Slashdot Mirror
A Quick Look At KDE SC 4.5 Beta 1
dmbkiwi writes "The latest in the 4.x series of the KDE Software Compilation is due to be released in early August 2010. With the first beta of this release recently unleashed, I thought I'd download the openSuse packages and see what 4.5's got in store for us."
What’s New? The Beta 1 release announcement lists only 4 major new features, which seems a little underwhelming.
These are:
One of the big upgrades that was scheduled for KDE SC 4.5 was porting the PIM (ie. kmail, korganizer, kaddressbook) applications to the Akonadi framework. Unfortunately, that process won’t be completed in time for 4.5.0, and will be delayed until 4.5.1. This is a little disappointing given that Akonadi has been full of promise for quite some time, with no real user visible outcomes. It would have been nice to see what Akonadi will bring to the party. However, it’s better to wait until all the kinks are ironed out. But unfortunately, it leaves the KDE 4.5 feature cupboard a little bare.
That being said, there are a whole bunch of little improvements that I’ll talk about later on in this article.
For large sets, this will be our guide even unto death, for the LORD will work for each type of data it is applied to...
Biggest problem with KDE is its massive memory usage and poor performance on low-end hardware. It's much worse than GNOME not to mention the actual lightweight DE's.
From what I've seen of the KDE devs, you'd be exactly wrong on that front. New features are always prioritized because they're exciting, while bugfixes get ignored. I don't have the link handy, but awhile back I saw a bug report regarding (iirc) icon opacity, that had stagnated for years. From everything I've seen, the devs aren't as interested in making sure everything works flawlessly as they are in being progressive.
A Quick Look At KDE SC 4.5 Beta 1
Wow, it was so quick I missed it on the way to the Service Temporarily Unavailable page.
The only thing holding me back from upgrading to KDE4 on my primary work computer (from KDE 3.5.10) is that I need an accelerated triple head display. From what I can tell this is just not possible with KDE4, while it is working fine with KDE 3.5.10.
I really think at this point in KDE4 that they need to work on bugfixes, sure new features are exiting, but what's the point if they don't work? There's a reason why I don't use KDE as my main desktop, it just sits next to my gnome/xfce/e17/whatever desktop, and every once in a while I boot in to KDE to play with it.
Feel free to mod me down, just know that unlike some Anonymous Cowards I'm not afraid to express my views as myself.
On a related note, Aaron Seigo had an interesting post on his blog (http://aseigo.blogspot.com/2010/05/i-dont-need-no-stinking-nepomuk-right.html) where he struggled (mostly in vain) to explain to people why akonadi and nepomuk were needed or even useful. A lot of comments were similar to yours... basically, just give us a stable KDE desktop to run apps and stop messing around with whizzbang buggy features and eye-candy.
Service Temporarily Unavailable
The server is temporarily unable to service your request due to maintenance downtime or capacity problems. Please try again later.
New feature: KDE can now be slashdotted.
These posts express my own personal views, not those of my employer
In truth, akonadi and nepomuk are just a waste of system resources. Not only are they not needed, they're buggy as hell. Seems to me the kde devs have gotten lost in minutiae and forgotten that the point of a DE really is to provide a transparent, appealing framework from which to run apps. If it gets in the way or demands you read a lot of documentation, it means you're doing it wrong.
Hell, it was less effort for me to script my own DE functionality around awesome wm than to learn kde4 so I could support my users who want it.
Caveat Utilitor
heya,
You know, I'm curious how many of the people complaining about bugginess and memory issues are running say, Kubuntu?
I'm on Arch Linux, and the KDE 4.x branch has been quite stable for me - the odd crash here and there, e.g. of Konsole, particularly early on, but nothing that really blew up the whole desktop.
And it's performed very well on my desktop, much more snappy/responsive than Gnome.
There's a lot of distributions that have done terribly, half-done jobs of packaging KDE. Kubuntu is a prime examble, seriously it's an absolute joke how terrible they've done. Last I heard, apparently it was because Kubuntu only had a single guy or something? That might just be a rumour, but I seriously think Canonical should just shelve the Kubuntu branch, instead of giving KDE a bad name.
Arch has been stable for me, and openSUSE was quite good for KDE as well. Don't know about other distributions, but I've heard that outside of those two, the rest are pretty much a joke - they just do a bad job of packaging KDE, or adding their own half-done patches, and pushing out low-quality KDE desktops.
