Some basic thoughts about KDE 4

Published Wednesday August 3rd, 2005
104 Comments on Some basic thoughts about KDE 4
Posted in KDE

There has been a lot of focus on the desktop lately, and how the desktop should evolve. Rephrasing the questions: what can/should/might KDE 4 become? There are many good ideas out there, some of which I hope will get implemented. But there’s also things I miss in the current debate, things that deal with the fundamentals of desktop computing. What a splendid opportunity to write my first blog entry, I thought, and so I did.

Some background for this article: I had the opportunity to observe an untrained desktop computer user doing real work using KDE 3.4 and Open Office. I also had the possibility to make suggestions to the workflow. The task was straight forward: Someone was sending dozens of doc files as e-mail attachment, and expected translations back. As easy the task sounds, I was surprised about the challenges that desktop computer users do face. I believe I learned valuable lessons from combining that experience with prior observations, and I would like to share my thoughts with you. In case you are a KDE developer who hasn’t yet had a chance to spend some time watching non-techie users using computers, please try to. It’s worth it, and it’s fun, no matter what desktop system they are using, KDE, Windows or Mac OS X. Watching real users doing real work on real computers is likely the best way to gain the experience necessary to create the next generation desktop system. I don’t want to read a manual before using a mobile phone, or operating a DVD player, and I don’t have to. Sending an email, receiving some files, translating them and sending them back is in the same league of complexity, and it shouldn’t require extensive training, or the need to understand how the system works under the hood.

Window management

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. They sit in front of 20″ screens staring at a 1600×1200 background image and then typing commands in an 80×25 text window with a tiny font, having to scroll up and down all the time to read the output. It takes significant pain before someone actually uses the mouse to maximize the window vertically, or puts it to a more central spot on the screen. The same happens with other users. They go a long way scrolling around in a tiny file manager window before they might decide to make the window larger.

Lesson learned: the ideal window manager is the one that you don’t use at all. This needs support from the applications. Applications don’t normally start up with a sane size. Maximizing isn’t always the solution, especially because the most often used applications don’t work very well when being maximized, among them web browsers, consoles, plain text editors (think an email composer). Every application that does line breaks based on the window width has this problem.

Another problem with window management today are secondary windows, also called dialogs. In KDE, we try hard to bind those closer to the application main window. For example, a dialog doesn’t have its own entry in the taskbar; when you raise a task, its dialogs are kept above the application main window; when you minimize a task, you also minimize its dialogs. But the connection isn’t perfect. While dialogs are placed in the middle of their main windows, they don’t adjust position and size when you resize or move the main window. And when you try to activate a modally shadowed main window, the only visual clue you get is that nothing happens. Ideally, dialogs and other stand-alone secondary toplevel windows should go away for standard end-user applications. Having things like Apple’s sheets and drawers that naturally extend and modify the main window itself feels more natural, more elegant and avoids confusion.


What I said about window management also applies to configuration. Nobody likes to configure anything. Sometimes people like to customize something, but that’s different. Customization means giving the computer a personal touch, with colors, images, or sounds. Changing the computer’s behaviour is entirely different. Configurability is costly, because it’s code that must be maintained, and it’s getting harder and harder to test all combinations. Even KDE’s own file dialog doesn’t work properly with all the 6 different completion modes that KDE offers for combo boxes. In DropDownList mode, you can no longer navigate with the keyboard to sub directories. Likely an easy bug to fix, but that’s not the point. The point is that with all that configurability, even a community that diverse and large as KDE’s cannot ensure basic functionality working in all modes. KDE 4 is our chance to get rid of all that clutter, I hope we make use of it. “Those who write the code, decide” is a good rule, but it only works if there is a strong maintainer or group of maintainers with a clear and aligned vision. Many parts of the KDE had this, others didn’t. And that’s where we see the issues today.

File manager

Konqueror today is a swiss army knife. That’s cool, people like swiss army knifes. I used to have one, too, and I loved the engineering behind it. Putting so much functionality in one single device is admirable. But I never used it unless I absolutely had to. At the campfire I had to, and I was happy to have it. At home or in my kitchen, I don’t have to. Instead I use whatever knife, screwdriver, or other tool that fits best for the job.

A document manager is not a webbrowser is not a text editor is not an image viewer is not a console is not a trash bin. Profiles are a way around it, but not the solution. Profiles are just like opened swiss army knifes. It’s like buying 4 swiss army knifes and having one lying around as a knife, one as a saw, one as a screwdriver, and one as scissors. Whatever you do to them, they are still swiss army knifes. They can magically convert into something else, and they will if you do something to them.

