In the last few days, KDE developers have been actively discussing usability issues for KDE 4. Here are some thoughts, and some more are here. What do you think?
This deap interest in usability shows how mature is KDE. Its community spent a lot of time to be sure they’ll have flexible technologies and a stable environment.
Unfortunately, as many persons think, KDE is slightly copying some of the MS Windows UI. But by discussing KDE’s usability, and thanks to the active and talented community, we will probably see more innovations and creativity in KDE 4.
I agree with Tomasz Torcz (the first comment to the first linked article). That does all sound like what Gnome did for the 2.0 series. And that makes me happy, because KDE is seriously in need of some simplification. The second article is very good as well, but draws just a little bit too much from OS X. I mean, don’t get me wrong, OS X is great, especially when it comes to usability. However, I’m not a big fan of appfolders, and, even if I was, that’s what ROX is for. As for the number of core applications, one must remember that this is linux, OS of choice, where you can pick from a million wms, document viewers, text editors, everything under the sun. Simplicity is good, but you don’t really want to remove ALL choices…
For me personally, even though KDE is packed with more features, and more powerful programs, but I seem to constantly have programs segfault, w/ 3.4 this has happened on at least 3 distros on 2 computers. from actual Kwin crashs to VNC desktop ones, I find stability and smoothness to be the problems with KDE. With all the features they give you I find the desktop more then usable, Yet the approach GNOME seems to be what I’m prefering, it just seems to be simplier and does exactly what I want a desktop to do: Let me get my work done, quickly, easily, comfortably, and without crashing.
What I actually want to see from the desktops is more freedesktop standards support. Apps communicate w/ Dbus, KDE and Gnome apps both being able to minimize to the tray in either(I think its getting good at that), .desktop support, etc. I’m really glad that Gnome 2.12 is finally supporting some standards, and KDE has progressed just as far.
I think KDE4’s goals should be QT4, stability, support of most if not all of freedesktop standards, some cleanups(even though I really like how much they have cleaned up in 3.x), and maybe adding extra features like an intergrated widget program.
I don’t know what you’re using, but I’ve had very few issues with KDE from about 3.3.2 onwards. It has been very stable, I think I’ve had like 2 or 3 Konqueror crashes in 6+ months, that’s damn good. 3rd party applications crashing? That’s not KDE’s issue I hate to tell you. If you want to argue otherwise, I’ll argue that Microsoft Windows is in deep shit then 😉
It is obvious from your post that you’re:
1. A Gnome user
2. Prefer Gnome (for a variety of reasons and personal preferences).
There’s nothing wrong with that, but it usually leads to inconclusive comparisons based on preferences and second hand guessing, rather than factual comments.
As to freedesktop “standards” – most of that’s crap. I’ve used Gnome a bit, and my honest appraisal is that it has serious UI problems. But then, you could say, that being a KDE user, I’m biased (or meet points 1. & 2. outlined above). Usage colours your perception.
As to KDE 4’s goals – QT4 is one of them. Every KDE release is motivated by being stable, as well as useful features etc. Freedesktop standards do nothing for myself, and I’d rather KDE focussed on it’s on UI enhancements, than something from freedesktop.
And the final comment – “cleanup”. I personally disagree. KDE is just fine as is, if anything, Gnome is oversimplified (which creates its own set of problems). You don’t have to use *every* feature, but it’s there if you ever need it. Users will generally use the features that they want, and ignore the rest. Have a look at Microsoft Word – you’ll see a tonne of icons up on the toolbar, how many users actually use all of them? Is MS Word bloated then?
Your arguments lack merit (in my eyes at least). It is simply more Gnomish trolling of a fine product.
This sentence…”Users will generally use the features that they want, and ignore the rest.” should actually read “Users will generally use the features they want, and THE REST WILL GET IN THE WAY.”
Yes in MY opinion I think MS Word is a bit bloated and the default toolbar layout stinks: How many people ever use the show/hide paragraph button?? I won’t go on and on..
I like the post below about Usability themes. That make sense to me…a different layout for different applications..office, home entertainment, home management/control….hmmmm more?
I guess I see getting in the way the same way as a pop-up window does when your about to click a link.
I’ll admit my first ever WM was KDE. I tried to use Gnome. Hated it’s UI nor can I understand how people like that as soon as you click an option it’s applied even tho you’ve not told it to yet.
That was the major reason I won’t use Gnome.
The other reason is lack of options. I like to fully customize. yes I probably don’t use over 50% of the options. nor can I see why I’d ever use them. But it’s nice to have that option. They don’t get in the way. don’t have to scroll to get to my options they are always right there visible. nicely organized in tabs and what not.
Alot more reading to see what option you are clicking but then again how often do you configure something?
it’s not like your always changing a huge number of options and because of that it’s slowing down your productivity.
As for the start of this thread on stablity. I’ve posted this before here and I’ll post it again. The True reason why I’ll never use Gnome again.
I use Gentoo Linux keep it very up to date. One week I decided that I was going to seriously try to use Gnome because I kept hearing how good it was. So I made myself a brand new system. Now because I keep things so up to date the same versions of the same programs where installed only it had a gnome Twist instead of a KDE twist going on.
I booted it GDM looked amazing. I was impressed. Gnome booted Ok All the GTK apps looked infinitely better than they normally did. Was impressed by the Cd icon having the Cd title. the volume increase/decrease was awesome.. Things are looking great I really started to see why I had been told good things. Then the trouble started when I tried to do the amount of things I normally did. Suddenly the response time was lagging my mouse was jumping, program crashed and then computer freezes.. so I had to reset and try again.. after 15 resets I was mad. I reset the computer because of serious crashes on average every 2 minutes. which taking into consideration of boot time that only gives you ~ 1-2 minute of computer use.
The same program versions installed. Tell me this is a kernel or X problem? Also not only did the kernel crash and X lock up multiple times but every gnome app crashed at least once before it locked up.
KDE has never locked up X or crashed the kernel. the programs do occationally crash.
Grant it this was over 2 years ago so I might give it another Go. However kde’s kioslaves have really got me hooked. plus the infectious gnome way of setting options right away before you tell it to.
Simplicity runs parallel with productivity not “lack of features” as it is sometimes suggested.
I read a great article about HIG once and I wish I could remember where it was so I could post the link to give credit. It posed a few simple questions about how to make an interface instantly more useable and quicker to access for the user. One of the points was about icons (number of them, size, etc.) and how they should be few and NOT small. (I laughed out loud when it poked at ‘power’ users loving the ‘many small icons’)
Since the icons should be few there should be a place to configure which icons to display (so we don’t loose precious features or functionality) and of course there should be ‘sane’ defaults for those people who don’t know what they need or care about a billion options until they need them……
The options should pick themselves when someone chooses a usability theme.
One of the points was about icons (number of them, size, etc.) and how they should be few and NOT small. (I laughed out loud when it poked at ‘power’ users loving the ‘many small icons’)
This is really the thing that finally made me move to GNOME. The biggest asthetic problem with KDE is the lack of spacing. There isn’t enough space between anything. Space is as important in an asthetic work as the elements that fill the space. In KDE, the elements are beautiful (especially the icons), but there is no empty space to bring it all together.
Using Gentoo Liunx you are not entitled to an opinion on the stability of the software you run.
If you can prove that you have compiled and configured the application and everything it interacts with in a supported way, you might claim it. But seriously noone has the time to read the proof just to read your opinion on the matter.
Yes, MS Word is a bloated product with a horrendous GUI. Sadly, OpenOffice.Org have largely copied them. Try Lotus Word Pro instead.
OO.org is another group that need to begin thinking abut useability. Bravo to the KDE team for putting this well and truly on the agenda. Though currently a GNOME user (for the very reason of useability) I am eagerly looking forward to what arises from the KDE 4.0 development cycle.
and no 3rd party applications. the VNC program IN KDE, Kwim(the window manager) has crashed on me. It’s happened in 3.4.0 and 3.4.1, I haven’t been able to get 3.4.2, but I will try it out.
I don’t use gnome primary, I actually use Blackbox with a few little apps(bbkeys, torsmo, etc.). And I end up using KDE more often because of its apps. Saying that KDE is on par w/ Gnome in standards, saying its more powerful, more configurable, is “more then usablable enough” and doesn’t need any more usability work that I can see, that has better apps, is by no measure Gnome trolling.
Also GNOME doesn’t support the standards any better, both are still short on quite a few. I just want to see both the gnome and kde desktops more unified and have elements of both work better with each other.
Quote: “First, everything I said was “my expirence”
and no 3rd party applications.”
I got the impression that you were referring to KDE usage with 3rd party applications.
Quote: “VNC program IN KDE”
I don’t use (or need) VNC, which application is it? That said, I don’t think it’s fair to blame KDE if a application is dodgy. That shouldn’t (and in my eyes doesn’t), affect usability etc of KDE as a whole.
Quote: “I don’t use gnome primary, I actually use Blackbox with a few little apps”
Ah ok. Again, I misread your original comment and perhaps overreacted to it. My apologies.
Quote: “And I end up using KDE more often because of its apps.”
Exactly. That’s what usability really is imho.
Quote: “Saying that KDE is on par w/ Gnome in standards, saying its more powerful, more configurable, is “more then usablable enough” and doesn’t need any more usability work that I can see, that has better apps, is by no measure Gnome trolling.”
Again, my apologies. Whenever there’s a KDE article on osnews.com we have the Gnomites trolling and trashing KDE to their hearts content, and it really pisses me off. Then they start their “but it’s not really GPL’d”, or “it doesn’t conform to standards”, or “it looks childish”, or “it’s too bloated” bullshit arguments.
Quote: “Also GNOME doesn’t support the standards any better”
If you listened to the Gnomites you’d think so. I’m no fool, so of course, I know what you’re saying and obviously agree with you.
Quote: “I just want to see both the gnome and kde desktops more unified and have elements of both work better with each other. ”
Interesting comment – in reality you are right. The sooner Linux loses the multiple window managers/desktop environments and just has *one* primary desktop that is pushed by all distros, the better. One package management system would hurt either. Slim down the number of applications out there as well, and motivate developers to join the larger (and generally better) application development models. Include much better backward compatibility for libraries, so that a program from now can run on a library from 3 or 5 years ago, and vice versa. None of these are really desktop environment based issues, but overall usage problems that can, and do interact with the desktop environment to some extent.
Quote: “I don’t think I’m trolling by any stretch of the imagination.”
Sometimes hindsight is a good thing. I misread Seth for being a Gnome troll, and reacted accordingly. Me bad! I still think KDE is a much better desktop environment than Gnome, although I have a good liking for XFCE. If only XFCE used the QT widget, my that would be super sexy!
Some very good points raised. I used KDE for a long time, loving the extreme configurability (I guess I could be described as a poweruser++).
The interoperability is great, the whole desktop is very integrated. However, discovering that isn’t quite as easy.
KDE needs usability work. Hard. Many interfaces are directly unfriendly, which is what the article comparing with OS X basically points out.
Eventually, I got tired of KDE. I had tried a Mac OS X desktop on a friend’s laptop, and I was astonished. Perfect example of great usability and eye-candy, that really gets the job done. In KDE, you spend a lot of time trying to figure out which button does what, and why that button is there at all.
As is pointed out in the first article, customization is something completely different from configuration. Configuration is something you do one time and one time only (unless something breaks).
I was especially amazed at the KDevelop-XCode-comparison! The KDE folks could really gather a lot of inspiration from such examples, especially regarding toolbars. Toolbars are plentiful to the point where it really gets annoying, when using KDevelop and Konqueror.
So: Simplification of the KDE UI is needed. If KDE does not clean up it’s clutter, I doubt it will be a very popular alternative in the future desktop market. Especially now, when the GNOME desktop is beginning to see some serious integration and usability work.
So: Simplification of the KDE UI is needed. If KDE does not clean up it’s clutter, I doubt it will be a very popular alternative in the future desktop market. Especially now, when the GNOME desktop is beginning to see some serious integration and usability work.
The main reason to use KDE is KIOSlaves. Once the Gnome people get their VFS working everywhere KDE will loose a lot of attraction. The problem is that in both the Gnome and the KDE camp developers are trying to make a good gui for Unix. When they really should focus on making a good GUI for people to get work donw.
One example: Ask your secretary how often she visits
/etc, /dev, /boot, /lib, /sbin /usr, /proc. Cbanses are that he/she never goes there. So why not hide these directories to ordinary non sysadmin users. Every time you suggest something like this, even on the usability lists, you get lots and lots of protests, saying that thes place ABSOLUTELY must be visible at all times to everybody. Of course these places might be useful to a sysadmin, but so far the most sysadmins I know usually use these libraries from the command line, not from the GUI.
If there really are things that normal users needs to changed in e.g /etc there should be a proper GUI to do it. Vieong the file structure in the konqueror or Nautilus is not good enough.
Other things that needs to be fixe is how to install software. Why not have some kind of virtual software foldet to wich packages could be dragged to install them, and then just drag them to the to remove them.
Whenever some dependency problem occurs, whatever needed should automagically be downloaded from the net if possible.
Another problem I see is that Gnome and KDE both seam to think that the user always work alone. He is suppposed to put his file in his home directory and should not access anything else. In a modern office you cooperate with your fellow group members and you share and exchange information. Why not create some kind of virtual structure to easy find what other users are prepared to share with you based on their group belonging.
“One example: Ask your secretary how often she visits
/etc, /dev, /boot, /lib, /sbin /usr, /proc. Cbanses are that he/she never goes there.”
So?
Does she visit c:windowssystem32 that much?
I don’t think so.
What’s your point?
The following quote is your point?
Quote:
“So why not hide these directories to ordinary non sysadmin users. Every time you suggest something like this, even on the usability lists, you get lots and lots of protests, saying that thes place ABSOLUTELY must be visible at all times to everybody.”
They ARE, yes ARE hidden !!!
You do not see /boot or /usr at any time in KDE, unless you specifically choose to see them. KDE starts in your home folder, you know /home/your_account
I fail to see your point here.
Quote:
“Whenever some dependency problem occurs, whatever needed should automagically be downloaded from the net if possible.”
Did you try a linux distribution in the last, say… 3 years? Automatic dependency resolving is somewhat standard in any modern package management of modern linux/unix distributions. If not, then you’re dealing with a rubbish distro.
Quote:
“Another problem I see is that Gnome and KDE both seam to think that the user always work alone. He is suppposed to put his file in his home directory and should not access anything else.”
Sorry, but saying this shows that you’ve never ever used KDE, and it even highly contradicts with the first paragraphs of your post.
Quote: “So why not hide these directories to ordinary non sysadmin users.”
Yeah, sure, like Microsoft Windows is (by default) designed to hide file extensions, and your system files, which is a friggen pain in the ass. That’s really great UI. Not. You’ve obviously never worked on a helpdesk before have you (that deals with the public)? As an example, you might have 2 files named lwremove.exe and lwremove.ini. With extensions turned off, they’re like “wtf, I have 2, which one!”. The customer has already started to panic, and then you have to spend extra time explaining to them etc, and getting them to click on the right item by using icon identification (ie. the icon I want you to click on looks like THIS). That’s disgraceful.
You need to edit/remove certain system files to fix a problem? Sure. You tell them to go to c:windowssystem32. “I don’t have it”. So, you have to spend another couple of minutes getting them to unhide it.
Is this the sort of crap you want? Go friggen use Microsoft Windows then. Your argument is completely blown out of the water by the fact that Apple’s Mac OS X doesn’t *hide* any system folders etc. All perfectly visible. Is OS X now badly designed? eh? Come on, make a reply to this one please.
Quote: “If there really are things that normal users needs to changed in e.g /etc”
Most things there is. Try configuring Microsoft Windows in many areas – guess what! There’s no GUI tool either, it’s drop to a dos command line and do it from there. Example – release your IP address. Do I hear you bitching about Microsoft Windows? No. Again, your argument has no validity.
Quote: “Why not have some kind of virtual software foldet to wich packages could be dragged to install them, and then just drag them to the to remove them. ”
uh huh! So you have used Mac OS X then. The problem with this, is that simply removing the application folder from /applications *doesn’t* remove all of the files from /system and /library, etc. Don’t believe me? Do a finder search. Both rpm and dpkg remove packages and config files when told. Oh, and Microsoft Windows is a shocker for this. Ever seen how much crap dll files are left over, or how much shit is left in the registry? Ever tried editing the registry? It’s really user friendly – *NOT*.
Quote: “Whenever some dependency problem occurs, whatever needed should automagically be downloaded from the net if possible.
”
Yeah sure. Debian’s apt-get does a pretty damn good job for me. Much better (and more accurate) than either OS X or Windows. Your argument is?
Quote: “Another problem I see is that Gnome and KDE both seam to think that the user always work alone.”
Um yeah. Sure. That’s called browsing the network. Windows has network neigbhorhood or my network places, OS X has go > connect (and you can browse if you want). Linux isn’t any harder. Simply edit /etc/samba/smb.conf and change your workgroup name (KDE has a gui for this as well I might add in kcontrol). Restart the smbd and nmbd daemons and voila. Open up smb4k (linneigbhorhood is dodgy and doesn’t support samba3 as far as I’m aware, and xsmbrowser is not as good as smb4k as far as I’m concerned). You can happily browse your network and mount shares etc. The whole idea of this is security. It’s the whole idea of a Unix like operating system – security is valued. Not like with Microsoft Windows, where it’s an added-on thought.
Your arguments are nonsensical at the best, idiotic at the worst. And – they don’t necessarily apply to KDE, or even Linux!
