Linked by Thom Holwerda on Mon 10th Mar 2008 17:52 UTC, submitted by irbis
KDE Ars takes a look at KDE 4.0.2. "When KDE 4.0 was officially released in January, there were a lot of gaping holes in basic functionality. During the past few months, the codebase has matured considerably, and the environment is steadily approaching the point where it will be sufficiently robust for widespread day-to-day use. Although there are still many features missing, version 4.0.2 - which was released last week - offers an improved user experience. We tested KDE 4.0.2 with the recently released Kubuntu 8.04 alpha 6." In addition, there is a new 'visual changelog' for KDE 4.1.
Permalink for comment 304282
To read all comments associated with this story, please click here.
RE[4]: Summary
by karl on Tue 11th Mar 2008 07:32 UTC in reply to "RE[3]: Summary"
Member since:

Your comment here is spot on. GNOME 2.0 was at once thrilling, looking much nicer with AA fonts and sporting a lean, sleek look, and thoroughly disappointing because only a fraction of *minimal* functionality was present. It took the better part of a year after GNOME 2.0 was released for it to gradually become really usable.

The worst part of this was 2 things 1) the total lack of configurability -prior to the gconf editor one had to use *really* obscure gconf-d commands to change the paltry undiscoverable options, and 2) many of the most useful apps for the GNOME environment only got updated to GTK2 much, much later-leaving us with a nice looking GNOME2 desktop populated with god awful ugly GTK1.2 applications.

The initial move to cairo did in fact slow things down tremendously-yet I do not regret this move at all for it has really, really paid off. Yet even now this shift to cairo is not fully complete today-I do not say this in reference to all the things which could be done with cairo, but in reference to a baseline of functionality which cairo provides-ie printing, being present in all GNOME applications. I look forward to 2.22 to see if this baseline functionality is fully implemented.

I myself am not a great KDE fan. I certainly appreciate and value much of the tech in KDE but have only sporadically used KDE desktop in the last 6 years. Prior to GNOME 2.0 I was either using KDE of enlightenment, although I found GNOME 1.4 with Nautilus promising it was pure mayhem and crashy back then.

I have continued to use a variety of KDE applications-things like K3B which is just the best burning application out there-bar none. I also use Kaffeine and on occasion I use Konqueror-because during most of GNOME 2.X development Nautilus has been doing the 2 steps forward, one step back dance, and GNOME-VFS, which is now being replaced by GIO, was simply unreliable, crash prone, horribly inefficient and otherwise terrible.

I look forward to KDE 4.1. And I do not understand, and have little sympathy for, the continuous harping on how incomplete/not-ready KDE 4.0x is. From what I have seen of KDE 4.0x things are shaping up nicely- the screenshots show off some first class graphics works and QT-4.x is really diving into some very powerful and experimental stuff -actually revolutionary stuff-stuff unlike anything Linux has seen before. Because I don't use KDE regularly I have no problem waiting another year until KDE really polishes off the tremendous advancements they have made. At that point I look forward to trying it out.

One of the revolutionary things QT-4.x is doing is changing how widgets are drawn to the screen. Both KDE(via QT) and GNOME(via Cairo) are pushing Xorg development and responding to changes in Xorg. When people complain about Cairo people forget that the majority of problems with Cairo were primarily due to a lack of Render support in most graphics drivers.

Paradoxically it was KDE guys who introduced the EXA architecture which finally made it easier to add the missing Render acceleration to most FOSS graphics drivers. QT will still be using Render for fonts, but is foregoing Render for the rendering of widgets and is tying QT directly to the drawing primitives provided through EXA(at least to the extent I have been able to parse Asiegos blogs).

Things I am really looking forward to is the initkit work of FDO-ie. using DBus, a shared tech of KDE and GNOME, to replace the aging init system for system administration. If this work takes off, in the not to distant future we will have simple GUI control of the services which constitute our Linux environments.

Also I am keen as to how Pulseaudio( and eventually libsydney) are going to play out with Phonon. Unfortunately Pulseaudio is too tied to Consolekit, which is totally undocumented and exclusively developed by Redhat/Fedora-making it really, really hard to just try out by downloading and doing configure/make/make install. Additionally it is so bleeding edge that you need the latest kernel, and latest libtool requiring bleeding edge GNU toolchain.

Why is this important? so that we can finally start to solve the persistent nightmare which is audio in Linux. Of course I also wish that the OSS/ALSA folks could work together and provide a low-level answer to those things which Pulseaudio, being high-level, cannot do. Dmix is an affront to any human sensibilities and Pulseaudio can shield us from direct encounters of the really, really ugly kind. KDE opted out of directly tackling these issues, leaving GNOME to do the grunt work of solving media issues with Pulseaudio and Gstreamer. I hope that Phonon plays well in this environment.

Reply Parent Score: 6