I get the feeling we do embedding and morphing not because its useful, but because we can. And we are ignoring its downsides. And the biggest downside in my opinion is that it takes mindshare away from dedicated applications which would do the job better.

A little example: you have a browser window that covers 40% of your screen, it contains pictures that you have taken over the weekend. You want to show them to someone else. In KDE today you would probably find an empty spot on the icon view, and then select Preview_In/Photobook. Or you are one of the few chosen ones that know the icon on the toolbar. What you then get is a fairly good embedded image viewer, but it’s still only an embedded image viewer. And it’s only using 40% of your screen. You can then maximize the window, or – if you are an expert power user – select Settings/Full_Screen_Mode. I found it after looking under “Window” and “View” first, by accident. But even then you don’t see any image in full size, you see Konqueror in full size. You still see the side bar, the navigation bar, and a tool bar. 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.

Konqueror’s future is being a dedicated web browser. What KDE needs in addition is a dedicated document manager. Great preview capabilities, yes, embedding, no. Konqueror is not a document manager, it’s a generic tree structure browser. And if all you have is a tree structure, everything looks like a kio-slave. Our trash bin in KDE 3.4 is a nice example for that rule: it’s a full blown browser window that uses the trash:// protocol. And the one main operation that you want to do with a trash bin (emptying it) is only available in the context menu of the icon, or the free space in between the items. Who is supposed to find that?!

I’ve been discussing design ideas with Zack and Simon, and hopefully we are able to present a prototype of how a modern file manager for KDE 4 could look like. And if that’s done properly, the desktop itself should probably be such a view. Today the desktop is a set of links that show huge tooltips with interesting contents like “myComputer.desktop is a Desktop Config File that is 170B large, owned by ettrich – users with -rw-r-r– and points to media://”. Thanks for letting me know…

A nifty feature I’d like to see in a file manager is that it shows me when I’m visiting a document currently. 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’m looking forward to aKademy, to heated discussions, and even hotter coding sessions!

Do you like this? Share it
Share on LinkedInGoogle+Share on FacebookTweet about this on Twitter

Posted in KDE


Tomasz Torcz says:

It’s funny, your thoughts sound exactly like motives behind GNOME 2.x. Simple but powerful, not overbloated, easy to use single purpose apps. I suggest reading some Havoc Pennington manifestos, they could be pretty insightful at this stage od planning.

unlimited says:

“getting rid of clutter”
Does that mean cutting functionality? That scares me. Why can’t we leave Gnome that kind of stuff? xD

Also I personally like the way I can browse my files and surf the web at the same time in one konqueror window with multiple tabs.

KDE developers, please don’t try to attract new users at the cost of existing ones.

Of course, a simplified default configuration would be perfectly fine with me, as long as I can still configure KDE the way that I like. πŸ™‚

Julio Gazquez says:

With all the respect you deserve as KDE (and LyX!) creator and developer, I feel your rant is too much a personal opinion. Year after year KDE has been about listening their users, not having a vision to shape KDE around. That is Gnome 1.4+ approach, and I guess that KDE users aren’t Gnome users because they don’t like Gnome. I know, there are a lot of developers and there will be a debate, however I don’t feel this “KDE should be this way” rant as positive.
Lot of people, including myself, are devotes of KDE because how KDE is. I love Konqueror as it is, maybe it can be better, but I like the paradigm, I want to browse directories, then do ftping with it, then check a man page or a web page and so on. Maybe you can enhance the interface, maybe you can have a generic browser with less capabilities for casual web browsing and a full featured web browser for more intense use (as kghostview and kpdf exists today independently of konqueror), but trying to go against what KDE users like and expect because of what non KDE users could want.

christian says:

Reading your other comments/flack here about your ideas being rather gnomish I feel you deserve kudos.
I mostly agree with you about the swiss army knife that konqueror has become. Don’t get me wrong, I love visiting some poorly labeled link in slashdot that is actually a pdf and having things just work. I also love the multi tabs and multi window features enabling konqueror to exist as a smb -> ftp -> fish transfer tool.
However, I have also seen in others and had problems myself with konqueror acting as a trash can. I think you had a nice summary of issues you see, and I am very excited as a user to see where these ideas lead kde. In short just because we don’t like where gnome went with its simple approaches does not mean their ideas were totally flawed. Thanks for dreaming and most importantly, designing, for the future.

Zack Cerza says:

Wow. I’m glad I’m not the only one that feels this way about KDE, and about Konqueror. Every time I sit down at a fresh KDE installation, I have to spend a total of over an hour configuring and customizing it to make sense to me. After that, things are decent … with some exceptions, a few of which I’ll whine about now.