Cheers,
Victor
Interesting timing on this story. KDE 4.5 beta 2 was released today.
http://kde.org/announcements/announce-4.5-beta2.php for the official announcement
He is probably just indicating that those with the most issues seem to strangely be coming from the kubuntu camp, fedora, opensuse etc seem to treat kde as more a first class citizen than second.
Then again it could just be typical ubuntu users are more from the newer to linux camp and thus complain more in general.
This was a very enlightening blog post (and comments) to read. It does explain a lot about why KDE4 is the way it is.
That said, what are the options? As far as widget toolkits go, I much prefer Qt - it's miles ahead of Gtk from programmer's perspective, and it's faster as well. But I'm not aware of any DE (not WM, DE - with file manager and so on) written in plain Qt, with no KDE4-style reinvention of the desktop wheel, and useless bells and whistles.
But okay, I can stick to GNOME for the time being, especially since I don't really develop for Linux full-time, and who cares what widgets apps use under the hood? All well and good, except until that relatively recent announcement of "Gnome Shell" to come in 3.0, with those awful screenshots. Oh. My. Fucking. God! It's like GNOME devs looked at the trainwreck that is KDE4, became envious, and devised their own cunning plan to mess up their clean and usable desktop, and overall screw over existing users as much as possible, for the sake of pushing through some brand new bright UI design and usability ideas. I suspect this will go about as good as their "spacial file browser" did in the past, except that one was relatively minor and could be trivially disabled; whereas Shell design has far-reaching implications for entire desktop, and even third-party apps.
I had preventively moved to Xfce for now, which seems to be free from that "reinvent the wheel again, our own special way" disease mentality (so far). It's okay, but I'm still open to alternatives. What other options are there? (again, DEs, not WMs, so please don't suggest OpenBox etc).
It's been forked - latest change was 4 days ago
http://github.com/gustavosbarreto/antico
I know that the vast majority of people don't care about it, but I honestly want the PIM finished, if they are going to integrate akonadi with it, then fine, but finish it already...
Other than that, it was about time to make a big release with mostly bug fixes in it, maybe it's me but I don't find it as unstable or as memory hungry as people are claiming here, it was some versions ago, no argument there, but now it's pretty decent, for me, what is left are mostly annoyances, and I have suffer a lot of them, but I keep the faith, I like the way it's going.
And honestly... once you make that desktop, you may as well as stop all development except for bug fixes, otherwise people will complain that it is turning into the next KDE.
heya,
Hmm, I really hope you know what you're talking about, and aren't just talking out of your rear-end...lol. Have you actually tried to use Kubuntu, then tried a different KDE 4.x distro and compared them?
I've been a KDE fan since the 3.5 days, and a Ubuntu fan from around those days as well. So it was a natural progression to use Kubuntu. I've basically tried every Kubuntu release since 7.04, until around 9.10, when I basically gave up on it. The 8.x branch, from memory, was particularly patchy for me. I've also tried openSUSE for some time, and spent a bit of time with Fedora. Of these, Arch Linux, either with it's stock KDE packages or KDEmod has been the best, and Kubuntu by far has been the worst.
And I'm not sure what you mean by "in the same place", but they most certainly aren't the vanilla KDE packages - part of the whole point of Ubuntu is they add their own patches to the vanilla packages.
Look, if you don't believe me, take a look at these two articles, and in particular the comments on them:
Kubuntu Gets Some Love
http://www.osnews.com/comments/22113
http://www.osnews.com/comments/22348
Comments like this one:
"Kubuntu has always been a little unstable in my experience - even with KDE3, but since the switch to KDE4 it's been nearly unusable. openSUSE and Mandriva, for example, have the resources to work through a lot of the issues that have come up with KDE4, and I think on a whole both have done a great job. Kubuntu seems at a disadvantage, and I can't imagine recommending it to anyone as a good distro to showcase KDE or even linux. It's just too buggy and frustrating. I have a lot of respect for the Kubuntu developers and their efforts, but they have a steep hill to climb."
Or this one:
"That's an appropriate comparison. I have to use Ubuntu at work and it's as if they go out of their way to damage KDE, so that people will get so disgusted with it that they'll switch to Gnome.
The problem with Kubuntu is not that it's being intentionally broken - it's just their shortage of resources. Blaming Kubuntu is stupid, because it's only so much a few guys can do.