In the end, if you don’t like it – either code it yourself, or hire someone to do it for you. In the words of Dr. Denis Leary – “shut the fuck up”.
That’s why themes were invented. Please use themes to death. I would use a developer theme (clutered, lot’s of info). My mom would use kitchen user simple uncluttered interface, same basic functionality though…. make everything configurable into themes and you can please 95% of the population (tha’s Windows not so secret after all no?).
So this — http://www.simplekde.org/ — is the fork in the road that speaks to many of us who feel KDE has become cluttered & bloated. I’m looking forward to it.
Well, this is accomplished by having a well designed ui that has sane defaults instead of no options.
Since most users never touch the default options, it makes sense to create a default that makes a lot of sense. However, since many people like kde’s configurability, let’s not remove all the options. This will please both camps. Wouldn’t KDE be great for the new user if s/he never had to go into the control center?
Also, I think we have to realize that there are two desktops for a reason. There is no reason to be all things for all people. Let gnome take the simplicity is better approach for those that like it. Why all this talk of kde should be like gnome or vice versa? Let them be different and let the people that like the different approaches use the DE of their choosing.
Two great blog posts. I especially liked Matthias Ettrich’s point:
Nobody likes to do window management. Wirth was right, although few people believed him when he designed Oberon. Users seem to go a long way to avoid having to adjust a windows position or size manually. This isn’t just untrained users, this is everybody. You can observe the same patterns on hackers everywhere…
I used to code witha bunch of windows opened up in xterms. They were always exactly 80×24. Why? Because that’s the default size of gnome-terminal! The only time I’d bother resizing is when I needed to see more compiler output at once. Now, I code in a single 1920×1200 window with one copy of VIM open showing three split-views side by side. The productivity benefits of not having to move windows around, resize them, always being able to see exactly what I need, they’re enormous.
Ultimately, I think the way forward is for apps to do more management of their own windows. Like the example I mentioned above — VIM knows better how to handle my code windows than metacity does, so let it do the management. I’d also think some sort of garbage collection for windows would be nice. I tend to end up with a dozen terminal windows or a dozen browser windows open at once. Finding old ones is usually too much trouble, so I tend to just keep opening new ones. Ultimately, the problem stems from the fact that what the user wants is a particular document (be it a webppage, source file, or whatever), but the only thing the window manager knows about is windows. A way for the user to experss what document they want, and for the window manager to bring that document to them would be very useful indeed.
Yes, the print configuration is a usability flaw*. But the OP said “a perfect example of great usability”, not a “great example of perfect usability”. There is a very significant difference between the two statements.
*) For those who’ve never used OS X’s print dialog. It’s got a combobox in the middle with a big list of options (like “layout”, “paper settings”, “printer settings”, etc). Choosing a different item in the combobox changes part of the dialog below it to refect the option. The thing is, its highly unusual to use a combobox for something like that. Usually, you see a set of tabs, or an icon bar or something. The combobox doesn’t even have a label showing what it does. You just have to click on it by accident to find out.
> The die hard KDE user likes it the way it is, klutered, they feel KDE is powerfull
I doubt that they would miss clutter. And usability doesn’t mean to make your desktop less powerfull (or only if your desktop is called GNOME and you need a reasoning for missing functionality). Look at Mac OS X or Pages as example: simple UI but not less powerfull because of that.
If KDE did some of these things, that would go a long way toward making me less offended by how incredibly ugly the whole desktop experience is now. If only KDE could get away from the anti-corporate development licence of QT and naming every application with a prominent K, perhaps we could achieve the nirvana of a single mainstream linux desktop environment someday. If they actually care about usability and not just tagging on every single concievable feature that anyone could imagine, KDE might actually become *gasp* useable. Please go forward with this!
If only KDE could get away from the anti-corporate development licence of QT
Heh … heh … which corporate development did you mean?
Corporate that doesn’t want to pay money for their commercial products?
From what I have seen and experienced, Trolltech’s licensing is bifurcated into two extremes:
1. The pseudo-GPL license for “open source” which prevents you from later dual-licensing your Qt-based products and covers only a subset of Qt. As this license claims code ownership rights — dictating what you can and cannot do with code you write — it is misleading if not an outright lie to use the term “GPL”.
2. The extremely expensive “commercial license” which has many restrictions on what you can do with the Qt source code.
There is no entry-level affordable commercial development license.
So the vast bulk of developers who would create an ecosystem of libraries, components, and widgets for Qt will never be able to.
Qt has fated itself to being the core of KDE and nothing much beyond this. There may be an app or two from time to time that uses Qt but these apps will be written by large companies that can afford to write their complete application stack themselves.
Trolltech has a quality framework in Qt starting with the 4.0 version (the 3.3 version was not high quality). I hope the company decides to lower the price points for Qt on the desktop and focus their revenue generation on the mobile phone companies. Otherwise there will never be any significant future for Qt unless the company is bought by a player with vision.
So the vast bulk of developers who would create an ecosystem of libraries, components, and widgets for Qt will never be able to.
The same vast network of proprietory but freely-available components that exists for GTK+, eh? Who cares about those components if nobody is able to use them?
Do you genuinely believe that a falsehood repeated with some frequency is thereby made true? Or is fraud your game?
The pseudo-GPL license for “open source” which prevents you from later dual-licensing your Qt-based products and covers only a subset of Qt. As this license claims code ownership rights — dictating what you can and cannot do with code you write — it is misleading if not an outright lie to use the term “GPL”.
What can and cannot be done with the Qt code base when licensed under the terms of the GPL is exactly the same as what can be done with every other single bit of code ever released under the GPL. That Qt is alternately available under licenses other than the GPL, proprietary or otherwise, has no bearing whatsoever in determining the what the precise terms of the GPL are when applied to Qt.
And as for Qt Free Edition being a subset of Qt, that is a misleading statement at best. The commercial and free editions of the X11 version of Qt are identical. The free edition of Qt on Windows is basically identical to what is avaible for X11 under both free and commerical licenses (excepting, of course, the necessary platform specific modifications). The commercial version of Qt for Windows also includes some integration features, most notably VS++ integration, which are not available under the free edition of Qt on Windows.
So the question is then what exactly constitutes the code base rightfully deemed Qt? To some extent the question is semantically vague, and thus no definitive answer is possible. But if we hold to the common understanding of the concepts involved, I feel it’s pretty easily demonstrated that you are either a) guilty of faulty thinking, or b) guilty of willful misrepresentation.
Only that code which is common to all versions of QT, roughly that which comprises the free edition, can be regarded as being a necessary condition for the code to properly be deemed as Qt. To hold otherwise would commit you to the view that since the Windows specific features of Qt are not available on either the X11 or the OSX platforms, Qt therefore can be said to not exist on either X11 or OSX. I trust that most people would regard this to be an absurd claim.
So is the code held in common by all variations of Qt, namely that code comprising the free edition, a sufficient condition for deeming it to be Qt? On X11 it certainly is and, the vagaries of natural languages notwithstanding, I think that justifies us in concluding that the code base constituting both the free edition and the X11 commercial edition is the best candidate for being regarded as Qt. Under this view, all other code would grouped under the title “+ platform extensions”, thus meaning that TT offers “Qt” on X11 and “Qt + platform extensions” everywhere else. Thus your claim that Qt Free Edition represents a subset of Qt is rendered false, as “Qt + platform extensions” is by definition then a superset of Qt.
Copying Windows UI is not necesarily a bad thing. Windows does quite a few things right despite what many may claim.
I consider myself a KDE power user. I can tell you I’m not afraid of change, and of course some parts of KDE need a rehaul. What I don’t want and I’m sure many power users don’t want is to lose funcionality just to achieve better usability. That’s what Gnome project has done, and what many of us dislike.
So the challenge lies in cleaning the clutter while keeping most of features and funcionality. Pretty big challenge, but I’m sure KDE guys are more than up to the task
They guy is talking about his own experience, KDE has crashed on my 5 times more than GNOME, its all about personal experience, maybe for you KDE has been stable, for me don’t.
while were are on usability… fonts anyone? please make a better font engine. I cannot stand the small size rendering of TT fonts with FreeType. Can anyone spend some time to actually produce better looking small size fonts and font engines? Again FreeType doesn’t cut it.
I most admit that I become somewhat concerned every time people talk about removing something from KDE.
Like the second article suggesting that Konservation, Kopete and KSirc, should only be one (I suppose he wants to keep Kopete, only) I find this to be wrong, I could understand KSirc and konservation, but Kopete is not to me an IRC-client (I don’t want my IM-client also to be my IRC-client). Also he suggests that playing and ripping CD’s should be done by ones favorite media player, don’t like that either, I don’t want to play CDs in the same application that plays my Oggs, neither do I want it to rip my CDs.
KAppfinder has to stay since not all applications have a desktop-file, and no filing a bug-report is not the solution, maybe it should be filed, but that wouldn´t solve the problem this instance, KAppfinder would and does.
I like the idea about the commen viewer, I think it could already be done since most of the functionality is already there in a kpart.
KGet should _not_ be removed from the application list, I use it often even though, I don’t use konqueror for web.
I’m strongly again removing any configurability from KDE, but I think Control Panel could use an overhaul.
KDE is of cause not perfect.
One thing I’d really like to get rit of is aRts, I’ve a soundcard with a hardware mixer, so I don’t need a soundserver taking ressources.
People talk about making KDE easier for normal users to use, so here is a little story…
I had a friend over the next morning while I sleept, she managed to mount a CD, watch video from it, play games, surf the net and stop all download activity (on purpose) by her self in KDE without wakeing me. She is a prime example of a user who normally need help with everything on her own computer, at home she runs Windows XP.
The only thing she didn’t managed to do was unmount the CD, because she’s used to push the button on the drive, which didn’t work on my computer that day.
They still need to focus on the clutter and mirroring other os’s enough so that companies and institutions would consider switching to such an environment. Once that happens they can talk about something else.
Isn’t it amazing how much more attractice OSX is in those screenshots? A lot of it has to do with the relative simplicity and lack of clutter.
(This is directed to Matthias Ettrich… I don’t want to post my email on QtDevelop, I tried a harmless fake “not@provided”, didn’t work)
Here am I, discussing the Monster with Doktor Frankenstein himself. Ooh, the irony of it all… (a little dramatic moment, drums fading to background, please)
Maybe I didn’t quite understand your real intentions, so pardon me in advance if so.
> Here?s how it should be (and surprisingly this is also how Windows Vista works): Having images in folder results in little button that says ?View slideshow?. You click it, you get a full screen viewer that shows your images. Simply, intuitive, what you want, what you expect, and no window management action required.
Apart from the problem you described (i.e., having to toggle the full-screen option), that’s exactly what I get in Konqueror 3.2.3. I have changed (in Konqueror’s Configuration, File Associations) “image” to use an external viewer (the default:Kuickshow). Now: just enter a folder and click on the first pic, it gets shown at full-screen. Move the mousewheel and voilà , you easily browse all that directory’s pics (no window management here). Is that what you wanted? (Though I recognize there’s no indication of this use, it’s pretty easy to discover this the first time you accidentaly moves the wheel — BTW, if you’re browsing the Internet this will present all the pictures in cache… pretty nifty, too, I might say).
I thought you knew about this. I must be misunderstanding the whole thing… 8-/
Maybe you’re referring to “default configurations” and how the user should get all this without having to change any options… I would agree with your point, if so.
> And while being at it, single click for opening documents is wrong. Initially I thought it was a good idea, too, but after having tried to use it for a while, I changed my mind. Unless I?m the only one who constantly starts applications by accident when I want to move files around, or delete them.
I gave a reasonable amount of consideration to this recently. To avoid opening things by chance, it’s enough to select things not by clicking, but by “drawing a square around selections” — much like one does “square selection” in GIMP. Works like a charm, and it’s a small price to have single-click selection (which is a real saver from RSI). Alas, one can use Ctrl-Click to select, which is even easier but of less intuitive usage. But, then, all this is a matter of taste, I guess… people can get fond of double-clicking, I suppose.
Again, I suppose you’re talking about first-day users… but then, doesn’t KDE ships with the doubleclick setting on?
I tried Kubuntu and had a ton of problems as well. 3 years using Mandrake and never had a single problem with KDE, though. I came to the conclusion that Kubuntu was rushed out a bit quickly (wasn’t it the very first distro with 3.4?)
I don’t know, it seems to me that usability for KDE is like an afterthought. Unlike Gnome which was built around the concept of natural ease of use, this wasn’t a priority in KDE before. You can’t just pin on the usability thing like a feature upgrade, every single little interface decision has to be rethought all over again. Its an overhaul, and naturally there will be resistance to change for the current users.
Plus, it’s questionable wether the usability of power users and usability of people who just need to get work done is reconcilable.
Less klutter, kandy kolors, and applications dubbed by k* would help for one
Also, the kicker is just really fugly.. and large, and the desktop icons etc look like crappy web links etc.
Those are always some of my complaints… plus yeah… too much copying Windows…. and ugly fat distracting widgets (though I’d say GNOME is no saint either with there incredibly dull widgets)
Whats needed is a simpler default interface, but one that leaves all the options available through an “advanced” menu or something. Gnome users always complain about how difficult it is to use KDE, but I actually find it simpler. I don’t like the defaults, and when I want to change something in KDE I look through a bunch of poorly designed interfaces. But that’s better than Gnome, where I have to Google about how to change the app (if it’s even possible), and then edit some obscure config file.
Whats needed is a simpler default interface, but one that leaves all the options available through an “advanced” menu or something. Gnome users always complain about how difficult it is to use KDE, but I actually find it simpler. I don’t like the defaults, and when I want to change something in KDE I look through a bunch of poorly designed interfaces. But that’s better than Gnome, where I have to Google about how to change the app (if it’s even possible), and then edit some obscure config file.
I use KDE full time currently on my PC. But I also have an Apple iBook. I really like the points the guy in the second blog made – reduce the clutter!
I used to really like KDE, but after using OS X I realised that there were some deep flaws in KDE. Things I could do quickly and easily on my ibook, were just painful on Linux/KDE. So much so that as soon as I can get around to it, I’m moving my PC to Ubuntu/GNOME as GNOME really seems to be “getting it” – at least in some departments. GNOME has some annoyances sure (I don’t really like nautilus, no decent CD Burning app, etc), but the cleanliness of it makes it far more pleasant to use.
No app should have more than one toolbar (by default). That is a great rule And trying to use the system with one mouse button! I don’t think it would be possible with KDE – which is a shame, because I’ve seen older people really struggle with multiple mouse buttons.
I hope that KDE4 will move me back into the KDE camp.
Interesting, I had the same experiences and finally gave up on KDE (and that after being a loyal KDE user for more than 3 years). Gnome is quite a bit cleaner and faster and things in most cases just seem to work, similar to OS-X.
I do hope that KDE will really work on a complete redesign, simplify and make it look more professional. In particular eliminate duplicate (sometimes triplicate) applications by default to clean up the menu. Instead of having three separate projects for a particular application, they should really be more forceful and scrap the ones that are not as good and focus on the ones that do work well.
The thing however that in the end really made me switch over was kmail. When they “upgraded” the mailbox format from 3.3 to 3.4, I almost lost thousands of archived emails. Upon checking, this was deemed part of “improved features”, something that for a production systems is just simply unacceptable. This is an example that the quest for more and more features and options in KDE comes often at the cost of compatibility, which for a production system is just unacceptable. I ended up moving to Thunderbird for emails, Firefox for browsing and Gnome as a desktop, a combination that I found to be more appealing and most importantly, more stable.
The thing however that in the end really made me switch over was kmail. When they “upgraded” the mailbox format from 3.3 to 3.4, I almost lost thousands of archived emails.
Ah yes, that happened to me when moving from 3.1 to 3.2. They didn’t change the mailbox format, but they fixed some bugs in it so it more closely matched the spec. So of course, my old buggy kmail had saved messages incorrectly, and the new kmail would not read them. I ended up writing a perl script to go through my emails and correct the errors in them – then I got kmail to re-index my mailbox. Really beyond the abilities of a non-techy user!
I’m not looking forward to trying to move my 15k+ emails to another mail client. I will likely choose thunderbird (can it read maildir format?), because it is cross platform (and I use it at work on Windows).
I read through the horror stories, and I am amazed that there is still some belief that Linux is ready for the mainstream desktop.
Until the focus is put on solving the user’s problems and not making life easy for the lazy developer (who obviously could not be bothered to write a data migration utility), Linux is not going anywhere on the desktop.
Quote: “Until the focus is put on solving the user’s problems and not making life easy for the lazy developer (who obviously could not be bothered to write a data migration utility), Linux is not going anywhere on the desktop.”
Maybe you should get off your own lazy ass and start coding then if you think you can do a better job?
I”m getting damnwell sick and tired of some people just expecting things to be done for them, the way that they like it, and then bitch when they don’t get it and dare put down the developers.
These guys kode (OK, I couldn’t resist the pun) in their own time, out of their own “pay”, for the fun of it, and the purpose of creation. If you want better, either do it yourself, or go and fork out money to hire a “professional” developer to write it for you. Good luck.
People like you are the height of rudeness in my eyes.
I use KDE and I have for some time, I find it to be quite powerful and I love how much it can easily be tweaked to my needs.
That being said, I deffinitly agree that KDE could be slimmed down a bit.
-only include a few games…. not 50.. or however many that is
-include amarok as the default media player and add video playback capabilities to it
-make the default settings less cluttered and slimmed down a bit… (not so slim as gnome though)
I think it would be a good idea to not only change the default settings, but to slightly change control center, maybe have very basic configuration on the surface, then have an advanced button that gives the user the full ability to do whatever he or she wants.