I recently started using Kyzis as an editor. Sure it’s weird, but it’s my preference. Making it the default editor for all text/* MIME types entailed well over a *hundred* clicks – select the file type, click “Move Up” about 7 times for each, repeat until carpal tunnel sets in. And when I remembered that KDE has an “embedded text editor service” I trolled through the control center until I found a thing called the “component chooser” which let me tell KDE to use Kyzis. And, hmm – after all that, I’m typing this comment in what looks like an embedded version of KEdit.

Speaking of the file associations editor, I’m pretty convinced that if it still exists in KDE 4 (which maybe it shouldn’t), raw MIME type names shouldn’t be the main descriptors for file types.

Most KParts, when embedded in Konqueror, are read-only. I can’t think of a good reason. Hmm – maybe you’re right. Maybe they should go away entirely.

At every release there is another application that feels that it is important enough to live in my notification area, and that by closing the window I actually mean “Turn into a very small icon, hide in the corner of my screen, and pop up a dialog box telling me you did so. But don’t offer to actually quit.”

Unlike some others, I don’t pretend to represent all, or the majority, or even 1% of KDE users. This is just one dude’s opinion. There’s probably a reason I don’t make any decisions in the KDE project. Oh, and your suggestion to watch *actual nontechnical users* while they work just made my day.

Inorog says:

One single small nag: you state the window manager does wrong but you don’t merely start to say how this can be fixed. And the problem is… I can’t imagine how either. Perhaps because it’s difficult. If you asked me, all apps should cover all screen by default and navigations should be done by browsing a list of the stacked apps. At worse, tiling. You mention Oberon, it’s an excellent example. That’s applicable to dialogs too. Modal dialogs should simply replace the current app display (you can’t use _that_ anyways). Non-modal dialogs shouldn’t exist. Drag’n’drop is a horrible anti-invention. But guess what? I feel that this kind of changes would be perceived simply too radical by most users.

Well, too bad I can’t be in Malaga. Too bad only for me. Thankfully, KDE lives very well without me. Thanks to unresting thinking, like yours. Great job!

mxcl says:

Julio, it may surprise you, but not everyone agrees with your perspective.

I find Mattias’s views to be based on sound experience and good sense. I can’t say I disagree with anything he has written above.

I don’t have so much faith in KDE 4 removing configuration clutter. I’m not even sure if the API will get a much needed cleaning job. But I do think, at the very least, the UI of the desktop, browser and file manager will be rennovated. I look forward to aKademy so I can get more of a feel for what might change, and what I can help with.

Scytale says:

About the modal dialogs and missing visual feedback: What if the main application did just gray out when displaying a modal dialog?

About limiting Konqueror to be just a web browser: _Hell_ _no_! It’s nearly perfect the way it is. I don’t want to be using application A for browsing, application B for file management, application C for PDF viewing and application D for SFTP transfers. That’s the Windows way, and IMHO this difference is one of the biggest advantages of KDE. The overall integration and embedding is what I love so much. Just yesterday I found a funny picture on the net, chose “copy picture” (note how it doesn’t say “copy picture _address_”, even though it does exactly that), chose “new message” in Kontact, clicked the paperclip, did a middle-click inside the “file” textbox in the “open” dialog, watched the URL of the picture appear, clicked “open” and was happy how the picture automatically got downloaded and attached to my e-mail. That’s how it should be. I don’t want to save that file somewhere, then attach it, then delete it again.

This is a warning: If you cut any integration or functionality from any of KDE’s applications, you will lose users, me being the first. See, for example, I need a new notebook. Those Apple iBooks look quite neat, expose is fun and stuff. But, what keeps me from buying one, is that I don’t think I could run KDE there as easy as on an x86 machine. I can’t work productively without KDE any more, and if you ask me why I love it so much, the first thing I say is “integration”.

So please, if you really really need to make KDE easier for Windows users, then do so, but _optionally_. Don’t try to remove any single “power user” feature. That would be very sad.

I’m not a KDE developer, I don’t even speak C++, but this post made me think about attending aKademy, just to discuss with people and keep KDE from drifting away from the power it has now.

Vide says:

I think you perfectly hit the nail about Konqueror. Konqueror is a swiss knife and we all do love it, but it to be even better it has to be a swiss knife without the “normal user” even know it. So, different and specialized UIs for different tasks (as you said, for example, trash bin, image gallery etc.) but with the possibility to use the power of IOSlaves in a basic Konqueror window.
Another example of “clutter” is the RMB menu in the sidebars…it’s plenty of options that do not fit with the current context (for example: why should I compress an application entry??!? or who needs the size of the print:/ folder?)

I really do hope your vision will come true, and no, it has nothing to do with the Gnome vision as someone said.

