Linked by Thom Holwerda on Fri 31st Aug 2007 19:48 UTC, submitted by diegocg
KDE The KDE project has delayed the release of KDE 4.0 by two months. "We, The Release Team, hereby announce that we are extending the KDE 4.0.0 schedule 2 months by inserting an extra 2 Betas, as follows: September 24: Beta3. October 22: Beta4. November 19: Total Release Freeze. November 21: RC1. December 5: RC2. December 20: 4.0.0 tagged. We feel that there are crucial elements of the release that need more development time. The feature freeze (less exemptions) remains in effect."
Thread beginning with comment 267785
To view parent comment, click here.
To read all comments associated with this story, please click here.
quenturi
Member since:
2006-04-10

I've been reading for years here things like: 'app x never crashed on me', 'app x crashed a zillion times on me so now I'm using (here you put whatever desktop you want), 'X desktop is faster and way more stable than desktop Y', 'no it's not'...

OSNews is crippled with that kind of things. I have no doubt people are telling the truth. However it might be time to understand apps, desktop and distros do NOT perform the same way on every computer.

Therefore I don't see the point telling 'my experience counts for something, yours doesn't'. They both weight the same in my humble opinion.

I'm a kde user (kubuntu and kde 3.5.7). I'm more or less satisfied but I do have some crashes here and there from different apps and most of my apps are too slow to start in my opinion (noticably slower than any app on Windows XP for example). My linux system is pretty good (dual core AMD, 2 Gb Dual channel DDR, 6600 GT PCIE, HDD SATA, SATA DVD and DVD Burner). Despite all that, I'm not sure it would be right to say kde is not stable, is slow and crashes all the time. Why ? Because it's my system and I know some other people reported kde works fine on their system. A conclusion that kde is good or bad, slow or fast etc. based on my only experience is pointless.

@ aseigo: please, do not reuse the 3.5 HIG (or a slightly modified version) for 4.x, rework it entirely.

Reply Parent Score: 4

Obscurus Member since:
2006-04-20

I agree. Without doing a detailed stability assessment on controlled hardware, one person's anecdote is as useless as another's.

That said, any large software project that allows end users to make very complex configuration changes is likely to have problems testing and sorting out bugs. In KDE, there are literally billions of possible combinations of configuration settings, across hundreds of programs and subsystems.

Microsoft don't limit the extent to which you can customise Windows because they are mean horrible people, they do it because finding and fixing bugs, and supporting end users, becomes exponentially harder and more expensive with each variable introduced.

The more variables there are to control, the harder it is to identify, troubleshoot and fix bugs.

Is KDE less stable than other DEs? I can't say for sure, as I have no solid data, but KDE is a lot more complex than any other DE, and all other things being equal, that fact alone means it is more likely to be buggy. My own experience is that older, simpler software is generally more stable than newer, more complex software. XFCE in my experience is much more stable than Gnome and especially KDE on the range of machines I have tested them on, I dare say that has something to do with the relative simplicity of XFCE compared to the other two.

Reply Parent Score: 2

smitty Member since:
2005-10-13

The more variables there are to control, the harder it is to identify, troubleshoot and fix bugs.

It's true, the more features software has the harder it is to fully test. However, there are ways to increase stability that KDE does better than any other desktop environment (that i know of) like reusing common code in libraries. For example, every GNOME application that is ssh aware needs to add code to view files over the network, while KDE apps can simply use the built-in fish KIO slave. So a lot of KDE's features actually end up being quite simple since you can assume the framework is fairly well tested and working.

Reply Parent Score: 8