Oh yes, one more thing, I think kio-locate should be integrated into kde by default, instead of typing “locate:” there should be a small Locate bar (like a google search bar) integrated into konqueror.
I think this would be a very nice feature and with profiling support compiled into the kernel.. it is fast at finding files….. almost as fast as I can hit enter.
KDE and Gnome are different. This is good! I used to use Gnome and then rediscovered KDE and discovered how its little features made my life nicer, now I find Gnome constricting and cramped. For me, KDE’s defaults are at least usable whereas Gnomes are not and changing things in Gnome is more complicated (read gconf). I do acknowledge that Gnome has its place and so does KDE, just as politics is ruined by both sides competing for the center, I fear that KDE and Gnome will pick up eachothers vices as well as eachothers strengths. I love how I can sit down at any Linux (or other NIX) workstation and choose a desktop that fits my likes, wether that may be KDE, Gnome or WindowMaker. If you don’t like KDE, stop bitching and visa versa.
I enjoy seeing this take place, espically with the growing amount of “Linux is hard to use thus not ready for grandma” comments. I wonder if KDE has tried a similar study that GNOME did, involving recording the user’s mouse and eye movements. This would judge how much the user searches through non-intuitive buttons and menus.
These topics are often generally useless – you have the KDE users who on the whole prefer to keep the desktop as it is. Then you have the GNOME users who don’t use KDE because of its usability problems and prefer the simplicity of their desktop. THen you have the OSX users who wonder why people use either. Then there’s people like me, who use GNOME even though they prefer some KDE apps (kdevelop, kate) – usability issues aside.
Alright, I’m a Gnome user since GTK2’s release, but before that I was a KDE user. I also agree KDE has too many useless tooltips, menus, options and menubars. I’m in no way suggesting they scrap everything however. I think they should work on a simple default interface, but give users an option to go “into advanced mode” which would provide all the extra functionality. This would appeal to both new and old.
Yes, an advanced mode is a good idea. There is no point in having every twiggle and twoggle at your finger tip when (as someone else pointed out) you don’t change those settings that often.
I also don’t think people like to have their train of thought interrupted when they have to ‘look’ for a particular feature button. This of course is impossible to have implemented perfectly, but I do think discussion can help clean things up. It really isn’t that hard to set a standard, but it is hard to have people agree on one….and stick to it…and not get mad and fork a project…etc.
KDE has a lot going for it, and I think it could really challenge some of the popularity of Gnome if the interface was a bit more sane and didn’t draw a bead of sweat at first glance. I go from Gnome to KDE to XFCE to ICE to whatever and always find myself going back to gnome to do any real work. The main reason??? I keep pushing buttons, clicking on things, hovering over icons, and exploring just to see whats going on. As much as I fing that fun, it just isn’t very productive.
——–I think the key is when you give somebody ‘beginner’ and ‘advanced’ modes, nobody ever selects beginner.———
I agree 111%. For those who desire simplicity, and “usability”(which is very subjective) the DE should come shipped in “beginner” mode. This group like it or not is the majority. For those of us who love the advanced and highly configurable options, it shouldn’t be hard to put into “advanced” mode. Perhaps one or two switches in a text file.(we’re geeks, we can handle that)
That would make it ok for everybody. I’m one of those who loves controlling everything.
“And while being at it, single click for opening documents is wrong. Initially I thought it was a good idea, too, but after having tried to use it for a while, I changed my mind. Unless I’m the only one who constantly starts applications by accident when I want to move files around, or delete them.”
Oh gods, yes. Single-click-open was a major pain in the ass until I managed to dig up and change the setting.
Kubuntu is a very poor KDE distro. I tried it and it’s dreadful, it’s not even a full KDE. There’s a lot of stuff missing. I would not recommend Kubuntu as a good way to experience KDE.
I don’t see any progress with “SimpleKDE” anymore. Until now every distribution packaging of KDE applies more patches to KDE (SUSE, Xandros, …) than SimpleKDE.
> Like the second article suggesting that Konservation, Kopete and KSirc, should only be one
I wonder about him complaining that his installation contains both Konversation and KSirc. Konversation is not part of KDE so he installed it himself. Arguably KSirc is not the best IRC client (anymore) and should not be shipped with KDE, or replaced by Konversation.
> I would not recommend Kubuntu as a good way to experience KDE.
Ack. Kubuntu is and becomes more a heavy modified distribution of KDE. Because of the customizations (and I don’t talk about shipping only a subset of KDE release) you cannot argue about (vanilla) KDE. It also includes, at least in Breezy, much experimental stuff (alternate control center, Katapult, …).
I allready like KDE the way it is now since 3.4* with some perceived sheduling.It’s fast enough and fucntional,and very easy to operate.If i really wanted a simple desktop i would install fluxbox (which i did and use occasionally to keep people quessing at what the heck is installed) and not gnome.So i hope the people of KDE will continue the good job and don’t let them steered to much by public opinion but rather by common sence and technical knowhow as usual.
How about changing the krappy naming convention with something sensible, Komrade? Yes, it’s trivial to change, but its got better ROI than the complicated fixes.
I think that both KDE and GNOME are too complicated for all but the most curious, explorative, and tolerant users, but KDE takes the cake in this sense.
If I see 10 buttons lined up across the top of an application, there is no way that I’m going to be able to infer the actions of these buttons from their icons without a minute or two of tooltip hovering. How intuitive do these button icons need to be in order to be useful? You’d be suprised. Many users have no clue what the justification buttons in MS word do, and I don’t know how the icons themselves could be any clearer.
KDE is loaded with such things. Buttons galore, tabs running down the sides, every feature prominently displayed. KDE applications scream: “look at how advanced I am, you can tell by the amount of menus and buttons and tabs that I can handle anything you dare to throw at me”… if the user can figure out what any of it does.
GNOME is a little bit better. All buttons say what they do underneath them, at least. There is very little configurability, but there is fairly capable theme engine. The interfaces are impressively intuitive and consistent (relative to KDE), but in some places there are glaring holes in functionality. Why isn’t there an option to burn an audio CD if all the items in nautilus CD burner are audio files? What good is a full-blown collaboration suite like Evolution if I can’t do simple things like subscribe to newsgroup?
The default interfaces presented by the DE must be even simpler than GNOME, let alone KDE, but they should be adaptable: there should be preferences for default interface set. For example, I should be able to easily tell Evolution that I don’t want to use the calendar or task list, and those elements should be removed from the default interface. Also, the interface should be designed such that elements are added and removed as they are needed for the usage context. For example, the media player should present a very simple interface when I click on an audio file in my browser. When I then open a video from the file manager, the media player should expand to show the playing audio track and the selected video in a playlist element. Then when the audio track finishes or I click Next, the player should expand to include a video canvas element.
The essence is that the user doesn’t want a swiss army knife or a collection of special-purpose knives. The user (ideally) wants a single knife that automatically becomes the ideal knife for the task at hand without having to find and unfold it. In practice, there should be more than a single application, of course. But some things that can be easily integrated. For example, a unified document viewer. Or a single document editor that can be anything from emacs to kate to quanta to openoffice to kdevelop depending on what kind of document is being edited. The interface complexity is dynamically adjusted to the simplest subset appropriate for the usage context.
It is apparent that application support is critical, but first the DE needs to expose an API that handles the interface customization. The application defines a base (default) interface and then defines supersets that inherit from and adds to simpler interfaces in a hierarchical manner. A major practical limitation is that this scheme requires a unified widget library (which we should have anyway).
We’re going to need some developer buy-in to get where we need to go…
The essence is that the user doesn’t want a swiss army knife or a collection of special-purpose knives. The user (ideally) wants a single knife that automatically becomes the ideal knife for the task at hand without having to find and unfold it.
YES!!! Especially for information browsing and gathering. The ultimate tool would be like the one described above.
Well, I don’t know why I read the whole thing. It was “same ol’, same ol'” things.
* QT’s not free eviiiils — Pu-lease! Can you inform yourself before posting a troll comment?
* Usability mantra — I see there’s a seriouse amount of people who things usability means cripping features. Well, do whatever you want. Don’t be surprised if one day you will be only able to click icon on desktop, and nothing more. That would be simple, no?
* KDE crashes — yes, yes. Windows crashes. Gnome crashes. KDE don’t crash on me very often, same goes for Windows. If something works for most of people, and it dosen’t for you, maybe it’s time to check out your system?
* Stock KDE should have sane defatuls — Why? You know what’s a secret of Ubuntu success? They managed to polish Gnome. It’s up to distro maintainers to set up your KDE or Gnome.
* OMG! Konqueror’s buttons — Hh, the horror! It takes 40 seconds to change it whatever you like. Hint: yes, you can do it with all the application! But insted you prefer to cut this feature, and give us uneditable ,,sane defaults”.
I should point out that Ubuntu’s GNOME is 99.9% the same as vanilla GNOME defaults. They changed the startup splash and made a theme that I (and others) consider to be ugly. In the first version of Ubunty (Warty), they deviated from vanilla GNOME 2.8 by introducing a Computer menu on the panel. This was a great improvement and inspired the new defaults for GNOME 2.10, which split out the original Computer menu into Places and Desktop. Ubuntu Hoary (current production release) follows the same menu structure except it changes the title of the Desktop menu to System. The only other change I’m aware of is the location of the trash to the right of the pager.
The secret of Ubuntu’s success, therefore, is not that they polished GNOME, but that they created the first free Debian-based distribution to feature a modern desktop environment. It could also be that people really like GNOME, but can’t stand Red Hat…
there appear to be two opposing views from the two articles, both of which have merit, but leave one confused as to which direction to pursue:
1) the anti swiss-army knife viewpoint where it is better to have several really good tools that do several jobs really well, than one mediocre tool that does several jobs in a mediocre fashion.
2) get rid of the klutter, why on earth do we have a thousand different apps for only fractionally different tasks? surely a little amalgamation will make our lives infinitely simpler?
can we ever reconcile the differences outlined above. and if so, can it be done piecemeal, or will it require a new design philosphy to emerger from KDE?
Another usability discussion about KDE,and another bunch of clueless GNOME losers having a flamefest.. how “productive”. Pray tell, you “productive” GNOME-losers, why don’t you do something “productive” in your preferred environment rather than whine endlessly about KDE’s “usability issues”?
Yes, KDE has issues, but most things mentioned here is mostly superficial bullshit issues brought up by gnome lusers who haven’t used KDE for more than max 5 minutes, and then decided that they didn’t like it.
Here is the kmix, kopete (on line in irc, msn and icq), konsole and konqueror running, and as you can see in the dump, it’s using ~ 90MB RAM. I doubt you’d get much lower with the equivalents for that in GNOME..
The second most popular thing to whine about is “klutter”. Well, please point it out for me in my snapshot, or stop trolling. The *defaults* are a totally over the top, but seriously, you are really a dumb fuck if you throw KDE into the trash because of the defaults.. Besides I have a faint recollection of that the defaults were chosen with the intention that the distributors should clean up the mess before presenting it to the end-user. Unfortunately few of them does.
So, what does bother me then?
The default settings for the toolbars are too much and takes some work to clean up, and I haven’t found a good way to separate the main toolbars for the “webbrowsing” respectively the “file-managing” profiles for konqueror, which is a bit annoying since you don’t necessarily need the same stuff in those two profiles, even if some of it are common.
The second beef I have with KDE and it’s apps are bugs, some of them quite old it seems. Kopete seems to be particularly troublesome. Off the top of my head I can think of at least three things that IMO are broken. In no particularly order: Fake transparency for the chatwindow is broken, and apparently has been since 0.10.2 (I use 0.10.3), if you set “base font color” to something nonstandard it doesn’t affect what you are writing, only what others write, and finally it appears to crash on exit from time to time. The latest kmail seems to have quite a few bugs being fixed right now, and I have one system here where disabling icons on the desktop makes the right and middle click menus on it stop working after being logged in for a while. These things are more annoying IMO since they aren’t fixable for a non-programmer user.
Now, all this are just my experiences/opinions, but summing it up IMO the only thing of the frequent complaints that carries any weight is the default “klutteredness”, but hey, it’s not like those toolbars are carved in stone. If you don’t like them, do something about them, use gnome, or stick to a terminal, it’s your choice, but stop trolling about the “klutter”, since as I’ve shown you can easily strip it out.
Oh, I forgot to mention that the issue with KDE “bloatedness” has nothing to do with memory footprint. It’s the UI bloat that’s disgusting.
And the number one problem with your position is that good defaults do matter. Most users never change their defaults. How many Windows XP machines still have that wavy sand wallpaper after years of use? Nobody likes configuring things. Some computer enthusiasts have a fetish for configuring things, and they like to spend time removing buttons and toolbars from their various applications, especially if the weather is nice and there are women around.
There is nothing more valuable in software design than good, clean, sensible defaults. But it seems like KDE’s defaults are made to satisfy some power users at the expense of everyone else. I suppose you think the unique use of bright pink titlebars in the default configuration for FVWM2 is a good thing in the long run? After all, you can always edit the hex codes in .fvwm2rc, and then you’ll suddenly feel the irresistable urge to become a guru and write a 1474-line config file like this one:
And the number one problem with your position is that good defaults do matter. Most users never change their defaults. How many Windows XP machines still have that wavy sand wallpaper after years of use? Nobody likes configuring things. Some computer enthusiasts have a fetish for configuring things, and they like to spend time removing buttons and toolbars from their various applications, especially if the weather is nice and there are women around.
There is nothing more valuable in software design than good, clean, sensible defaults.
Did you read what I wrote? I agreed that defaults matter, and that it takes a bit too long to configure it to some thing more useable. However, defaults are not *all* that matters, especially when they are as easy to change as in KDE.. Not like GNOME where you have to google all over the net if you are dissatisfied with the defaults, if they even are possible to change at all. *That’s* a real sinker. Furthermore I also wrote recollection of that the defaults were chosen with the intention that the distributors should clean up the mess before presenting it to the end-user. To put it more blunutly, if you are dissatisfied with the apperance, blame your distibutor. After all, these these popular GNOME based distributions certainly has applied a fair bit of spit ‘n polish to it, and there really is no law against them doing the same for KDE. Defaults are important, but being put off by them, and trashing the entire thing just because of them is just plain stupid.
It’s nothing advanced in gconf. It’s a silly unusable registry-like crap. Not intuitive, not easily accessible, not professional, not anything. Just a patch on a rather big hole. And quite a lousy patch that is. I’d rather take a text editor over gconf. Hell, I’d even take the “cluttered” and “bloated” kde control center any day over gconf.
And it’s also a lousy day when you consider that if you take gconf as the “advanced” mode, then what are you left with in “beginner” mode ? Right.
You’ve never browsed the .gconf directory. If you want you can edit the files in there with a text editor.
Again: The only thing registry like about gconf is the default UI used to edit it. If you don’t like that, there are other tools that expose more than the default UIs, but less than gconf.
You’ve never browsed the .gconf directory. If you want you can edit the files in there with a text editor.
Have you actually opend one of the files in the .gconf directory and tried to edit it? The files are in XML, which is no way very readable and actually making editing in a text editor even worse than with gconf.
Ah yes, that happened to me when moving from 3.1 to 3.2. They didn’t change the mailbox format, but they fixed some bugs in it so it more closely matched the spec
Very strange, never did that to me and my oldest mails on the system are from the 2.2 days (I lost my older mails due to a disk crash. I still have them on a backup CD but never felt the need to copy them back on hd =).
Are you sure that neither of you (a) could import the old mails into the new kmail and/or (b) used a KDE beta instead of a final. Beta versions of Kmail have been known to lose mails.
Very strange, never did that to me and my oldest mails on the system are from the 2.2 days (I lost my older mails due to a disk crash. I still have them on a backup CD but never felt the need to copy them back on hd =).
Are you sure that neither of you (a) could import the old mails into the new kmail and/or (b) used a KDE beta instead of a final. Beta versions of Kmail have been known to lose mails.
It happened when I moved from Mandrake 9.2 to 10.0. This was just my own discovery. I installed 10.0 onto a spare partition to test it, before putting it on my “live” partition. I copied my mail over, and kmail would read the indexes okay and display items in the inbox (subject, sender, ..), etc. But if I clicked on an email, the sender would change to “unknown”, and the subject and the message would be blank.
I traced this myself to some of my emails not conforming to the maildir spec. Since I have only ever used kmail, I assumed that somewhere along the line (i’ve been using kmail since KDE 2.2.2) kmail stuffed up and wrote the emails out wrong. Anyway, on KDE 3.1, kmail would read these broken emails fine. On KDE 3.2, it wouldn’t.
It happened when I moved from Mandrake 9.2 to 10.0.
How would you interpret that?
The interpretion of that is rather easy. It’s not a KMail issue at all, it’s a well known Mandrake issue. When the maildir spec was finalized and KMail got support for it implemented, Mandrake choose to patch Kmail to not to follow it. They choose to use .mail as directory for mail rather than Mail(or mail) as the spec says(In my opinion much more correct, hiding the maildatabase rather than polute the home dir). Sadly the patches from Mandrake did create some migration issues not present in the unmodified KMail.
Now, unlike the other reply I don’t really know about the specific circumstances of this case but I’ve used Mandrake for a year or so and I have to agree that you should be careful before you blame problems like yours on the developers of the software. I liked Mandrake for many reasons but it’s tendency to lose data due to various Mandrake-specific patches were alarming.