Vide says:

Scytale, I don’t see a “let limitate the power of KDE” in what ettrich is saying, but more a: “let use the power we have in a more intelligent way”. Integration is the key, everyone know it, but integration with a tons of non-related-to-context options in your face is not a good thing (and this is what is happening in some cases with Konqueror)

Pinaraf says:

Don’t kill KDE with KDE 4 : I really don’t want konqueror to be downgraded to a nautilus-like application nor a firefox-like application !
Konqueror is great for power users, and I didn’t see any non-power user lost in konqueror… My brother really likes konqueror because it can embed a media player, it can show archives…

Simon Farnsworth says:

Regarding the issue with modal dialogs not moving or changing when the application does; how hard would it be to make them relative to the owning application, not absolutely positioned?

Vide says:

Pinaraf and all the others: tell me, for example, what do you really do with all the Konqueror options while in the trash bin (trash:/). What parts of this “swiss knife” do you use in this context?

Moreover, the slideshow example from ettrich is almost perfect: I really do think “whoah, Konqueror should act like this, it’s the more natural manner” when seeing XP in action. Why do you need to see the toolbar, the sidebar etc etc when all you want is an automatic slideshow to show your holydays photos to your relateds and friends? Ettrich got very serious points, and all your cryings “do not dumb down KDE” doesn’t really have any clue. It’s about making Konqui and KDE smarter, not dumber.

fromoze says:

Just my 2 cents about clicks:

May be you’re right about the problem about the problem of opening apps. But I personally hate to use 2clicks for browsing and opening apps (not files). I will be really happy having a mix option: 1 click for apps and browsing and 2 clicks fot files. Do you?

Andre says:

Don’t we have all applications which get embedded in konqueror also as standalone apps? So what we really have isnt a swiss army knife which is everything but nothing good, but a real knife, a real saw, a real screwdriver, and real scissors, which can be assembled to be an all around handy swiss army knife.
But i agree that the menus (especially file context menus) should mutate into something sensible for the actual file type. Having the option to view a slideshow in an external imageviewer doesnt exclude the possibility to view it embedded. You just have to present both options in a sensible way. In the external slideshow there could be a button to directly jump back to konqueror, so that the user doesnt loose the link to his starting point. Konqueror has service menus ( but creating those manually is too much pain for users.
I like to use filebrowsers like Krusader (try it, its really great.). I can use such a profile for konqueror, but it isnt the same. And the image example is quite a good one for this case, too. For comfortably viewing images embedded, the panels are simply too small. Therefor im using Krusader for file management, and konqueror for web browsing.

Jakub says:

Looks like the single point I can agree with is this about modal dialog: when user tries to do something with shadowed main window, modal dialog should be raised and activated. Simple, and no more ‘that damn app is frozen’ comments.
Double click: hell no! I don’t see any point in just activating document for sake of activating. Either you want to open it (so you just click), open context menu (then right click – still no problem) or move.
Configurability: basically you are saying that whole components/functions should be ripped out instead of fixing bugs in them. Not a good way to move forward I think.
Konqueror: embedding _is_ useful. This prevents you from having tons of windows cluttering your screen.
Building document viewer without embedding capabilities is just crazy – how do you expect it to handle all those formats? Stick everything and a kitchen sink into viewer itself?
What konqueror needs is good detection of directory contents (does it contain pictures? music? or maybe cvs checkout?) and automatic switching to right view. Definitely not dumbing it down.
Your example with pictures also seems wrong to me: if user cannot find toolbar button to switch into photobook mode how is he supposed to find a button to enable slideshow? What is the great difference. That, and I really don’t want a slideshow – I want a sidebar with thumbnails so I don’t have to go over all pictures just to show several interesting ones.
About fullscreen: you are right, it is not konqueror windows with all toolbars etc. that should be shown fullscreen but only active view.

Ulrich Miller says:

Hi guys,
first of all a word of introduction: I am a big KDE fan but currently not involved in development, which perhaps might change in future if I find more time.
From a user point of perspective I can say that I would really appreciate, if some of the ideas of Matthias would find its way into KDE. Don’t make the mistake and try to use the overload of functionality and configurability as an attribute which is needed to differenciate between KDE and GNOME (there are better ones, e. g. the tight intergration, the brilliant qt library, the marvellous applications…).
Whenever I talk to end users the main critics about KDE is the overload of configurability and functionality.


ps: thanks to all for the great work.

Henk Poley says:

I do not know where to find these slideshows Vide is talking about, but here are my 2 euro cents:

The way forward with this is to make individual wrapper applications around the kparts that konqueror is using now. Please don’t touch Konqueror unless it makes things more clear in Konqueror or removes ‘dumb’ stuff (size of virtual folders or something).