It's a shame, really. It would be in KDE upstreams best interest to see that Kubuntu works well, because that's the distro they are going to get the majority of the users from. If you have a company policy specifying "Ubuntu", that's what you are going to use - and install kubuntu-desktop metapackage to get the kde environment."
or this one:
"Your attitude towards Kubuntu might be unfriendly, but it is kind-of deserved imho. Esp if you're a translator I can understand the frustration. I know and respect the two KDE-canonical employees (Aurelien and Riddell) but agree that Ubuntu puts less work in Kubuntu as it's user base would justify."
And there are more. As you can see, my views on Kubuntu's lack of polish is a fairly common one. Perhaps it's improved as of Lucid, I'm not sure. But the one time I did try the Lucid (Beta, mind you) Kubuntu live Cd, it refused to boot (IBM Lenovo X200 Tablet), and when I tried a later RC, it was horrible and broken (particularly plasma).
Cheers,
Victor
From the bug reports, it seems like KDE still can't handle silly things nobody ever uses, like persistent printer settings or SSL certificates. Both of those are regressions from KDE 3.5, and it seems like KDE tries to mimic Mozilla when it comes to usability.
But yeah, we totally need more UI bling. Not like there was work to do.
Fight hunger. Filet a politician and send him to a 3rd world country of your choice.
In truth, akonadi and nepomuk are just a waste of system resources. Not only are they not needed, they're buggy as hell. Seems to me the kde devs have gotten lost in minutiae and forgotten that the point of a DE really is to provide a transparent, appealing framework from which to run apps. If it gets in the way or demands you read a lot of documentation, it means you're doing it wrong.
Hell, it was less effort for me to script my own DE functionality around awesome wm than to learn kde4 so I could support my users who want it.
'In truth, graphics and sound are just a waste of system resources. Not only are they not needed, they're buggy as hell. Seems to me the kde devs have gotten lost in minutiae and forgotten that the point of a shell really is to provide a transparent, appealing framework from which to run apps. If it gets in the way or demands you read a lot of documentation, it means you're doing it wrong.
Hell, it was less effort for me to script my own shell functionality around bash shell than to learn kde4 so I could support my users who want it.'
Should Free Software really be playing a game of catch-up to proprietary software? The two main proprietary OSs are Mac, which defaulted to a GUI in 1984, and Windows which did the same in 1985. KDE and Gnome, the two main Free Software GUIs, came out around 1997 and 1998, 13 years later.
Now, GUIs were new fangled way of interacting in the 80s. What's the new fangled equivalent at the moment? One candidate is Internet-accessible services/databases/RDF/LinkedData/SemanticWeb/etc. At the moment these are almost entirely Web-centric: everyone's solution to interoperability and ubiquitous access seems to be to dump more stuff on the Web. Let's see how that might pan out:
Success: Everything is now done in the browser. The desktop paradigm dies, taking projects like KDE with it, and everything becomes the browser tab.
Failure: It doesn't work out properly. We're left with a mess of incompatible, buggy sites that make trying to get anything done a nightmare. It's all thrown out as a bad idea because it didn't work on the Web. We move back to the desktop, which has none of the networked-database goodness.
Now, what happens if a desktop like KDE integrates this technology into itself? In the success case that the Web takes over, KDE does not die. It becomes even more useful since it can interact with all of these Web equivalents because it's all standard. No more "doesn't work with Linux/BSD/etc.", because everything's on the Web, and KDE is an extension of that Web outside the confines of HTML and the browser.
In the failure case that this Web migration breaks down then not all hope is lost, since this stuff is still available in every other application that doesn't happen to be Web-based.
In reality, of course, there'll be a middle ground. However, you can bet that in the next few releases of OSX and Windows there will be equivalents to Nepomuk turning up. Of course the Windows one will look like a combination of Exchange and the registry, such that small database updates can break the system, and nobody sans Microsoft can interact with it. The OSX one will probably be more standard, stable and useful, but the database will only allow updates from the iTunes Data Store. Do we really have to wait for such crippled systems to cement themselves in place before we realise that we want one too, or do we grab some EU funding now and try to do it right?
It doesn't stop you from using KDE4, or KDE3, or sitting in a console and doing your image editing via Emacs on an XPM file.
That's really not better.
Yes that's all well and fine, but my point is that that particular functionality has no business being an integral part of the DE. Why is kde4 trying to be an OS? Just provide the DE, or at least make it properly modular so that all this extra crap isn't a requirement. It's too much like Apple or MS, they're trying to stuff this notion of "the kde way" down everyone's throats.
I guess it's just one more example of how mainstream Linux has lost sight of the UNIX philosophy.
Caveat Utilitor