I’m not saying that it couldn’t have been KDE’s fault just that that’s not necessarily the case =)
well.. you may be right.. perhaps set advanced mode as a command line argument, so… instead of startx, startx -advanced or kdm -advanced
anyway… you get the idea… whatever the command would be, this would make it so you have to boot to the command line and have a little knowlege of the command line to get advanced mode in kde
besides….. i think most geeks boot to CLI anyway instead directly into a gui… i know i do.
I have to agree that some apps have way too many functions in their toolbars: Konqueror is a good example. If Konqueror would cut a few rarely clicked things like “print,” “font adjustment(2),” and “security,” and the useless “up;” then maybe one could actually get the address and search bar on that same line! That gives more browsing space.
But on Quanta I think there should be an exception. I don’t use html editors, but I do some hand html. And if you want a visual way to edit html (not WYSIWYG) then you will need quick access to all the common tags.
That means work area includes many button functions.
You can get both toolbars in Konqueror on the same line, even with larger icons. Usually I just keep the Back, Forward, Up, Reload and Stop buttons, and then it fits. However, there are a few problems:
1. The “Up” button is useful in file manager mode, but rather useless when browsing the web. You can’t configure Konqueror so that it is only displayed when it is useful.
2. Whenever an URL gets too long, the address toolbar is displayed below the main toolbar instead of staying at the same place. That’s really annoying.
So I would suggest that to make KDE ultimately user-friendly (since almost everyone complains only about Konqueror), these three things should be done:
1. Remove all buttons except said ones in the default mode
How do you rate rarely clicked things correctly? Your list are not correct according to my user habits. As I use both “font adjustment(2)” and “up” several times each day. Me and a great many others sees the one you labled as useless, the up button, as one of the features making Konqueror great. For me it’s the second most missed feature when I have to use inferior browsers.
I wrote a tool (in kdenonbeta) that lets you record a log of which interface elements are in use. If enough people want to run it in their konqueror for a while, we could actually get some real-world data on this. There’s some info about it in an old blog post I wrote at the time:
what are your problems with gconf? every application has its section with all its settings and every setting has a description and a (localized) explanation of what it does and of which are the accepted values. there are .conf files really incomprehensible, gconf settings are almost always clear.
How much do you simply things so it would have an elegant design? For example nautilus doesn’t even have an invert selection. Nor something I guess considered somewhat advanced as a two pane browser mode. Does os X’s browser have two pane or is it considered too advanced an option that gets in the way of simplified design?
Then there are the non obvious things that make for a smoother experience. For ex. in konq, previews for pics only show in icon mode, not detailed mode, whereas they always show in naut even if all the icons are so tiny that the preview becomes useless and just adds overhead. However naut does remember if you set a particular folder (say for images) to icon mode whereas you might normally browse files in details mode. Btw, is there a way currently to get kde to preview video files?
I know this focuses on point 10 of the article which can be hotly debated. But I don’t mind for the most part kde’s approach. They just have to go all the way with some things. For example I wanted my bookmark toolbar folder to link to locations on my drive, but in web browsing mode show my frequently visted web sites. This should be saved separately with the profiles and not be shared as it is now.
talking about usability and configurability people who defend the “101 feature approach” call them self “power user”.
But what is a “power user”? For me a power user is a person who do more with his computer than the normal user which just surf the internet, write some mail, etc.
But doing more means for me, having a desktop that just works is much more important than for the normal user. The normal user maybe want to tweak the style of the system between surfing the internet and writing a letter, just because he doesn’t have to do much exciting things and if there is a periode were he don’t know what do do -> he will tweak the desktop and “play” with the system.
But a “power user” has so much exciting things to do, why should he waste his time with tweaking the desktop. I think a “power user” will prefere a simple desktop who doesn’t get in the way and let them concentrate on the exciting tasks.
I don’t see that a “power user” exploring all the day 101 buttons of a system and tweak it to death, that’s so boring and just something like “occupational therapy”. I think a “power user” is a user who make exciting things with his computer which the normal user does not.
It’s all fairly academic anyway. I love this though:
The component should have at least one compliant implementation available under an Open Source license that also promotes a “No Strings Attached” environment for developers. This means that the developer would be able to develop and deploy their software however they choose using at least one standard implementation. This is interpreted to mean that at least one implementation is available under a license that meets the Open Source Definition but is does not prohibit propriatry usage. The rationale for this criteria is very similar to that of the LGPL.
Note the this is interpreted to mean… bit bunged on at the end and how it doesn’t actually mean the first at all, but they’ve cocked it up. It talks about one implementation, which Qt quite clearly is because it’s dual-licensed. You get one implementation under mutliple licenses to allow you to do what you want with it. It’s very open to interpretation though, which many people seem to be doing a lot of.
If the LSB wants to create a commandment that says “Thou shalt not pay anything” then just spell it out, OK? That’s what it’s really about, but maybe that wouldn’t look right.
I just wonder what would happen if an ISV created a popular piece of proprietary software using GTK and other ‘LSB’ software and it became a de facto standard and everyone used it ;-).
What is confusing to me is that Qt is not all the way GPL.
If you start developing using the GPL version of Qt, you cannot then go and dual-license your application.
So by using the GPL edition of Qt, you are effectively signing away some of your code ownership rights to Trolltech. I say Trolltech because Trolltech is the only company that could give you an exception to this licensing term.
If Trolltech allowed developers to start with the GPL and then migrate to the commercial license when they wanted to produce closed-source apps, I think it would be an overall benefit to Trolltech for economic as well as legal reasons (reduce problems with Qt/TT licensing).
The other thing that is unclear is the role of “Qt Solutions”. These are important pieces of functionality, especially for using Qt for Windows development. As far as I know, “Qt Solutions” are not available except via the commercial license.
So we have Qt/TT’s GPL license as something that contains extra restrictions and covers a subset of the product.
What I don’t see is why Qt/TT would not go all the way GPL instead of part way. It seems they are most of the way there.
When you look at these issues, what are your thoughts?
What is confusing to me is that Qt is not all the way GPL.
Yes, it most certainly is.
If you start developing using the GPL version of Qt, you cannot then go and dual-license your application.
And if I start developing using the GPL licensed SomeRandomLib.so, I cannot just then go and dual-license my application either.
So by using the GPL edition of Qt, you are effectively signing away some of your code ownership rights to Trolltech.
Insofar as your code is derivative of somebody else’s code, it’s not entirely yours to do with however you please. Copyright law makes it thus, not the GPL, and certainly not Trolltech.
I say Trolltech because Trolltech is the only company that could give you an exception to this licensing term.
I say the copyright holder because the copyright holder is the only [party] that could give you an exception to this licensing term.
The other thing that is unclear is the role of “Qt Solutions”. These are important pieces of functionality, especially for using Qt for Windows development. As far as I know, “Qt Solutions” are not available except via the commercial license.
How dare Trolltech offer one product under the GPL while not offering another in the same exact fashion. Guess you’ll have to buy it then, eh?
So we have Qt/TT’s GPL license as something that contains extra restrictions
False
covers a subset of the product.
Which product? Perhaps the one Trolltech choose to release under the GPL?
If you are working in a commercial environment and are required to keep the source code of your product closed, you must use the commercial version instead. Trolltech’s commercial license terms do not allow you to start developing proprietary software using the Open Source edition.
and this:
If you are writing commercial/proprietary software, or if you plan to commercialize the software you are writing with Qt, you must use a commercial version of Qt.Please refer to the Open Source Downloads info page for reasons why.
From what I can see, Trolltech’s licensing will not allow you to start developing dual-license code with the Open Source Edition. You must use the commercial edition from the beginning otherwise you need some sort of special privilege from Trolltech.
Thus Trolltech’s “Open Source Edition” contains licensing terms that are beyond the scope of the GPL and are actually copyright/code ownership rights.
Also as the Open Source Edition does not support Visual Studio, it is not the same product as the Commercial Edition. It is all rather bizarre.
I’m a long time OS9 user, also KDE and lately Gnome (for the aesthetics mostly). I’m starting to think seriously about going to WindowMaker. Don’t really need a task bar, a right click menu for launching would take less space and do just as well. Don’t really need the enormous memory consumption. And in getting ready for this I’ve discovered something for my less technical friends – all they really want is the same thing, maybe iconised to launch apps. The problem with KDE is it assumes you want to spend your time working on configuring your desktop environment. Most people don’t. The problem with both OSX and XP is that they think everyone should have the same desktop environment. Why? They don’t wear the same clothes. For me, I think its going to be WM: quick, simple, and out of the way of the applications. Well, I have also read good things about Ion, which may be WM but more so…
Trolltech’s licensing will not allow you to start developing dual-license code with the Open Source Edition.
If your code are dual licensed, it’s no longer purely GPL and you can not use a GPL’ed library like the Open Source Edition. It’s the way the GPL works.
Also as the Open Source Edition does not support Visual Studio, it is not the same product as the Commercial Edition.
Wrong! It’s exactly the same, expect it does not contain the makefiles(or whatever they are called in VS) for Visual Studio. It’s no problem using VS, but you have to do some work yourself. And if you run into problems you are on your own and get no help or support from TT. Besides there are bound to be licensing issues when using a GPL incompatible compiler like the on in VS. How compatible are the license of it’s standard C++ libs with the GPL.
Besides ask RMS what kind of tools you should prefer to use when developing Free SW.
Besides ask RMS what kind of tools you should prefer to use when developing Free SW.
Ask Richard M. Stalin what tools I should use to support the glorious revolution?
Haven’t we been through this before?
The GPL is just the classical ideology bait and switch. Sell it to the workers, get the workers to make it happen, and then betray the workers and thrown them in the pit forever.
In the world of Linux software, the giant mega corps already make the vast bulk of the profits. The writing is already on the wall but the workers are too dumbed down to notice. They will give their heart and soul for “one last chance” only to find that yes, they can still be sold farther down the river.
Yes KDE is in severe use of UI cleanup … after a while of using KDE you get used to the messyness … but compare it to GNOME ,OSX or even Windows Desktop .. KDE is simply a mess in comparision .
What has to happen in my opinion is UI cleanup & simplification ala GNOME (like hiding menu-entries & icons when they can not be used) but for Option-dialogs /Controlcenter to have an advanced tab .
A few basic simple settings to choose from & an advanced tab – that should satisfy both camps .
I also very much agree with Matthias Ettrich’s point about people trying to avoid doing things to windows .
The magic Alt-key is a great invention which makes things a lot easier.. especially if one has to deal with pop-ups or just wants to move window for merely a second too look @ for example a howto .
I think the Taskbar in KDE should give a lot more info on the Program/window it is showing .. dealing with the problem mentioned on here with soo many windows open & loosing where one can find a document in all of this .
Info might be of course the name of the documents/file .. last used .. maybe a preview of each window when the mouse is over the taskbar program entry .. as in Konqueror .
I think kompose … the KDE version of OSX expose should be a part of default KDE as it can solve the one-million windows open problem in my opinion … I love the inclusion of Superkaramba in KDE 3.5 … although its more eye-candy than productivity enhancement IMO … although Linux/Windows/OSX is heavily going to be an eye-candy competition with what the “features” of Windows Vista are .
One of the things I love about Windowmaker is the fact that the minimize button is at the top left .. meaning I dont have to be afraid of closing the application when Im only trying to minmize it as in KDE … another thing … one gets a small box with the applications icon in it for every application running under Windowmaker on your desktop … maybe something like this could be used under KDE for the taskbar .
Or maybe have no tasbar at all and just use kompose for “windowmanagement” which would give one the whole screen of space for extra info for each program running & have windows autoresize to fill the whole screen or if there are two windows on a desktop ( might not need virtual desktops any more with kompose for most things although there are exceptions ) they resize so its 50/50 for each .
Some very useful things have in my opinion been already inplemented in windowlab .
Or maybe somehow use the scroll-whell on most mice when over a windowborder to show a list of all windows availlable to choose from or something along that line .
KDE stability isnt the best either in my opinion.
Konqueror crashes on me sometimes .. especially since I have been using a tablet mouse .
When KWin crashes it doesnt restart which makes everything useless … in GNOME the windowmanager restart when it actually crashed (after running full-time for two days ) took something like 2 seconds .
I love the new “Get cool stuff” feature in KDE 3.4 .
Hopefully it will be extended to other parts of KDE .
Somehow the UI has to activly adapt more to the user … dont ask me how .. I havent got a clue ……
Post Scribbles :
Usability themes ? Very great Cool idea !!!
“There are other systems that use KDE. FreeBSD, OpenBSD, NetBSD, etc. ” – THX 4 reminding … I easily forget that there is an open-source world outwidth Linux .
re_re wrote :
“Oh yes, one more thing, I think kio-locate should be integrated into kde by default, instead of typing “locate:” there should be a small Locate bar (like a google search bar) integrated into konqueror.
I think this would be a very nice feature and with profiling support compiled into the kernel.. it is fast at finding files….. almost as fast as I can hit enter.”
-> agree certainly with that .
Just looked at simplekde.org … like what I see as long as the extra options are in an extra tab or so .
Could someone please get rid of the tree-view in filemanagers …. I really dislike it & that subtrees stay open after I have changed to a higher tree .. I allways have the urge to then carefully close the still open subtree … please ..
instead maybe have columns .
Matthias Ettrich said that he wants Konqueror to be a browser instead of a file-manager … I certainly disagree with that idea … there are already tons of standalone browsers out there & on Linux people use mostly Firefox as far as I can guess .
Konqueror would be the best filemanager out there with UI cleanup .. more stabilty …& maybe some UI features from midnight-commander .
“Gnome has a rather good ‘advanced’ mode…Its called Gconf.” -> Sorry but thats allmost like telling someone to play with the windows registry to change settings .
I personally do not like KDE 3.4 as much as I did GNOME 2.10. I actually liked KDE 2.5 a lot better than 3.4 in Ubuntu when I installed it. I thought that Ubuntu and nearly all distributions I’ve used whilst
I personally do not like KDE 3.4 as much as I did GNOME 2.10. I actually liked KDE 2.5 a lot better than 3.4 in Ubuntu when I installed it. I thought that Ubuntu and nearly all distributions I’ve used whilst using KDE were quite slow and crashy/buggy.
Submitted by 
This deap interest in usability shows how mature is KDE. Its community spent a lot of time to be sure they’ll have flexible technologies and a stable environment.
Unfortunately, as many persons think, KDE is slightly copying some of the MS Windows UI. But by discussing KDE’s usability, and thanks to the active and talented community, we will probably see more innovations and creativity in KDE 4.
I agree with Tomasz Torcz (the first comment to the first linked article). That does all sound like what Gnome did for the 2.0 series. And that makes me happy, because KDE is seriously in need of some simplification. The second article is very good as well, but draws just a little bit too much from OS X. I mean, don’t get me wrong, OS X is great, especially when it comes to usability. However, I’m not a big fan of appfolders, and, even if I was, that’s what ROX is for. As for the number of core applications, one must remember that this is linux, OS of choice, where you can pick from a million wms, document viewers, text editors, everything under the sun. Simplicity is good, but you don’t really want to remove ALL choices…
For me personally, even though KDE is packed with more features, and more powerful programs, but I seem to constantly have programs segfault, w/ 3.4 this has happened on at least 3 distros on 2 computers. from actual Kwin crashs to VNC desktop ones, I find stability and smoothness to be the problems with KDE. With all the features they give you I find the desktop more then usable, Yet the approach GNOME seems to be what I’m prefering, it just seems to be simplier and does exactly what I want a desktop to do: Let me get my work done, quickly, easily, comfortably, and without crashing.
What I actually want to see from the desktops is more freedesktop standards support. Apps communicate w/ Dbus, KDE and Gnome apps both being able to minimize to the tray in either(I think its getting good at that), .desktop support, etc. I’m really glad that Gnome 2.12 is finally supporting some standards, and KDE has progressed just as far.
I think KDE4’s goals should be QT4, stability, support of most if not all of freedesktop standards, some cleanups(even though I really like how much they have cleaned up in 3.x), and maybe adding extra features like an intergrated widget program.
I don’t know what you’re using, but I’ve had very few issues with KDE from about 3.3.2 onwards. It has been very stable, I think I’ve had like 2 or 3 Konqueror crashes in 6+ months, that’s damn good. 3rd party applications crashing? That’s not KDE’s issue I hate to tell you. If you want to argue otherwise, I’ll argue that Microsoft Windows is in deep shit then 😉
It is obvious from your post that you’re:
1. A Gnome user
2. Prefer Gnome (for a variety of reasons and personal preferences).
There’s nothing wrong with that, but it usually leads to inconclusive comparisons based on preferences and second hand guessing, rather than factual comments.
As to freedesktop “standards” – most of that’s crap. I’ve used Gnome a bit, and my honest appraisal is that it has serious UI problems. But then, you could say, that being a KDE user, I’m biased (or meet points 1. & 2. outlined above). Usage colours your perception.
As to KDE 4’s goals – QT4 is one of them. Every KDE release is motivated by being stable, as well as useful features etc. Freedesktop standards do nothing for myself, and I’d rather KDE focussed on it’s on UI enhancements, than something from freedesktop.
And the final comment – “cleanup”. I personally disagree. KDE is just fine as is, if anything, Gnome is oversimplified (which creates its own set of problems). You don’t have to use *every* feature, but it’s there if you ever need it. Users will generally use the features that they want, and ignore the rest. Have a look at Microsoft Word – you’ll see a tonne of icons up on the toolbar, how many users actually use all of them? Is MS Word bloated then?