People will start to use these new ‘programs’ if they find them usefull. And in the mean time you haven’t thrown anything away that apparently lots of people (inclusing me ;-)) like to use.

btw, about the filetype preferences. There really should be something that can globably override, or change settings on particular MIME types.

seezer says:

Vide, he definatly talks about konqueror and its status of a swiss-knife. and i don’t think he is that big fan of swiss knifes πŸ™‚
And those who compare some of ettrich’s ideas to those of gnome aren’t that far either, in my opinion:
He says configurability is bad. But that was the one and only big reason i turned to a kde fellow instead of a gnome guy. In Gnome you can somehow feel, that the devs think they know what’s right for you. And you just feel this when you hit things where you got another opinion of ‘right’ -> “come on guys, why can’t i change this?.. damn you!”
I think (again, only my point of view), configurability is great and attractive to users.
Whenever you think “hey, can’t this..” and you go to control center and just change it to what you want, you get another “damn. i took the right desktop” feeling.
Customization of some backgrounds or windowdecorations and crap is what those Windows guys are happy about. And nearly the only thing they can do (without using big shellreplacements etc). But we (the KDE users) can customize and configure quite many aspects, and thats just what we (or I) want.
For those kind of computer users, ettrich is talking about, there should be good (perhaps better) default configurations, sure. But those users won’t even try to click through control center and search how they can change to sloppy focus or should they care?

unlimited says:

One thing is convenience. Currently I can just click on the button “photobook” in konqueror while in a folder with images and I get an embedded slideshow. Then I can just click an the “back” button on my mouse to quit, or I can open a new tab to browse my files or to surf the web and research stuff.

As you can see, the current situation isn’t all disadvantageous. That’s why it’s important if the default behaviour gets changed, not to remove the “old” features. Because if you do otherwise, you are going the way of the dodo… eh I mean Gnome. πŸ™‚

Jakub says:

About windows and modal dialogs: simple (?) way to fix this problem would be raising and activating the dialog if user tries to activate mainwindow that can’t be activated because of modal dialog presence.
About single click: well, I disagree completely. I never had any problem with accidential launching something when trying to move it. After all it is not so simple – to launch you have to release button before moving an item. But I have seen several people accustomed to double click and launching something two times by accident. Simple solution would be disactivating item for short time (100ms?) after a click.
About configurability: basically you say it is better to throw out function/component instead of fixing bugs. Not a good way to move forward I think.
Konqueror: I really like embedded view – much better than lots of windows cluttering my screen. Your picture album example does not like believable to me: if user is unable click toolbar button to change into view suitable for viewing pictures how can you expect him to click button enabling slideshow? I really see no difference.
I agree that Konqueror can be made better – but not by dumbing it down but by better way of detecting directory contents (picture album? music? svn checkout?) and changing to sensible view mode.
Profiles _do_ make sense: they are just not used enough for now.

Michael Thaler says:

First, I think people often confuse usability and familiarity. Lots of people are familiar with Windows. They know how to do things in Windows because they worked with it for years and learned how to use it. If these people try to do the same things with KDE, they often do not succeed because in KDE things often work slightly different – a priori this have nothing to do with usability.

Second I do not agree with the examples you mentioned. When I bought my first cellphone two years ago, I instantly knew how to call other people. But I knew this because I was familiar with phones before, I learned how to use them. But to use more of its functionality, I had to read the manual. When I bought my first camera, I also knew how to take pictures, because I used a camera before. But to understand the more advanced functionality, I had to read the manual, too. We are familiar with things like phones or cameras, that is the reason why we can use at least the basic functionality without referring to a manual. But for the advanced functionality we still need to read the manuals. Imagine someone from some isle, who had never used a (cell)phone or a camera before. Do you really think he could use these tools without learning first how to use them? I am pretty sure, he could not.

I just don’t think that it is possible to use something as complicated as a computer without first learning how to use it. And it does not make any sense. And simplifying computers and operating systems in a way that people can use them without knowing anything about how computers and software work, does not make sense.

That said, I totally agree that a browser, a file manager and a document viewer should be different applications. I would definitely like to see a really polished browser in KDE4 and a really polished file manager and also a really polished document viewer. These applications could still be kparts for those who really like embedding them, but the default should be different applications. I think this is also better for new users, because various applications with simple interfaces that are only intended to do one thing are certainly easier to learn and understand then one big monster application that is capable of doing a lot of things.

Gogs says:

How about removing the nineteen million toolbars in Konqueror (e.g. icon view toolbar, list view toolbar and so on).

