Username or EmailPassword
My take on this issue is that while KDE4 is somewhat heavier than other environments, the difference in CPU usage or memory consumption does not matter much on most of the hardware people use.
The problem with KDE is that it puts strain on the underlying subsystems and hardware drivers. From my experience, it works fairly well on legacy hardware, where most of the new stuff is disabled anyway, or it requires top-notch drivers supporting all the new and fancy features. The ugly truth is that most of the linux hardware drivers are half baked and require application writers to carefully maneuver their way through supported features.
On my system KDE4.7 is constantly using 100% of one (of four) CPU cores. This of course makes the whole system run like crap but I doubt this is how it is supposed to work (I mean, no one could screw the implementation up so badly, even if he/she tried very hard). More likely, there is some conflict between KDE and some of the subsystems (audio in my case).
Ultimately, though, it is the job of KDE4 programmers and distributors to make their product work well. The users have all rights to complain if it is unusable out of the box and it is not their job to fix it.
You make a good point. If the user experiences that a CPU core is constantly burned while using KDE, then they should be reporting it through KDE. In some sense KDE is to 'blame' in the sense that they stand between the underlying system and the user, and they should protect the user from such things.
Buggy video drivers, that don't work well in certain circumstances? I believe KDE should be adding ugly hacks around this in their system, while at the same time petitioning to get things fixed in the video drivers so the hacks can eventually be removed.
I think that it's the only way to get the users what they need over the long and short term.
Yup, my point exactly.
In an "ideal world" all the users trying KDE4 would have enough knowledge and patience to identify the exact cause of the problem (and perhaps fix it - that's often equivalent to identifying the culprit), go to a valid bug tracker and never ever mention in public that KDE is running slow or is unstable (even if that's exactly what they experience).
But in the real life the users will spend 1~2 hours on "trying new KDE4", hit an issue they can't resolve themselves, and spread the word to anyone daring to ask about how KDE4 is slow and unstable.
But, is it fair? Well, IMHO it's not fair that KDE4 gets all the blame. But they do deserve bad publicity. After all it is their product that wasn't robust enough and their poor design and testing practices that produced to it.
Besides, who cares?, Ultimately, the goal is to make a better DE satisfying needs of more and more users and not to cover problems with excuses.