Your arguments lack merit (in my eyes at least). It is simply more Gnomish trolling of a fine product.
Dave
This sentence…”Users will generally use the features that they want, and ignore the rest.” should actually read “Users will generally use the features they want, and THE REST WILL GET IN THE WAY.”
Yes in MY opinion I think MS Word is a bit bloated and the default toolbar layout stinks: How many people ever use the show/hide paragraph button?? I won’t go on and on..
I like the post below about Usability themes. That make sense to me…a different layout for different applications..office, home entertainment, home management/control….hmmmm more?
Out of curiousity _how_ do they get in the way?
do they prevent you from changing an option?
I guess I see getting in the way the same way as a pop-up window does when your about to click a link.
I’ll admit my first ever WM was KDE. I tried to use Gnome. Hated it’s UI nor can I understand how people like that as soon as you click an option it’s applied even tho you’ve not told it to yet.
That was the major reason I won’t use Gnome.
The other reason is lack of options. I like to fully customize. yes I probably don’t use over 50% of the options. nor can I see why I’d ever use them. But it’s nice to have that option. They don’t get in the way. don’t have to scroll to get to my options they are always right there visible. nicely organized in tabs and what not.
Alot more reading to see what option you are clicking but then again how often do you configure something?
it’s not like your always changing a huge number of options and because of that it’s slowing down your productivity.
As for the start of this thread on stablity. I’ve posted this before here and I’ll post it again. The True reason why I’ll never use Gnome again.
I use Gentoo Linux keep it very up to date. One week I decided that I was going to seriously try to use Gnome because I kept hearing how good it was. So I made myself a brand new system. Now because I keep things so up to date the same versions of the same programs where installed only it had a gnome Twist instead of a KDE twist going on.
I booted it GDM looked amazing. I was impressed. Gnome booted Ok All the GTK apps looked infinitely better than they normally did. Was impressed by the Cd icon having the Cd title. the volume increase/decrease was awesome.. Things are looking great I really started to see why I had been told good things. Then the trouble started when I tried to do the amount of things I normally did. Suddenly the response time was lagging my mouse was jumping, program crashed and then computer freezes.. so I had to reset and try again.. after 15 resets I was mad. I reset the computer because of serious crashes on average every 2 minutes. which taking into consideration of boot time that only gives you ~ 1-2 minute of computer use.
The same program versions installed. Tell me this is a kernel or X problem? Also not only did the kernel crash and X lock up multiple times but every gnome app crashed at least once before it locked up.
KDE has never locked up X or crashed the kernel. the programs do occationally crash.
Grant it this was over 2 years ago so I might give it another Go. However kde’s kioslaves have really got me hooked. plus the infectious gnome way of setting options right away before you tell it to.
Hmmm…How do they get in the way….hmmm
Simplicity runs parallel with productivity not “lack of features” as it is sometimes suggested.
I read a great article about HIG once and I wish I could remember where it was so I could post the link to give credit. It posed a few simple questions about how to make an interface instantly more useable and quicker to access for the user. One of the points was about icons (number of them, size, etc.) and how they should be few and NOT small. (I laughed out loud when it poked at ‘power’ users loving the ‘many small icons’)
Since the icons should be few there should be a place to configure which icons to display (so we don’t loose precious features or functionality) and of course there should be ‘sane’ defaults for those people who don’t know what they need or care about a billion options until they need them……
The options should pick themselves when someone chooses a usability theme.
One of the points was about icons (number of them, size, etc.) and how they should be few and NOT small. (I laughed out loud when it poked at ‘power’ users loving the ‘many small icons’)
This is really the thing that finally made me move to GNOME. The biggest asthetic problem with KDE is the lack of spacing. There isn’t enough space between anything. Space is as important in an asthetic work as the elements that fill the space. In KDE, the elements are beautiful (especially the icons), but there is no empty space to bring it all together.
> In KDE, the elements are beautiful (especially the
> icons)
Gosh, the icons? CrystalSVG has enough lens flare to burn my retina after 10 minutes of usage!
Lens flare? Do you even know what that means?
Using Gentoo Liunx you are not entitled to an opinion on the stability of the software you run.
If you can prove that you have compiled and configured the application and everything it interacts with in a supported way, you might claim it. But seriously noone has the time to read the proof just to read your opinion on the matter.
Until you do, this is just FUD.
Yes, MS Word is a bloated product with a horrendous GUI. Sadly, OpenOffice.Org have largely copied them. Try Lotus Word Pro instead.
OO.org is another group that need to begin thinking abut useability. Bravo to the KDE team for putting this well and truly on the agenda. Though currently a GNOME user (for the very reason of useability) I am eagerly looking forward to what arises from the KDE 4.0 development cycle.
First, everything I said was “my expirence”
and no 3rd party applications. the VNC program IN KDE, Kwim(the window manager) has crashed on me. It’s happened in 3.4.0 and 3.4.1, I haven’t been able to get 3.4.2, but I will try it out.
I don’t use gnome primary, I actually use Blackbox with a few little apps(bbkeys, torsmo, etc.). And I end up using KDE more often because of its apps. Saying that KDE is on par w/ Gnome in standards, saying its more powerful, more configurable, is “more then usablable enough” and doesn’t need any more usability work that I can see, that has better apps, is by no measure Gnome trolling.
Also GNOME doesn’t support the standards any better, both are still short on quite a few. I just want to see both the gnome and kde desktops more unified and have elements of both work better with each other.
Why don’t you read them:
http://www.freedesktop.org/wiki/Standards
(btw KDE 3.4 was quicker then gnome to jump on some of these, i.e. the .desktop one)
I don’t think I’m trolling by any stretch of the imagination.
Quote: “First, everything I said was “my expirence”
and no 3rd party applications.”
I got the impression that you were referring to KDE usage with 3rd party applications.
Quote: “VNC program IN KDE”
I don’t use (or need) VNC, which application is it? That said, I don’t think it’s fair to blame KDE if a application is dodgy. That shouldn’t (and in my eyes doesn’t), affect usability etc of KDE as a whole.
Quote: “I don’t use gnome primary, I actually use Blackbox with a few little apps”
Ah ok. Again, I misread your original comment and perhaps overreacted to it. My apologies.
Quote: “And I end up using KDE more often because of its apps.”
Exactly. That’s what usability really is imho.
Quote: “Saying that KDE is on par w/ Gnome in standards, saying its more powerful, more configurable, is “more then usablable enough” and doesn’t need any more usability work that I can see, that has better apps, is by no measure Gnome trolling.”
Again, my apologies. Whenever there’s a KDE article on osnews.com we have the Gnomites trolling and trashing KDE to their hearts content, and it really pisses me off. Then they start their “but it’s not really GPL’d”, or “it doesn’t conform to standards”, or “it looks childish”, or “it’s too bloated” bullshit arguments.
Quote: “Also GNOME doesn’t support the standards any better”
If you listened to the Gnomites you’d think so. I’m no fool, so of course, I know what you’re saying and obviously agree with you.
Quote: “I just want to see both the gnome and kde desktops more unified and have elements of both work better with each other. ”
Interesting comment – in reality you are right. The sooner Linux loses the multiple window managers/desktop environments and just has *one* primary desktop that is pushed by all distros, the better. One package management system would hurt either. Slim down the number of applications out there as well, and motivate developers to join the larger (and generally better) application development models. Include much better backward compatibility for libraries, so that a program from now can run on a library from 3 or 5 years ago, and vice versa. None of these are really desktop environment based issues, but overall usage problems that can, and do interact with the desktop environment to some extent.
Quote: “I don’t think I’m trolling by any stretch of the imagination.”
Fair enough, point made.
Dave
@dave…I don’t think you need to backpedal so much. You called it as you read it.
Sometimes hindsight is a good thing. I misread Seth for being a Gnome troll, and reacted accordingly. Me bad! I still think KDE is a much better desktop environment than Gnome, although I have a good liking for XFCE. If only XFCE used the QT widget, my that would be super sexy!
Dave
Some very good points raised. I used KDE for a long time, loving the extreme configurability (I guess I could be described as a poweruser++).
The interoperability is great, the whole desktop is very integrated. However, discovering that isn’t quite as easy.
KDE needs usability work. Hard. Many interfaces are directly unfriendly, which is what the article comparing with OS X basically points out.
Eventually, I got tired of KDE. I had tried a Mac OS X desktop on a friend’s laptop, and I was astonished. Perfect example of great usability and eye-candy, that really gets the job done. In KDE, you spend a lot of time trying to figure out which button does what, and why that button is there at all.
As is pointed out in the first article, customization is something completely different from configuration. Configuration is something you do one time and one time only (unless something breaks).
I was especially amazed at the KDevelop-XCode-comparison! The KDE folks could really gather a lot of inspiration from such examples, especially regarding toolbars. Toolbars are plentiful to the point where it really gets annoying, when using KDevelop and Konqueror.
So: Simplification of the KDE UI is needed. If KDE does not clean up it’s clutter, I doubt it will be a very popular alternative in the future desktop market. Especially now, when the GNOME desktop is beginning to see some serious integration and usability work.
– Simon
So: Simplification of the KDE UI is needed. If KDE does not clean up it’s clutter, I doubt it will be a very popular alternative in the future desktop market. Especially now, when the GNOME desktop is beginning to see some serious integration and usability work.
The main reason to use KDE is KIOSlaves. Once the Gnome people get their VFS working everywhere KDE will loose a lot of attraction. The problem is that in both the Gnome and the KDE camp developers are trying to make a good gui for Unix. When they really should focus on making a good GUI for people to get work donw.
One example: Ask your secretary how often she visits
/etc, /dev, /boot, /lib, /sbin /usr, /proc. Cbanses are that he/she never goes there. So why not hide these directories to ordinary non sysadmin users. Every time you suggest something like this, even on the usability lists, you get lots and lots of protests, saying that thes place ABSOLUTELY must be visible at all times to everybody. Of course these places might be useful to a sysadmin, but so far the most sysadmins I know usually use these libraries from the command line, not from the GUI.
If there really are things that normal users needs to changed in e.g /etc there should be a proper GUI to do it. Vieong the file structure in the konqueror or Nautilus is not good enough.
Other things that needs to be fixe is how to install software. Why not have some kind of virtual software foldet to wich packages could be dragged to install them, and then just drag them to the to remove them.
Whenever some dependency problem occurs, whatever needed should automagically be downloaded from the net if possible.
Another problem I see is that Gnome and KDE both seam to think that the user always work alone. He is suppposed to put his file in his home directory and should not access anything else. In a modern office you cooperate with your fellow group members and you share and exchange information. Why not create some kind of virtual structure to easy find what other users are prepared to share with you based on their group belonging.
Quote:
“One example: Ask your secretary how often she visits
/etc, /dev, /boot, /lib, /sbin /usr, /proc. Cbanses are that he/she never goes there.”
So?
Does she visit c:windowssystem32 that much?
I don’t think so.
What’s your point?
The following quote is your point?
Quote:
“So why not hide these directories to ordinary non sysadmin users. Every time you suggest something like this, even on the usability lists, you get lots and lots of protests, saying that thes place ABSOLUTELY must be visible at all times to everybody.”
They ARE, yes ARE hidden !!!
You do not see /boot or /usr at any time in KDE, unless you specifically choose to see them. KDE starts in your home folder, you know /home/your_account
I fail to see your point here.
Quote:
“Whenever some dependency problem occurs, whatever needed should automagically be downloaded from the net if possible.”
Did you try a linux distribution in the last, say… 3 years? Automatic dependency resolving is somewhat standard in any modern package management of modern linux/unix distributions. If not, then you’re dealing with a rubbish distro.
Quote:
“Another problem I see is that Gnome and KDE both seam to think that the user always work alone. He is suppposed to put his file in his home directory and should not access anything else.”
Sorry, but saying this shows that you’ve never ever used KDE, and it even highly contradicts with the first paragraphs of your post.
Quote: “So why not hide these directories to ordinary non sysadmin users.”
Yeah, sure, like Microsoft Windows is (by default) designed to hide file extensions, and your system files, which is a friggen pain in the ass. That’s really great UI. Not. You’ve obviously never worked on a helpdesk before have you (that deals with the public)? As an example, you might have 2 files named lwremove.exe and lwremove.ini. With extensions turned off, they’re like “wtf, I have 2, which one!”. The customer has already started to panic, and then you have to spend extra time explaining to them etc, and getting them to click on the right item by using icon identification (ie. the icon I want you to click on looks like THIS). That’s disgraceful.
You need to edit/remove certain system files to fix a problem? Sure. You tell them to go to c:windowssystem32. “I don’t have it”. So, you have to spend another couple of minutes getting them to unhide it.
Is this the sort of crap you want? Go friggen use Microsoft Windows then. Your argument is completely blown out of the water by the fact that Apple’s Mac OS X doesn’t *hide* any system folders etc. All perfectly visible. Is OS X now badly designed? eh? Come on, make a reply to this one please.
Quote: “If there really are things that normal users needs to changed in e.g /etc”
Most things there is. Try configuring Microsoft Windows in many areas – guess what! There’s no GUI tool either, it’s drop to a dos command line and do it from there. Example – release your IP address. Do I hear you bitching about Microsoft Windows? No. Again, your argument has no validity.
Quote: “Why not have some kind of virtual software foldet to wich packages could be dragged to install them, and then just drag them to the to remove them. ”
uh huh! So you have used Mac OS X then. The problem with this, is that simply removing the application folder from /applications *doesn’t* remove all of the files from /system and /library, etc. Don’t believe me? Do a finder search. Both rpm and dpkg remove packages and config files when told. Oh, and Microsoft Windows is a shocker for this. Ever seen how much crap dll files are left over, or how much shit is left in the registry? Ever tried editing the registry? It’s really user friendly – *NOT*.
Quote: “Whenever some dependency problem occurs, whatever needed should automagically be downloaded from the net if possible.
”
Yeah sure. Debian’s apt-get does a pretty damn good job for me. Much better (and more accurate) than either OS X or Windows. Your argument is?
Quote: “Another problem I see is that Gnome and KDE both seam to think that the user always work alone.”
Um yeah. Sure. That’s called browsing the network. Windows has network neigbhorhood or my network places, OS X has go > connect (and you can browse if you want). Linux isn’t any harder. Simply edit /etc/samba/smb.conf and change your workgroup name (KDE has a gui for this as well I might add in kcontrol). Restart the smbd and nmbd daemons and voila. Open up smb4k (linneigbhorhood is dodgy and doesn’t support samba3 as far as I’m aware, and xsmbrowser is not as good as smb4k as far as I’m concerned). You can happily browse your network and mount shares etc. The whole idea of this is security. It’s the whole idea of a Unix like operating system – security is valued. Not like with Microsoft Windows, where it’s an added-on thought.
Your arguments are nonsensical at the best, idiotic at the worst. And – they don’t necessarily apply to KDE, or even Linux!
In the end, if you don’t like it – either code it yourself, or hire someone to do it for you. In the words of Dr. Denis Leary – “shut the fuck up”.
Dave
Exactly. Hiding directories can have problems if done the Microsoft way, but think of the original Mac. That system proved this can work.
Has KDE been ignoring GNOME for years? It’s not like simplicity is some secret.
KDE developers are in a big dilema:
One part of them know KDE doesn’t have the clarity in their apps. and menus, but how to fix it?
The die hard KDE user likes it the way it is, klutered, they feel KDE is powerfull and that’s way they like it.
But how to reach users who likes simplicity w/o arm the number of user they have now?
It will be a hard task.
That’s why themes were invented. Please use themes to death. I would use a developer theme (clutered, lot’s of info). My mom would use kitchen user
simple uncluttered interface, same basic functionality though…. make everything configurable into themes and you can please 95% of the population (tha’s Windows not so secret after all no?).
good luck and let’s get to work now…
So this — http://www.simplekde.org/ — is the fork in the road that speaks to many of us who feel KDE has become cluttered & bloated. I’m looking forward to it.
Well, this is accomplished by having a well designed ui that has sane defaults instead of no options.
Since most users never touch the default options, it makes sense to create a default that makes a lot of sense. However, since many people like kde’s configurability, let’s not remove all the options. This will please both camps. Wouldn’t KDE be great for the new user if s/he never had to go into the control center?
Also, I think we have to realize that there are two desktops for a reason. There is no reason to be all things for all people. Let gnome take the simplicity is better approach for those that like it. Why all this talk of kde should be like gnome or vice versa? Let them be different and let the people that like the different approaches use the DE of their choosing.
oggb4mp3
forgot my password
There were indeed a reason for the two desktops. Qt’s licence. But that issue changed. But it wasn’t about usability.
Two great blog posts. I especially liked Matthias Ettrich’s point:
Nobody likes to do window management. Wirth was right, although few people believed him when he designed Oberon. Users seem to go a long way to avoid having to adjust a windows position or size manually. This isn’t just untrained users, this is everybody. You can observe the same patterns on hackers everywhere…
I used to code witha bunch of windows opened up in xterms. They were always exactly 80×24. Why? Because that’s the default size of gnome-terminal! The only time I’d bother resizing is when I needed to see more compiler output at once. Now, I code in a single 1920×1200 window with one copy of VIM open showing three split-views side by side. The productivity benefits of not having to move windows around, resize them, always being able to see exactly what I need, they’re enormous.