Me personally, I don’t like the icon size increase/decrease icons, so I customise Konq’s toolbar and remove them. Then, I change from say icon view to list view – and voila! the increase/decrease icons are back. I then customise toolbars (again!!) and find that there’s a separate ‘list view’ toolbar. So…. I remove the icons from list view. Later on, maybe I change to ‘list with details (or something, can’t remember what exactly)’ and what happens??

Yep, those F***ING icons are back AGAIN. And so on and so on.


fp26 says:

I think some solutions exists.

As for the RMB menu, I think the best way is configuration.
Create a configuration entry that says:
add/remove such and such entry when in “given context”
such that:

“Empty trash” is “added” when all selection list items or url are “^trash://”
“Slide show” is “added” when some selection list items or some files in current url are “png|gif|jpe?g|bmp|xpm|…” or maybe “$(picture_files)”
“Size” is “removed” when all selection list items or url are “^print://”

Some combobox, boolean and some regexp textfield.

But you also need rules when multiple items are selected,
what happens if you select Trash and Home and you press the RMB?
Does “Empty trash” should be there or not… I think it shouldn’t.

So we need a lower bound of allowed actions.

“Why should I compress an application entry??!?”
That sounds reasonable for one entry,
but what happens if you select a bunch of them
and some picture, should you be able to compress all of them?

Also it is possible that you want to compress many application entry
to send to someone else or simply to backup your “current settings”,
because you’re doing some weird stuff in there.

So limitation, yes. But you have to think “at large”.
An options should only be removed
if “it makes no sense at all in every possible scenario”
and that you can “guarantee that”… πŸ˜›
else leave it there.

The size of “print://”, I think makes no sense,
since it’s a KIOslave, not actual files…
unless you some ‘real’ config files exist in there,
which can be copied around, compress, etc.
and in this case, should those files be copied from “there”,
or should only be copied/moved/compress from ~/.kde/…..

As for default application size, I think the best is to have
a “minimum setting” plus a “normal setting”
as a ratio of current desktop size.

For instance: 700×400 or 87% desktop width by 66% desktop height which ever is higher. These should be configurable also.

Another way, is settings for a bunch of desktop size and a ratio setting.
So you have a default settings for:
640×480, 800×600, 1024×768, 1152×864, 1280×1024, 1600×1200
and any others are ratio based.

So, it’s not about really removing features,
it’s about removing features that “cannot be used” basically. =P
Also, 90% of frequently used feature should “directly accessible”,
while the remaining 10% can be moved to an advanced sub-menu…

In other words, if you just happen to need that “10% use case”,
you don’t want to be in a Gnomish situation where it can’t be done,
you want to be able to do it…”as easy as before”,
but maybe just one click-away; however,
it shouldn’t be in the way of common task.
In others words, I might do it just once a month or once a year,
but don’t do like WinXP and “removed the feature”,
because when I need it, I REALLY NEED IT NOW!!! =P

Hope you guys see the idea! πŸ˜‰

Just my two cents…

sns says:

This page renders incorrectly in IE, Firefox and Opera. The main column backgound stops half way across the righthand sidebar column….

justto letyouknow says:

The background to the main column of this blog stops halfway across the righthand sidebar when rendered in ie, firefox or opera.

Brian says:

>Unless I’m the only one who constantly starts
>applications by accident when I want to move files
>around, or delete them.

I though I was the only one!

Fls says:

“Cut functionality”, doesn’t sound bad, but the question remains: which functionality?

I have a methaphor for that. I need to tidy up my bedroom, but I’m not sure what stuff should be removed. The one keyrule I use is: “Which of this stuff have I been using during the last year?” If the answer is negative, I throw it away(okay, exceptions exist…).

I don’t think this would be a bad criteria to decide which configuration-options should still exist. Further, should there still be to many options, one could decide to add a simple and an advanced mode to “kcontrol”.

The Badger says:

Jakub writes about single-clicking and not having a problem with it, and I too prefer it over double-clicking any day; indeed, double clicking, aside from bringing on all sorts of muscle damage, also incurs yet more configuration cost in the form of yet more dialogues for tuning the double-click delay, click distance, and so on. (And the double vs. triple-click behaviour for word and line selection that most toolkits like to support often ends in irritating charades involving the wrong thing getting selected nanoseconds after you stop clicking, having seen the right thing momentarily become selected.)

What is disturbing with Konqueror’s navigation is the way simple file manipulation operations are turned into some kind of knife-edge sport where folders are opened for you and then slammed shut spontaneously before you realise that all those files you were dragging (incurring yet more muscle damage) have now started to get copied into the wrong directory. No wonder the cut/copy/paste operations were introduced to Windows, KDE and other environments!

Meanwhile, if you’re looking to simplify menu clutter, I’d advise that you take a long hard look at Eclipse and then try your best to not emulate the mess it makes of dynamic menu configuration. Konqueror only sometimes replicates some of the bizarre issues – eg. duplicated menus – that Eclipse seemingly exhibits routinely.

