't reflect reality. Creating a good desktop OS requires a ton of focus and concerted effort and, yes, hierarchy.
JMO of course.