Ultimately, I think the way forward is for apps to do more management of their own windows. Like the example I mentioned above — VIM knows better how to handle my code windows than metacity does, so let it do the management. I’d also think some sort of garbage collection for windows would be nice. I tend to end up with a dozen terminal windows or a dozen browser windows open at once. Finding old ones is usually too much trouble, so I tend to just keep opening new ones. Ultimately, the problem stems from the fact that what the user wants is a particular document (be it a webppage, source file, or whatever), but the only thing the window manager knows about is windows. A way for the user to experss what document they want, and for the window manager to bring that document to them would be very useful indeed.
> That does all sound like what Gnome did for the 2.0 series.
You’re wrong. It’s not about removing 80% of functionality. Usability is not about removing functionality.
> I had tried a Mac OS X [..] Perfect example of great usability
Mac OS X has also its usability flaws, iirc I heard the print configuration is one of them.
Yes, the print configuration is a usability flaw*. But the OP said “a perfect example of great usability”, not a “great example of perfect usability”. There is a very significant difference between the two statements.
*) For those who’ve never used OS X’s print dialog. It’s got a combobox in the middle with a big list of options (like “layout”, “paper settings”, “printer settings”, etc). Choosing a different item in the combobox changes part of the dialog below it to refect the option. The thing is, its highly unusual to use a combobox for something like that. Usually, you see a set of tabs, or an icon bar or something. The combobox doesn’t even have a label showing what it does. You just have to click on it by accident to find out.
> The die hard KDE user likes it the way it is, klutered, they feel KDE is powerfull
I doubt that they would miss clutter. And usability doesn’t mean to make your desktop less powerfull (or only if your desktop is called GNOME and you need a reasoning for missing functionality). Look at Mac OS X or Pages as example: simple UI but not less powerfull because of that.
“Mac OS X has also its usability flaws, iirc I heard the print configuration is one of them.”
Then again, we’re not discussing Mac OS X’s usability flaws, but trying to find good usability solutions and then implement them in KDE. Right?
Finally it seems someone understands that the “Just throw it in wherever it fits” mentality doesn’t cut it.
If KDE did some of these things, that would go a long way toward making me less offended by how incredibly ugly the whole desktop experience is now. If only KDE could get away from the anti-corporate development licence of QT and naming every application with a prominent K, perhaps we could achieve the nirvana of a single mainstream linux desktop environment someday. If they actually care about usability and not just tagging on every single concievable feature that anyone could imagine, KDE might actually become *gasp* useable. Please go forward with this!
perhaps we could achieve the nirvana of a single mainstream linux desktop environment someday.
Don’t hold your breath for that one.
If only KDE could get away from the anti-corporate development licence of QT
Heh … heh … which corporate development did you mean?
Corporate that doesn’t want to pay money for their commercial products?
If only KDE could get away from the anti-corporate development licence of QT
Heh … heh … which corporate development did you mean?
Corporate that doesn’t want to pay money for their commercial products?
From what I have seen and experienced, Trolltech’s licensing is bifurcated into two extremes:
1. The pseudo-GPL license for “open source” which prevents you from later dual-licensing your Qt-based products and covers only a subset of Qt. As this license claims code ownership rights — dictating what you can and cannot do with code you write — it is misleading if not an outright lie to use the term “GPL”.
2. The extremely expensive “commercial license” which has many restrictions on what you can do with the Qt source code.
There is no entry-level affordable commercial development license.
So the vast bulk of developers who would create an ecosystem of libraries, components, and widgets for Qt will never be able to.
Qt has fated itself to being the core of KDE and nothing much beyond this. There may be an app or two from time to time that uses Qt but these apps will be written by large companies that can afford to write their complete application stack themselves.
Trolltech has a quality framework in Qt starting with the 4.0 version (the 3.3 version was not high quality). I hope the company decides to lower the price points for Qt on the desktop and focus their revenue generation on the mobile phone companies. Otherwise there will never be any significant future for Qt unless the company is bought by a player with vision.
So the vast bulk of developers who would create an ecosystem of libraries, components, and widgets for Qt will never be able to.
The same vast network of proprietory but freely-available components that exists for GTK+, eh? Who cares about those components if nobody is able to use them?
Do you genuinely believe that a falsehood repeated with some frequency is thereby made true? Or is fraud your game?
The pseudo-GPL license for “open source” which prevents you from later dual-licensing your Qt-based products and covers only a subset of Qt. As this license claims code ownership rights — dictating what you can and cannot do with code you write — it is misleading if not an outright lie to use the term “GPL”.
What can and cannot be done with the Qt code base when licensed under the terms of the GPL is exactly the same as what can be done with every other single bit of code ever released under the GPL. That Qt is alternately available under licenses other than the GPL, proprietary or otherwise, has no bearing whatsoever in determining the what the precise terms of the GPL are when applied to Qt.
And as for Qt Free Edition being a subset of Qt, that is a misleading statement at best. The commercial and free editions of the X11 version of Qt are identical. The free edition of Qt on Windows is basically identical to what is avaible for X11 under both free and commerical licenses (excepting, of course, the necessary platform specific modifications). The commercial version of Qt for Windows also includes some integration features, most notably VS++ integration, which are not available under the free edition of Qt on Windows.
So the question is then what exactly constitutes the code base rightfully deemed Qt? To some extent the question is semantically vague, and thus no definitive answer is possible. But if we hold to the common understanding of the concepts involved, I feel it’s pretty easily demonstrated that you are either a) guilty of faulty thinking, or b) guilty of willful misrepresentation.
Only that code which is common to all versions of QT, roughly that which comprises the free edition, can be regarded as being a necessary condition for the code to properly be deemed as Qt. To hold otherwise would commit you to the view that since the Windows specific features of Qt are not available on either the X11 or the OSX platforms, Qt therefore can be said to not exist on either X11 or OSX. I trust that most people would regard this to be an absurd claim.
So is the code held in common by all variations of Qt, namely that code comprising the free edition, a sufficient condition for deeming it to be Qt? On X11 it certainly is and, the vagaries of natural languages notwithstanding, I think that justifies us in concluding that the code base constituting both the free edition and the X11 commercial edition is the best candidate for being regarded as Qt. Under this view, all other code would grouped under the title “+ platform extensions”, thus meaning that TT offers “Qt” on X11 and “Qt + platform extensions” everywhere else. Thus your claim that Qt Free Edition represents a subset of Qt is rendered false, as “Qt + platform extensions” is by definition then a superset of Qt.
Copying Windows UI is not necesarily a bad thing. Windows does quite a few things right despite what many may claim.
I consider myself a KDE power user. I can tell you I’m not afraid of change, and of course some parts of KDE need a rehaul. What I don’t want and I’m sure many power users don’t want is to lose funcionality just to achieve better usability. That’s what Gnome project has done, and what many of us dislike.
So the challenge lies in cleaning the clutter while keeping most of features and funcionality. Pretty big challenge, but I’m sure KDE guys are more than up to the task
This circle jerk will be fruitless. Any discussion about usability needs to involve users.
Corporate that doesn’t want to pay money for their commercial products?
I don’t understand your point,
I make commercial software for MS Windows platform and I don’t need to pay a license to Microsoft neather to GPL my beloved code.
Why shouldn’t Trolltech be able to make money on their products?
Yes, you are. You are paying for your Windows license. Also, Visual Studio costs money, a lot of money.
You don’t have to GPL your code to use QT, don’t spread FUD.
Why shouldn’t Trolltech be able to make money on their products?
Nobody said it shouldn’t, but since KDE is using Qt the license issue is there. Denie it won’t help.
They guy is talking about his own experience, KDE has crashed on my 5 times more than GNOME, its all about personal experience, maybe for you KDE has been stable, for me don’t.
> They guy is talking about his own experience, KDE has
> crashed on my 5 times more than GNOME, its all about
> personal experience, maybe for you KDE has been stable,
> for me don’t.
Then stop compiling it and use a proper distro.
while were are on usability… fonts anyone? please make a better font engine. I cannot stand the small size rendering of TT fonts with FreeType. Can anyone spend some time to actually produce better looking small size fonts and font engines? Again FreeType doesn’t cut it.
I use KDE every day, and I love the way it is.
I most admit that I become somewhat concerned every time people talk about removing something from KDE.
Like the second article suggesting that Konservation, Kopete and KSirc, should only be one (I suppose he wants to keep Kopete, only) I find this to be wrong, I could understand KSirc and konservation, but Kopete is not to me an IRC-client (I don’t want my IM-client also to be my IRC-client). Also he suggests that playing and ripping CD’s should be done by ones favorite media player, don’t like that either, I don’t want to play CDs in the same application that plays my Oggs, neither do I want it to rip my CDs.
KAppfinder has to stay since not all applications have a desktop-file, and no filing a bug-report is not the solution, maybe it should be filed, but that wouldn´t solve the problem this instance, KAppfinder would and does.
I like the idea about the commen viewer, I think it could already be done since most of the functionality is already there in a kpart.
KGet should _not_ be removed from the application list, I use it often even though, I don’t use konqueror for web.
I’m strongly again removing any configurability from KDE, but I think Control Panel could use an overhaul.
KDE is of cause not perfect.
One thing I’d really like to get rit of is aRts, I’ve a soundcard with a hardware mixer, so I don’t need a soundserver taking ressources.
People talk about making KDE easier for normal users to use, so here is a little story…
I had a friend over the next morning while I sleept, she managed to mount a CD, watch video from it, play games, surf the net and stop all download activity (on purpose) by her self in KDE without wakeing me. She is a prime example of a user who normally need help with everything on her own computer, at home she runs Windows XP.
The only thing she didn’t managed to do was unmount the CD, because she’s used to push the button on the drive, which didn’t work on my computer that day.
have been in OS/400 since 1988. Hmm, i’m getting old.
But, on a serious note, i think that’s a great idea.
one must remember that this is linux, OS of choice, where you can pick from a million wms, document viewers, text editors, everything under the sun.
There are other systems that use KDE. FreeBSD, OpenBSD, NetBSD, etc. It is really annoying to constantly see KDE/Linux.
And please dont say that KDE was made specifically for linux.
Open source is more than linux.
Really!!!
They still need to focus on the clutter and mirroring other os’s enough so that companies and institutions would consider switching to such an environment. Once that happens they can talk about something else.
Isn’t it amazing how much more attractice OSX is in those screenshots? A lot of it has to do with the relative simplicity and lack of clutter.
That’s precisely what they’re talking about! Did you even read the posts???
(This is directed to Matthias Ettrich… I don’t want to post my email on QtDevelop, I tried a harmless fake “not@provided”, didn’t work)
Here am I, discussing the Monster with Doktor Frankenstein himself. Ooh, the irony of it all… (a little dramatic moment, drums fading to background, please)
Maybe I didn’t quite understand your real intentions, so pardon me in advance if so.
> Here?s how it should be (and surprisingly this is also how Windows Vista works): Having images in folder results in little button that says ?View slideshow?. You click it, you get a full screen viewer that shows your images. Simply, intuitive, what you want, what you expect, and no window management action required.
Apart from the problem you described (i.e., having to toggle the full-screen option), that’s exactly what I get in Konqueror 3.2.3. I have changed (in Konqueror’s Configuration, File Associations) “image” to use an external viewer (the default:Kuickshow). Now: just enter a folder and click on the first pic, it gets shown at full-screen. Move the mousewheel and voilà , you easily browse all that directory’s pics (no window management here). Is that what you wanted? (Though I recognize there’s no indication of this use, it’s pretty easy to discover this the first time you accidentaly moves the wheel — BTW, if you’re browsing the Internet this will present all the pictures in cache… pretty nifty, too, I might say).
I thought you knew about this. I must be misunderstanding the whole thing… 8-/
Maybe you’re referring to “default configurations” and how the user should get all this without having to change any options… I would agree with your point, if so.
> And while being at it, single click for opening documents is wrong. Initially I thought it was a good idea, too, but after having tried to use it for a while, I changed my mind. Unless I?m the only one who constantly starts applications by accident when I want to move files around, or delete them.
I gave a reasonable amount of consideration to this recently. To avoid opening things by chance, it’s enough to select things not by clicking, but by “drawing a square around selections” — much like one does “square selection” in GIMP. Works like a charm, and it’s a small price to have single-click selection (which is a real saver from RSI). Alas, one can use Ctrl-Click to select, which is even easier but of less intuitive usage. But, then, all this is a matter of taste, I guess… people can get fond of double-clicking, I suppose.
Again, I suppose you’re talking about first-day users… but then, doesn’t KDE ships with the doubleclick setting on?
Lee Nooks.
Then stop compiling it and use a proper distro.
I used Kubuntu a KDE distro, I tried ARK linux too among others, same results.
I tried Kubuntu and had a ton of problems as well. 3 years using Mandrake and never had a single problem with KDE, though. I came to the conclusion that Kubuntu was rushed out a bit quickly (wasn’t it the very first distro with 3.4?)
I don’t know, it seems to me that usability for KDE is like an afterthought. Unlike Gnome which was built around the concept of natural ease of use, this wasn’t a priority in KDE before. You can’t just pin on the usability thing like a feature upgrade, every single little interface decision has to be rethought all over again. Its an overhaul, and naturally there will be resistance to change for the current users.
Plus, it’s questionable wether the usability of power users and usability of people who just need to get work done is reconcilable.
Less klutter, kandy kolors, and applications dubbed by k* would help for one
Also, the kicker is just really fugly.. and large, and the desktop icons etc look like crappy web links etc.
Those are always some of my complaints… plus yeah… too much copying Windows…. and ugly fat distracting widgets (though I’d say GNOME is no saint either with there incredibly dull widgets)
Whats needed is a simpler default interface, but one that leaves all the options available through an “advanced” menu or something. Gnome users always complain about how difficult it is to use KDE, but I actually find it simpler. I don’t like the defaults, and when I want to change something in KDE I look through a bunch of poorly designed interfaces. But that’s better than Gnome, where I have to Google about how to change the app (if it’s even possible), and then edit some obscure config file.
Whats needed is a simpler default interface, but one that leaves all the options available through an “advanced” menu or something. Gnome users always complain about how difficult it is to use KDE, but I actually find it simpler. I don’t like the defaults, and when I want to change something in KDE I look through a bunch of poorly designed interfaces. But that’s better than Gnome, where I have to Google about how to change the app (if it’s even possible), and then edit some obscure config file.
Amen!
I use KDE full time currently on my PC. But I also have an Apple iBook. I really like the points the guy in the second blog made – reduce the clutter!
I used to really like KDE, but after using OS X I realised that there were some deep flaws in KDE. Things I could do quickly and easily on my ibook, were just painful on Linux/KDE. So much so that as soon as I can get around to it, I’m moving my PC to Ubuntu/GNOME as GNOME really seems to be “getting it” – at least in some departments. GNOME has some annoyances sure (I don’t really like nautilus, no decent CD Burning app, etc), but the cleanliness of it makes it far more pleasant to use.
No app should have more than one toolbar (by default). That is a great rule
And trying to use the system with one mouse button! I don’t think it would be possible with KDE – which is a shame, because I’ve seen older people really struggle with multiple mouse buttons.
I hope that KDE4 will move me back into the KDE camp.
Interesting, I had the same experiences and finally gave up on KDE (and that after being a loyal KDE user for more than 3 years). Gnome is quite a bit cleaner and faster and things in most cases just seem to work, similar to OS-X.
I do hope that KDE will really work on a complete redesign, simplify and make it look more professional. In particular eliminate duplicate (sometimes triplicate) applications by default to clean up the menu. Instead of having three separate projects for a particular application, they should really be more forceful and scrap the ones that are not as good and focus on the ones that do work well.
The thing however that in the end really made me switch over was kmail. When they “upgraded” the mailbox format from 3.3 to 3.4, I almost lost thousands of archived emails. Upon checking, this was deemed part of “improved features”, something that for a production systems is just simply unacceptable. This is an example that the quest for more and more features and options in KDE comes often at the cost of compatibility, which for a production system is just unacceptable. I ended up moving to Thunderbird for emails, Firefox for browsing and Gnome as a desktop, a combination that I found to be more appealing and most importantly, more stable.
The thing however that in the end really made me switch over was kmail. When they “upgraded” the mailbox format from 3.3 to 3.4, I almost lost thousands of archived emails.
Ah yes, that happened to me when moving from 3.1 to 3.2. They didn’t change the mailbox format, but they fixed some bugs in it so it more closely matched the spec. So of course, my old buggy kmail had saved messages incorrectly, and the new kmail would not read them. I ended up writing a perl script to go through my emails and correct the errors in them – then I got kmail to re-index my mailbox. Really beyond the abilities of a non-techy user!
I’m not looking forward to trying to move my 15k+ emails to another mail client. I will likely choose thunderbird (can it read maildir format?), because it is cross platform (and I use it at work on Windows).
I read through the horror stories, and I am amazed that there is still some belief that Linux is ready for the mainstream desktop.
Until the focus is put on solving the user’s problems and not making life easy for the lazy developer (who obviously could not be bothered to write a data migration utility), Linux is not going anywhere on the desktop.
Quote: “Until the focus is put on solving the user’s problems and not making life easy for the lazy developer (who obviously could not be bothered to write a data migration utility), Linux is not going anywhere on the desktop.”
Maybe you should get off your own lazy ass and start coding then if you think you can do a better job?
I”m getting damnwell sick and tired of some people just expecting things to be done for them, the way that they like it, and then bitch when they don’t get it and dare put down the developers.
These guys kode (OK, I couldn’t resist the pun) in their own time, out of their own “pay”, for the fun of it, and the purpose of creation. If you want better, either do it yourself, or go and fork out money to hire a “professional” developer to write it for you. Good luck.
People like you are the height of rudeness in my eyes.