Mitchell Mebane says:

How about moving tab controls, pane splitting controls, etc into the window manager?

ac says:

For what it’s worth, Matthias, I think your observations are dead-on correct. We’ll see how the solutions work out, though–it’s always harder to fix a problem properly than to identify it.

Window management/dialogs are a great example. I definitely see the same problems as you, but I’m not sure solutions exist which are better that the way things are right now. Sheets and drawers are definitely a different way to approach this, but I’m not sure it’s better. Ultimately the developers make this call though πŸ˜‰

Decoupling the browser and file manager–I couldn’t agree more! The first thing I always do at a new KDE installation is disable the “embedded” viewer for all filetypes–because dedicated viewers for local files are always better. I agree that Konqui is stealing away the opportunity for some of these dedicated applications to show what they can do. Embedded viewers make a lot of sense in a web browser, but for filemanagers, I think embedded viewers are the exception, rather than the rule. “Show slide show” is a great example of how to handle these exceptions.

Your observations about the disconnect between the trash:// URL and actual user needs, and desktop tooltip overload are also quite accurate. There seem to be some obvious solutions there.

Keep up the good work. I hope some of these ideas make it into the final product.

ta says:

Couple of comments:
1) If I want to see something (pdf, image, whatever) embedded, I left click in Konqueror, and when I need a separate window (i.e. the file I want to see is that important) then I middle click to open the respective program. This is a simple problem -to me at least-.
2) Saving attachments, translating and sending them is an interesting problem.
* Currently, what you do is: Create a new folder, save them by one by one -if you do not know right click->save attachments option- then open/switch to file manager, find the folder you have saved and start working on them with the respective wordprocessor.
* (Pure speculation:) A better way might be
a) Create an attachement section with a separator bar for e-mails in Kmail (these days attachments are integral part of e-mail unlike a decade ago) put save all attachments button to the separator bar as well as view and save selected attachement buttons. Advantage: no right click menu to discover/browse, everything immediately in front of you.
b) Automatic Folder Management:
When you save attachements/files of an application a virtual folder/link is generated with attributes: Kmail, document types, date, user name, document meta infos AND user given folder name
A new mode for Konqueror where last saved documents/last run applications and their last accessed documents are shown as well as additional search functionality is added (I haven’t used but this is similar to new desktop search of MacOS X, I guess).
Then, those translation documents are immediately accessible just with a refresh of Konqueror and you can start working on them.

A side idea for search: There is finite amount of info in a HDD, so a tree like visual search can be implemented.
Multimedia->Audio->Type or Date or User Keyword
Generating Application (s)->Date or Metadata
User Keyword (Folder name) -> Type-> Application etc.
* I want to find pdf,latex documents related to project X sorted by date
* I want to find downloaded pdf files (download flag) on subject Y (folder name/user keyword).

Google style: write keywords to search box
Visual style: Left side tree of Konqueror but multiple tabs for different possible root attributes, search continues by clicking on the attribute
(Advantage: similarity to folder search)

Anyway, just a few thoughts…
Thanks to everyone contributing to KDE. It working with computers a pleasure as it is. I am sure it will get even better…

You make some very good points, but on the issue of Konqueror I think you have it exactly backwards.

Konqueror is a much better file manager than web browser. In part that’s due to the limitations of KHTML, but more because read-write KIOSlaves are so amazingly cool. Easily the #1 thing I love about KDE (and there are many) is that the exact same program and window is/can be used for local files, FTP, SFTP, SMB, etc. Managing files across the network is insanely transparent. Then add in the compression IOSlaves (tar:/), and you have everything you need for “file manipulation and organization” in one place with one, very rich interface. That’s very very good!

If anything, I’d rather see the web browser split off from the file manager into a separate application. Insert various calls for KFireFox here, which no doubt will annoy KHTML developers and for that I am sorry. But the only reason I use Konqueror as a web browser is for the better integration into the rest of the system (handling of mailto:, for instance). If Firefox had that level of integration out of the box, I’d drop Konqueror as a browser in a heart-beat. That level of integration doesn’t require that the browser BE the file manager, just that the browser speak KIO, which is one of the two key components that makes KDE so awesome in the first place. (KParts being the other one.)

Konqueror of the 3.x series is a showcase of KParts and KIO. That’s fine. Its next evolution, I agree, is to split into logically-grouped applications tailored for different activities that require different interfaces and logic. I just see “file futzing” as the main one for “Konqueror” to retain, while web browser gets split one way and, perhaps, photobook gets split another.