Dave
I use KDE and I have for some time, I find it to be quite powerful and I love how much it can easily be tweaked to my needs.
That being said, I deffinitly agree that KDE could be slimmed down a bit.
-only include a few games…. not 50.. or however many that is
-include amarok as the default media player and add video playback capabilities to it
-make the default settings less cluttered and slimmed down a bit… (not so slim as gnome though)
I think it would be a good idea to not only change the default settings, but to slightly change control center, maybe have very basic configuration on the surface, then have an advanced button that gives the user the full ability to do whatever he or she wants.
Oh yes, one more thing, I think kio-locate should be integrated into kde by default, instead of typing “locate:” there should be a small Locate bar (like a google search bar) integrated into konqueror.
I think this would be a very nice feature and with profiling support compiled into the kernel.. it is fast at finding files….. almost as fast as I can hit enter.
KDE and Gnome are different. This is good! I used to use Gnome and then rediscovered KDE and discovered how its little features made my life nicer, now I find Gnome constricting and cramped. For me, KDE’s defaults are at least usable whereas Gnomes are not and changing things in Gnome is more complicated (read gconf). I do acknowledge that Gnome has its place and so does KDE, just as politics is ruined by both sides competing for the center, I fear that KDE and Gnome will pick up eachothers vices as well as eachothers strengths. I love how I can sit down at any Linux (or other NIX) workstation and choose a desktop that fits my likes, wether that may be KDE, Gnome or WindowMaker. If you don’t like KDE, stop bitching and visa versa.
Seth
(A KDE fan, but I hope not zealot)
I enjoy seeing this take place, espically with the growing amount of “Linux is hard to use thus not ready for grandma” comments. I wonder if KDE has tried a similar study that GNOME did, involving recording the user’s mouse and eye movements. This would judge how much the user searches through non-intuitive buttons and menus.
————-I enjoy seeing this take place, espically with the growing amount of “Linux is hard to use thus not ready for grandma” comments.————
The funny thing is…… That’s exactly how it is in my family. My grandma uses linux. KDE. Fedora.
I get maybe 3 tech calls a year since the migration from windows. Calls are always bound to happen regardless of the platform. Mac. BSD. Whatever.
So far this year I’ve had one. Password/email issue.
These topics are often generally useless – you have the KDE users who on the whole prefer to keep the desktop as it is. Then you have the GNOME users who don’t use KDE because of its usability problems and prefer the simplicity of their desktop. THen you have the OSX users who wonder why people use either. Then there’s people like me, who use GNOME even though they prefer some KDE apps (kdevelop, kate) – usability issues aside.
Confused? You should be!
I was a KDE user for a little over 2 years, then finally gave up and switched to Gnome after seeing the bloat increasing with new releases.
Alright, I’m a Gnome user since GTK2’s release, but before that I was a KDE user. I also agree KDE has too many useless tooltips, menus, options and menubars. I’m in no way suggesting they scrap everything however. I think they should work on a simple default interface, but give users an option to go “into advanced mode” which would provide all the extra functionality. This would appeal to both new and old.
Yes, an advanced mode is a good idea. There is no point in having every twiggle and twoggle at your finger tip when (as someone else pointed out) you don’t change those settings that often.
I also don’t think people like to have their train of thought interrupted when they have to ‘look’ for a particular feature button. This of course is impossible to have implemented perfectly, but I do think discussion can help clean things up. It really isn’t that hard to set a standard, but it is hard to have people agree on one….and stick to it…and not get mad and fork a project…etc.
KDE has a lot going for it, and I think it could really challenge some of the popularity of Gnome if the interface was a bit more sane and didn’t draw a bead of sweat at first glance. I go from Gnome to KDE to XFCE to ICE to whatever and always find myself going back to gnome to do any real work. The main reason??? I keep pushing buttons, clicking on things, hovering over icons, and exploring just to see whats going on. As much as I fing that fun, it just isn’t very productive.
Its called Gconf. Sure the included editor might not be to many people’s taste, but there are plenty of third party tools that expose more options.
I think the key is when you give somebody ‘beginner’ and ‘advanced’ modes, nobody ever selects beginner.
——–I think the key is when you give somebody ‘beginner’ and ‘advanced’ modes, nobody ever selects beginner.———
I agree 111%. For those who desire simplicity, and “usability”(which is very subjective) the DE should come shipped in “beginner” mode. This group like it or not is the majority. For those of us who love the advanced and highly configurable options, it shouldn’t be hard to put into “advanced” mode. Perhaps one or two switches in a text file.(we’re geeks, we can handle that)
That would make it ok for everybody. I’m one of those who loves controlling everything.
“And while being at it, single click for opening documents is wrong. Initially I thought it was a good idea, too, but after having tried to use it for a while, I changed my mind. Unless I’m the only one who constantly starts applications by accident when I want to move files around, or delete them.”
Oh gods, yes. Single-click-open was a major pain in the ass until I managed to dig up and change the setting.
Kubuntu is a very poor KDE distro. I tried it and it’s dreadful, it’s not even a full KDE. There’s a lot of stuff missing. I would not recommend Kubuntu as a good way to experience KDE.
Dave
I don’t see any progress with “SimpleKDE” anymore. Until now every distribution packaging of KDE applies more patches to KDE (SUSE, Xandros, …) than SimpleKDE.
> please make a better font engine.
Outside the scope of KDE or any other desktop.
> Like the second article suggesting that Konservation, Kopete and KSirc, should only be one
I wonder about him complaining that his installation contains both Konversation and KSirc. Konversation is not part of KDE so he installed it himself. Arguably KSirc is not the best IRC client (anymore) and should not be shipped with KDE, or replaced by Konversation.
> I would not recommend Kubuntu as a good way to experience KDE.
Ack. Kubuntu is and becomes more a heavy modified distribution of KDE. Because of the customizations (and I don’t talk about shipping only a subset of KDE release) you cannot argue about (vanilla) KDE. It also includes, at least in Breezy, much experimental stuff (alternate control center, Katapult, …).
A plea to KDE
(paraphrased from Jack Nicholson’s ”
As Good as It Gets” character quote…)
Dont be like Gnome! Don’t you be like Gnome!
“Also, The genius of Mac UI is overblown fluff. More on this on 11 o’clock news.”
I allready like KDE the way it is now since 3.4* with some perceived sheduling.It’s fast enough and fucntional,and very easy to operate.If i really wanted a simple desktop i would install fluxbox (which i did and use occasionally to keep people quessing at what the heck is installed) and not gnome.So i hope the people of KDE will continue the good job and don’t let them steered to much by public opinion but rather by common sence and technical knowhow as usual.
How about changing the krappy naming convention with something sensible, Komrade? Yes, it’s trivial to change, but its got better ROI than the complicated fixes.
I think that both KDE and GNOME are too complicated for all but the most curious, explorative, and tolerant users, but KDE takes the cake in this sense.
If I see 10 buttons lined up across the top of an application, there is no way that I’m going to be able to infer the actions of these buttons from their icons without a minute or two of tooltip hovering. How intuitive do these button icons need to be in order to be useful? You’d be suprised. Many users have no clue what the justification buttons in MS word do, and I don’t know how the icons themselves could be any clearer.
KDE is loaded with such things. Buttons galore, tabs running down the sides, every feature prominently displayed. KDE applications scream: “look at how advanced I am, you can tell by the amount of menus and buttons and tabs that I can handle anything you dare to throw at me”… if the user can figure out what any of it does.
GNOME is a little bit better. All buttons say what they do underneath them, at least. There is very little configurability, but there is fairly capable theme engine. The interfaces are impressively intuitive and consistent (relative to KDE), but in some places there are glaring holes in functionality. Why isn’t there an option to burn an audio CD if all the items in nautilus CD burner are audio files? What good is a full-blown collaboration suite like Evolution if I can’t do simple things like subscribe to newsgroup?
The default interfaces presented by the DE must be even simpler than GNOME, let alone KDE, but they should be adaptable: there should be preferences for default interface set. For example, I should be able to easily tell Evolution that I don’t want to use the calendar or task list, and those elements should be removed from the default interface. Also, the interface should be designed such that elements are added and removed as they are needed for the usage context. For example, the media player should present a very simple interface when I click on an audio file in my browser. When I then open a video from the file manager, the media player should expand to show the playing audio track and the selected video in a playlist element. Then when the audio track finishes or I click Next, the player should expand to include a video canvas element.
The essence is that the user doesn’t want a swiss army knife or a collection of special-purpose knives. The user (ideally) wants a single knife that automatically becomes the ideal knife for the task at hand without having to find and unfold it. In practice, there should be more than a single application, of course. But some things that can be easily integrated. For example, a unified document viewer. Or a single document editor that can be anything from emacs to kate to quanta to openoffice to kdevelop depending on what kind of document is being edited. The interface complexity is dynamically adjusted to the simplest subset appropriate for the usage context.
It is apparent that application support is critical, but first the DE needs to expose an API that handles the interface customization. The application defines a base (default) interface and then defines supersets that inherit from and adds to simpler interfaces in a hierarchical manner. A major practical limitation is that this scheme requires a unified widget library (which we should have anyway).
We’re going to need some developer buy-in to get where we need to go…
The essence is that the user doesn’t want a swiss army knife or a collection of special-purpose knives. The user (ideally) wants a single knife that automatically becomes the ideal knife for the task at hand without having to find and unfold it.
YES!!! Especially for information browsing and gathering. The ultimate tool would be like the one described above.
We’re going to need some developer buy-in to get where we need to go…
Better buy a free-lance brain surgeon.
Well, I don’t know why I read the whole thing. It was “same ol’, same ol'” things.
* QT’s not free eviiiils — Pu-lease! Can you inform yourself before posting a troll comment?
* Usability mantra — I see there’s a seriouse amount of people who things usability means cripping features. Well, do whatever you want. Don’t be surprised if one day you will be only able to click icon on desktop, and nothing more. That would be simple, no?
* KDE crashes — yes, yes. Windows crashes. Gnome crashes. KDE don’t crash on me very often, same goes for Windows. If something works for most of people, and it dosen’t for you, maybe it’s time to check out your system?
* Stock KDE should have sane defatuls — Why? You know what’s a secret of Ubuntu success? They managed to polish Gnome. It’s up to distro maintainers to set up your KDE or Gnome.
* OMG! Konqueror’s buttons — Hh, the horror! It takes 40 seconds to change it whatever you like. Hint: yes, you can do it with all the application! But insted you prefer to cut this feature, and give us uneditable ,,sane defaults”.
I should point out that Ubuntu’s GNOME is 99.9% the same as vanilla GNOME defaults. They changed the startup splash and made a theme that I (and others) consider to be ugly. In the first version of Ubunty (Warty), they deviated from vanilla GNOME 2.8 by introducing a Computer menu on the panel. This was a great improvement and inspired the new defaults for GNOME 2.10, which split out the original Computer menu into Places and Desktop. Ubuntu Hoary (current production release) follows the same menu structure except it changes the title of the Desktop menu to System. The only other change I’m aware of is the location of the trash to the right of the pager.
The secret of Ubuntu’s success, therefore, is not that they polished GNOME, but that they created the first free Debian-based distribution to feature a modern desktop environment. It could also be that people really like GNOME, but can’t stand Red Hat…
there appear to be two opposing views from the two articles, both of which have merit, but leave one confused as to which direction to pursue:
1) the anti swiss-army knife viewpoint where it is better to have several really good tools that do several jobs really well, than one mediocre tool that does several jobs in a mediocre fashion.
2) get rid of the klutter, why on earth do we have a thousand different apps for only fractionally different tasks? surely a little amalgamation will make our lives infinitely simpler?
can we ever reconcile the differences outlined above. and if so, can it be done piecemeal, or will it require a new design philosphy to emerger from KDE?
Another usability discussion about KDE,and another bunch of clueless GNOME losers having a flamefest.. how “productive”. Pray tell, you “productive” GNOME-losers, why don’t you do something “productive” in your preferred environment rather than whine endlessly about KDE’s “usability issues”?
Yes, KDE has issues, but most things mentioned here is mostly superficial bullshit issues brought up by gnome lusers who haven’t used KDE for more than max 5 minutes, and then decided that they didn’t like it.
The the most frequent complaint seems to be “bloat”. What is this bloat thing exactly? (see the snapshot) here http://img292.imageshack.us/img292/4112/snapshot5vh.png
Here is the kmix, kopete (on line in irc, msn and icq), konsole and konqueror running, and as you can see in the dump, it’s using ~ 90MB RAM. I doubt you’d get much lower with the equivalents for that in GNOME..
The second most popular thing to whine about is “klutter”. Well, please point it out for me in my snapshot, or stop trolling. The *defaults* are a totally over the top, but seriously, you are really a dumb fuck if you throw KDE into the trash because of the defaults.. Besides I have a faint recollection of that the defaults were chosen with the intention that the distributors should clean up the mess before presenting it to the end-user. Unfortunately few of them does.
So, what does bother me then?
The default settings for the toolbars are too much and takes some work to clean up, and I haven’t found a good way to separate the main toolbars for the “webbrowsing” respectively the “file-managing” profiles for konqueror, which is a bit annoying since you don’t necessarily need the same stuff in those two profiles, even if some of it are common.
The second beef I have with KDE and it’s apps are bugs, some of them quite old it seems. Kopete seems to be particularly troublesome. Off the top of my head I can think of at least three things that IMO are broken. In no particularly order: Fake transparency for the chatwindow is broken, and apparently has been since 0.10.2 (I use 0.10.3), if you set “base font color” to something nonstandard it doesn’t affect what you are writing, only what others write, and finally it appears to crash on exit from time to time. The latest kmail seems to have quite a few bugs being fixed right now, and I have one system here where disabling icons on the desktop makes the right and middle click menus on it stop working after being logged in for a while. These things are more annoying IMO since they aren’t fixable for a non-programmer user.
Now, all this are just my experiences/opinions, but summing it up IMO the only thing of the frequent complaints that carries any weight is the default “klutteredness”, but hey, it’s not like those toolbars are carved in stone. If you don’t like them, do something about them, use gnome, or stick to a terminal, it’s your choice, but stop trolling about the “klutter”, since as I’ve shown you can easily strip it out.
Oh, I forgot to mention that the issue with KDE “bloatedness” has nothing to do with memory footprint. It’s the UI bloat that’s disgusting.
And the number one problem with your position is that good defaults do matter. Most users never change their defaults. How many Windows XP machines still have that wavy sand wallpaper after years of use? Nobody likes configuring things. Some computer enthusiasts have a fetish for configuring things, and they like to spend time removing buttons and toolbars from their various applications, especially if the weather is nice and there are women around.
There is nothing more valuable in software design than good, clean, sensible defaults. But it seems like KDE’s defaults are made to satisfy some power users at the expense of everyone else. I suppose you think the unique use of bright pink titlebars in the default configuration for FVWM2 is a good thing in the long run? After all, you can always edit the hex codes in .fvwm2rc, and then you’ll suddenly feel the irresistable urge to become a guru and write a 1474-line config file like this one:
http://dev.gentoo.org/~taviso/fvwm2rc.html
I’m trolling for the kure.
And the number one problem with your position is that good defaults do matter. Most users never change their defaults. How many Windows XP machines still have that wavy sand wallpaper after years of use? Nobody likes configuring things. Some computer enthusiasts have a fetish for configuring things, and they like to spend time removing buttons and toolbars from their various applications, especially if the weather is nice and there are women around.
There is nothing more valuable in software design than good, clean, sensible defaults.
Did you read what I wrote? I agreed that defaults matter, and that it takes a bit too long to configure it to some thing more useable. However, defaults are not *all* that matters, especially when they are as easy to change as in KDE.. Not like GNOME where you have to google all over the net if you are dissatisfied with the defaults, if they even are possible to change at all. *That’s* a real sinker. Furthermore I also wrote recollection of that the defaults were chosen with the intention that the distributors should clean up the mess before presenting it to the end-user. To put it more blunutly, if you are dissatisfied with the apperance, blame your distibutor. After all, these these popular GNOME based distributions certainly has applied a fair bit of spit ‘n polish to it, and there really is no law against them doing the same for KDE. Defaults are important, but being put off by them, and trashing the entire thing just because of them is just plain stupid.
Thanks for the link,looks nice when loaded.
“Gnome has a rather good ‘advanced’ mode”
It’s nothing advanced in gconf. It’s a silly unusable registry-like crap. Not intuitive, not easily accessible, not professional, not anything. Just a patch on a rather big hole. And quite a lousy patch that is. I’d rather take a text editor over gconf. Hell, I’d even take the “cluttered” and “bloated” kde control center any day over gconf.
And it’s also a lousy day when you consider that if you take gconf as the “advanced” mode, then what are you left with in “beginner” mode ? Right.
You’ve never browsed the .gconf directory. If you want you can edit the files in there with a text editor.
Again: The only thing registry like about gconf is the default UI used to edit it. If you don’t like that, there are other tools that expose more than the default UIs, but less than gconf.
You’ve never browsed the .gconf directory. If you want you can edit the files in there with a text editor.
Have you actually opend one of the files in the .gconf directory and tried to edit it? The files are in XML, which is no way very readable and actually making editing in a text editor even worse than with gconf.
Ah yes, that happened to me when moving from 3.1 to 3.2. They didn’t change the mailbox format, but they fixed some bugs in it so it more closely matched the spec
Very strange, never did that to me and my oldest mails on the system are from the 2.2 days (I lost my older mails due to a disk crash. I still have them on a backup CD but never felt the need to copy them back on hd =).
Are you sure that neither of you (a) could import the old mails into the new kmail and/or (b) used a KDE beta instead of a final. Beta versions of Kmail have been known to lose mails.
Very strange, never did that to me and my oldest mails on the system are from the 2.2 days (I lost my older mails due to a disk crash. I still have them on a backup CD but never felt the need to copy them back on hd =).
Are you sure that neither of you (a) could import the old mails into the new kmail and/or (b) used a KDE beta instead of a final. Beta versions of Kmail have been known to lose mails.
It happened when I moved from Mandrake 9.2 to 10.0. This was just my own discovery. I installed 10.0 onto a spare partition to test it, before putting it on my “live” partition. I copied my mail over, and kmail would read the indexes okay and display items in the inbox (subject, sender, ..), etc. But if I clicked on an email, the sender would change to “unknown”, and the subject and the message would be blank.
I traced this myself to some of my emails not conforming to the maildir spec. Since I have only ever used kmail, I assumed that somewhere along the line (i’ve been using kmail since KDE 2.2.2) kmail stuffed up and wrote the emails out wrong. Anyway, on KDE 3.1, kmail would read these broken emails fine. On KDE 3.2, it wouldn’t.
How would you interpret that?
It happened when I moved from Mandrake 9.2 to 10.0.
How would you interpret that?
The interpretion of that is rather easy. It’s not a KMail issue at all, it’s a well known Mandrake issue. When the maildir spec was finalized and KMail got support for it implemented, Mandrake choose to patch Kmail to not to follow it. They choose to use .mail as directory for mail rather than Mail(or mail) as the spec says(In my opinion much more correct, hiding the maildatabase rather than polute the home dir). Sadly the patches from Mandrake did create some migration issues not present in the unmodified KMail.
Now, unlike the other reply I don’t really know about the specific circumstances of this case but I’ve used Mandrake for a year or so and I have to agree that you should be careful before you blame problems like yours on the developers of the software. I liked Mandrake for many reasons but it’s tendency to lose data due to various Mandrake-specific patches were alarming.
I’m not saying that it couldn’t have been KDE’s fault just that that’s not necessarily the case =)
well.. you may be right.. perhaps set advanced mode as a command line argument, so… instead of startx, startx -advanced or kdm -advanced
anyway… you get the idea… whatever the command would be, this would make it so you have to boot to the command line and have a little knowlege of the command line to get advanced mode in kde
besides….. i think most geeks boot to CLI anyway instead directly into a gui… i know i do.
could also make advanced mode so insane that normal people see it and are like HOLY CRAP.. and immedately shut it off. heh, jk lol
another option would be to start controlcenter from a shell with an advanced argument kcontrol -advanced for instance
I have to agree that some apps have way too many functions in their toolbars: Konqueror is a good example. If Konqueror would cut a few rarely clicked things like “print,” “font adjustment(2),” and “security,” and the useless “up;” then maybe one could actually get the address and search bar on that same line! That gives more browsing space.
But on Quanta I think there should be an exception. I don’t use html editors, but I do some hand html. And if you want a visual way to edit html (not WYSIWYG) then you will need quick access to all the common tags.
That means work area includes many button functions.
You can get both toolbars in Konqueror on the same line, even with larger icons. Usually I just keep the Back, Forward, Up, Reload and Stop buttons, and then it fits. However, there are a few problems:
1. The “Up” button is useful in file manager mode, but rather useless when browsing the web. You can’t configure Konqueror so that it is only displayed when it is useful.
2. Whenever an URL gets too long, the address toolbar is displayed below the main toolbar instead of staying at the same place. That’s really annoying.
So I would suggest that to make KDE ultimately user-friendly (since almost everyone complains only about Konqueror), these three things should be done:
1. Remove all buttons except said ones in the default mode
2. Move the “Up” button to the File Manager KPart
3. Fix point 2 from above.
Konqueror would cut a few rarely clicked things
How do you rate rarely clicked things correctly? Your list are not correct according to my user habits. As I use both “font adjustment(2)” and “up” several times each day. Me and a great many others sees the one you labled as useless, the up button, as one of the features making Konqueror great. For me it’s the second most missed feature when I have to use inferior browsers.
I wrote a tool (in kdenonbeta) that lets you record a log of which interface elements are in use. If enough people want to run it in their konqueror for a while, we could actually get some real-world data on this. There’s some info about it in an old blog post I wrote at the time:
http://www.kdedevelopers.org/node/376
Rich.
I presume its because some pages have pretty small fonts? If thats the case you can set a minimal font size in settings.
I wish other more popular browsers had that…
what are your problems with gconf? every application has its section with all its settings and every setting has a description and a (localized) explanation of what it does and of which are the accepted values. there are .conf files really incomprehensible, gconf settings are almost always clear.
How much do you simply things so it would have an elegant design? For example nautilus doesn’t even have an invert selection. Nor something I guess considered somewhat advanced as a two pane browser mode. Does os X’s browser have two pane or is it considered too advanced an option that gets in the way of simplified design?
Then there are the non obvious things that make for a smoother experience. For ex. in konq, previews for pics only show in icon mode, not detailed mode, whereas they always show in naut even if all the icons are so tiny that the preview becomes useless and just adds overhead. However naut does remember if you set a particular folder (say for images) to icon mode whereas you might normally browse files in details mode. Btw, is there a way currently to get kde to preview video files?
I know this focuses on point 10 of the article which can be hotly debated. But I don’t mind for the most part kde’s approach. They just have to go all the way with some things. For example I wanted my bookmark toolbar folder to link to locations on my drive, but in web browsing mode show my frequently visted web sites. This should be saved separately with the profiles and not be shared as it is now.
talking about usability and configurability people who defend the “101 feature approach” call them self “power user”.
But what is a “power user”? For me a power user is a person who do more with his computer than the normal user which just surf the internet, write some mail, etc.
But doing more means for me, having a desktop that just works is much more important than for the normal user. The normal user maybe want to tweak the style of the system between surfing the internet and writing a letter, just because he doesn’t have to do much exciting things and if there is a periode were he don’t know what do do -> he will tweak the desktop and “play” with the system.
But a “power user” has so much exciting things to do, why should he waste his time with tweaking the desktop. I think a “power user” will prefere a simple desktop who doesn’t get in the way and let them concentrate on the exciting tasks.
I don’t see that a “power user” exploring all the day 101 buttons of a system and tweak it to death, that’s so boring and just something like “occupational therapy”. I think a “power user” is a user who make exciting things with his computer which the normal user does not.
I love sidebars. Please keep them. They encapsulate a whole lot of functionality without having to go to a menu.
Probably the best posting to read is from Scott Collins at Trolltech which addresses some misconceptions:
http://freestandards.org/pipermail/lsb-discuss/2004-November/002230…
It’s all fairly academic anyway. I love this though:
The component should have at least one compliant implementation available under an Open Source license that also promotes a “No Strings Attached” environment for developers. This means that the developer would be able to develop and deploy their software however they choose using at least one standard implementation. This is interpreted to mean that at least one implementation is available under a license that meets the Open Source Definition but is does not prohibit propriatry usage. The rationale for this criteria is very similar to that of the LGPL.
Note the this is interpreted to mean… bit bunged on at the end and how it doesn’t actually mean the first at all, but they’ve cocked it up. It talks about one implementation, which Qt quite clearly is because it’s dual-licensed. You get one implementation under mutliple licenses to allow you to do what you want with it. It’s very open to interpretation though, which many people seem to be doing a lot of.
If the LSB wants to create a commandment that says “Thou shalt not pay anything” then just spell it out, OK? That’s what it’s really about, but maybe that wouldn’t look right.
I just wonder what would happen if an ISV created a popular piece of proprietary software using GTK and other ‘LSB’ software and it became a de facto standard and everyone used it ;-).
What is confusing to me is that Qt is not all the way GPL.
If you start developing using the GPL version of Qt, you cannot then go and dual-license your application.
So by using the GPL edition of Qt, you are effectively signing away some of your code ownership rights to Trolltech. I say Trolltech because Trolltech is the only company that could give you an exception to this licensing term.
If Trolltech allowed developers to start with the GPL and then migrate to the commercial license when they wanted to produce closed-source apps, I think it would be an overall benefit to Trolltech for economic as well as legal reasons (reduce problems with Qt/TT licensing).
The other thing that is unclear is the role of “Qt Solutions”. These are important pieces of functionality, especially for using Qt for Windows development. As far as I know, “Qt Solutions” are not available except via the commercial license.
So we have Qt/TT’s GPL license as something that contains extra restrictions and covers a subset of the product.
What I don’t see is why Qt/TT would not go all the way GPL instead of part way. It seems they are most of the way there.
When you look at these issues, what are your thoughts?
What is confusing to me is that Qt is not all the way GPL.
Yes, it most certainly is.
If you start developing using the GPL version of Qt, you cannot then go and dual-license your application.
And if I start developing using the GPL licensed SomeRandomLib.so, I cannot just then go and dual-license my application either.
So by using the GPL edition of Qt, you are effectively signing away some of your code ownership rights to Trolltech.
Insofar as your code is derivative of somebody else’s code, it’s not entirely yours to do with however you please. Copyright law makes it thus, not the GPL, and certainly not Trolltech.
I say Trolltech because Trolltech is the only company that could give you an exception to this licensing term.
I say the copyright holder because the copyright holder is the only [party] that could give you an exception to this licensing term.
The other thing that is unclear is the role of “Qt Solutions”. These are important pieces of functionality, especially for using Qt for Windows development. As far as I know, “Qt Solutions” are not available except via the commercial license.
How dare Trolltech offer one product under the GPL while not offering another in the same exact fashion. Guess you’ll have to buy it then, eh?
So we have Qt/TT’s GPL license as something that contains extra restrictions
False
covers a subset of the product.
Which product? Perhaps the one Trolltech choose to release under the GPL?
Trolltech’s website says this:
If you are working in a commercial environment and are required to keep the source code of your product closed, you must use the commercial version instead. Trolltech’s commercial license terms do not allow you to start developing proprietary software using the Open Source edition.
and this:
If you are writing commercial/proprietary software, or if you plan to commercialize the software you are writing with Qt, you must use a commercial version of Qt.Please refer to the Open Source Downloads info page for reasons why.
From what I can see, Trolltech’s licensing will not allow you to start developing dual-license code with the Open Source Edition. You must use the commercial edition from the beginning otherwise you need some sort of special privilege from Trolltech.
Thus Trolltech’s “Open Source Edition” contains licensing terms that are beyond the scope of the GPL and are actually copyright/code ownership rights.
Also as the Open Source Edition does not support Visual Studio, it is not the same product as the Commercial Edition. It is all rather bizarre.
> What is confusing to me is that Qt is not all the way GPL.
It is all the way GPL. Period.
> If you start developing using the GPL version of Qt, you cannot then go and dual-license your application.
You can dual-license your application with any GPL-compatible licenses when using Qt Open Source Edition.
> If Trolltech allowed developers to start with the GPL and then migrate to the commercial license
That is a restriction of the commercial license, not the GPL.
> So we have Qt/TT’s GPL license as something that contains extra restrictions
Wrong.
> and covers a subset of the product.
The product is Qt. “Qt Solutions” are an additional offer/product.
I’m a long time OS9 user, also KDE and lately Gnome (for the aesthetics mostly). I’m starting to think seriously about going to WindowMaker. Don’t really need a task bar, a right click menu for launching would take less space and do just as well. Don’t really need the enormous memory consumption. And in getting ready for this I’ve discovered something for my less technical friends – all they really want is the same thing, maybe iconised to launch apps. The problem with KDE is it assumes you want to spend your time working on configuring your desktop environment. Most people don’t. The problem with both OSX and XP is that they think everyone should have the same desktop environment. Why? They don’t wear the same clothes. For me, I think its going to be WM: quick, simple, and out of the way of the applications. Well, I have also read good things about Ion, which may be WM but more so…
Trolltech’s licensing will not allow you to start developing dual-license code with the Open Source Edition.
If your code are dual licensed, it’s no longer purely GPL and you can not use a GPL’ed library like the Open Source Edition. It’s the way the GPL works.
Also as the Open Source Edition does not support Visual Studio, it is not the same product as the Commercial Edition.
Wrong! It’s exactly the same, expect it does not contain the makefiles(or whatever they are called in VS) for Visual Studio. It’s no problem using VS, but you have to do some work yourself. And if you run into problems you are on your own and get no help or support from TT. Besides there are bound to be licensing issues when using a GPL incompatible compiler like the on in VS. How compatible are the license of it’s standard C++ libs with the GPL.
Besides ask RMS what kind of tools you should prefer to use when developing Free SW.
Besides ask RMS what kind of tools you should prefer to use when developing Free SW.
Ask Richard M. Stalin what tools I should use to support the glorious revolution?
Haven’t we been through this before?
The GPL is just the classical ideology bait and switch. Sell it to the workers, get the workers to make it happen, and then betray the workers and thrown them in the pit forever.
In the world of Linux software, the giant mega corps already make the vast bulk of the profits. The writing is already on the wall but the workers are too dumbed down to notice. They will give their heart and soul for “one last chance” only to find that yes, they can still be sold farther down the river.
Poor fucks.
my too many cents of comment :
Yes KDE is in severe use of UI cleanup … after a while of using KDE you get used to the messyness … but compare it to GNOME ,OSX or even Windows Desktop .. KDE is simply a mess in comparision .
What has to happen in my opinion is UI cleanup & simplification ala GNOME (like hiding menu-entries & icons when they can not be used) but for Option-dialogs /Controlcenter to have an advanced tab .
A few basic simple settings to choose from & an advanced tab – that should satisfy both camps .
I also very much agree with Matthias Ettrich’s point about people trying to avoid doing things to windows .
The magic Alt-key is a great invention which makes things a lot easier.. especially if one has to deal with pop-ups or just wants to move window for merely a second too look @ for example a howto .
I think the Taskbar in KDE should give a lot more info on the Program/window it is showing .. dealing with the problem mentioned on here with soo many windows open & loosing where one can find a document in all of this .
Info might be of course the name of the documents/file .. last used .. maybe a preview of each window when the mouse is over the taskbar program entry .. as in Konqueror .
I think kompose … the KDE version of OSX expose should be a part of default KDE as it can solve the one-million windows open problem in my opinion … I love the inclusion of Superkaramba in KDE 3.5 … although its more eye-candy than productivity enhancement IMO … although Linux/Windows/OSX is heavily going to be an eye-candy competition with what the “features” of Windows Vista are .
One of the things I love about Windowmaker is the fact that the minimize button is at the top left .. meaning I dont have to be afraid of closing the application when Im only trying to minmize it as in KDE … another thing … one gets a small box with the applications icon in it for every application running under Windowmaker on your desktop … maybe something like this could be used under KDE for the taskbar .
Or maybe have no tasbar at all and just use kompose for “windowmanagement” which would give one the whole screen of space for extra info for each program running & have windows autoresize to fill the whole screen or if there are two windows on a desktop ( might not need virtual desktops any more with kompose for most things although there are exceptions ) they resize so its 50/50 for each .
Some very useful things have in my opinion been already inplemented in windowlab .
Or maybe somehow use the scroll-whell on most mice when over a windowborder to show a list of all windows availlable to choose from or something along that line .
KDE stability isnt the best either in my opinion.
Konqueror crashes on me sometimes .. especially since I have been using a tablet mouse .
When KWin crashes it doesnt restart which makes everything useless … in GNOME the windowmanager restart when it actually crashed (after running full-time for two days ) took something like 2 seconds .
I love the new “Get cool stuff” feature in KDE 3.4 .
Hopefully it will be extended to other parts of KDE .
Somehow the UI has to activly adapt more to the user … dont ask me how .. I havent got a clue ……
Post Scribbles :
Usability themes ? Very great Cool idea !!!
“There are other systems that use KDE. FreeBSD, OpenBSD, NetBSD, etc. ” – THX 4 reminding … I easily forget that there is an open-source world outwidth Linux .
re_re wrote :
“Oh yes, one more thing, I think kio-locate should be integrated into kde by default, instead of typing “locate:” there should be a small Locate bar (like a google search bar) integrated into konqueror.
I think this would be a very nice feature and with profiling support compiled into the kernel.. it is fast at finding files….. almost as fast as I can hit enter.”
-> agree certainly with that .
Just looked at simplekde.org … like what I see as long as the extra options are in an extra tab or so .
Could someone please get rid of the tree-view in filemanagers …. I really dislike it & that subtrees stay open after I have changed to a higher tree .. I allways have the urge to then carefully close the still open subtree … please ..
instead maybe have columns .
Matthias Ettrich said that he wants Konqueror to be a browser instead of a file-manager … I certainly disagree with that idea … there are already tons of standalone browsers out there & on Linux people use mostly Firefox as far as I can guess .
Konqueror would be the best filemanager out there with UI cleanup .. more stabilty …& maybe some UI features from midnight-commander .
“Gnome has a rather good ‘advanced’ mode…Its called Gconf.” -> Sorry but thats allmost like telling someone to play with the windows registry to change settings .
……..
I personally do not like KDE 3.4 as much as I did GNOME 2.10. I actually liked KDE 2.5 a lot better than 3.4 in Ubuntu when I installed it. I thought that Ubuntu and nearly all distributions I’ve used whilst
I personally do not like KDE 3.4 as much as I did GNOME 2.10. I actually liked KDE 2.5 a lot better than 3.4 in Ubuntu when I installed it. I thought that Ubuntu and nearly all distributions I’ve used whilst using KDE were quite slow and crashy/buggy.