Alex says:

This blog post matches my thoughts exactly. After watching many new Linux users struggle with KDE, I have learned that the “Swiss army knife” approach KDE takes with Konqueror and other programs just doesn’t work for usability, at least not initially. It’s not just the “jack of all trades, master of none” problem, the biggest problem is that people don’t expect it and don’t know where to search in a program. When you have one program that does everything, all the tasks become more obscure and hand holding more difficult. For example, let’s say I want to burn a CD, if Konqueror were to do this, it surely would be a KIO-slave from some obscure menu. However if something like K3b was to do it, as soon as the user opens the program he is asked what kind of CD, he has a dedicated interface, etc. He knows what to do and that he is in the right place to do it.

In addition, configurability is a mixed blessing. Users want sane defaults and hardly mess with those options. This means that a lot of resources are set aside just to please a minority of tweakers. This sacrifices stability, ease of use and homogenity (so that skills ar emore transferrable and things can be expected to work correctly on any KDE desktop) The level of configurability of KDE can best be described as “bloat” for the majority of users. GNOME has done a good thing in this regard.

Also, don’t forget that most users come from Windows, so try not to shock them with a spatial file manager or inverted “yes” or “no” buttons. ;-p. But do take some ques from GNOME 2.12 they have done fantastic work in the area of usability for the most part.


james says:

You are quite right of course, but good luck convincing the KDE community. For some reason KDE has attracted tons of people who think that more options is better. Their logic is if there are more options then more people can be satisfied because it works just how they like it. They fail to realize that most people would be satisfied better by simply choosing the best option and removing the rest. This is further complicated by the fact that removing options makes users mad. Once a feature has been added, removing it will always upset somebody, even if it was a stupid option. That’s why I think that KDE can never become the simple, usable interface we all really want. Instead, the only way to achieve true usability is to fork KDE and start removing options like mad. Call it “KDE lite”. Users of the old KDE will be happy with all their useless options, and sensible people can move to the new KDE (though they will always be asking for their “one favorite option” and they must be ignored for KDE lite to succeed). This is a radical step which would unfortunately fragment the developer community, but I think a radical step is necessary if real usability is to be achieved in KDE. Otherwise we will simply continue to be drowned in options.

Chris Parker says:

I think that the biggest problem with GNOME 2.0 was not the new “simple” interface, but with the overall lack of functionality. When the complexity was taken out of GNOME, it became very apparent that GNOME was not integrated in the least. Sure, there was a cool underlying object system, both none of the applications really shared components with each application living in its own little sandbox. What is a GNOME app? Gtk and GObject, and maybe GConf.

KDE has a great infrastructure. You remove the clutter and what you still have a great DE.

Stefan says:

Someone complained about the file associations and I agree. They’re a PITA, actually it’s even worse they’re a disaster.

Recently I installed SuSE 9.3 (wanted to try it again, the last SuSE I’ve used was 6.3) and just removing Realplayer as the standard player for mp3s and oggs (wtf were they thinking?!) took forever. You have to remove it from and x- and to make that permanent you have to become root and edit system wide files because KDE ignores your settings as long as those files are present. Also the “open in embedded viewer” gets ignored unless you change it for all, the group and all mimetypes that are somewhat similar to the one you want to edit.

Now once that works we can talk about other things. Dedicated applications are fine but I don’t see why that influences Konqueror. Just make the seperate applications the default and Konqueror an alternative instead of the other way around. I love gwenview, I use it regularily but I also use the gwenview kpart it depends on the situation. Why can’t there be a simple kfm, a trash viewer, a document viewer etc. but with the necessary kparts for konqueror.
That way unexperienced users and the usability-freaks can use Kbrowser, Kfm, KControl-for-dummies and Kviewer while those who like KDE the way it is now can use Konqueror and KControl-would-you-like-fries-with-your-kitchen-sink. With proper kparts the duplication of code can be kept to a minimum.

(The reason I’ve talked about file associations first is that for example if I click on a jpeg it opens in Konqueror by default because I’ve found no way to distinguish between local and net use and I don’t want jpegs on webpages to open in a seperate application but I’d prefer them to open in gwenview locally. It’s simply not possible to seperate Konqueror the filemanager and Konqueror the browser due to this. It’s also not gonna be possible to reliably choose between easy default apps and more complex, flexible, modular and integrated apps like konqueror because the whole mimetype system is broken.)


Dawit Alemayehu says:

Outside of the point you were trying to make, what exactly is broken with the Dropdown List completion mode in KFile dialog ? I either do not see it or I am not getting your description of it…

koos says:

Good points IMO and to add why I only prefer embedding above starting a new app when previewing a link is that the embedded viewer starts faster πŸ™‚
The leadership point you mention is one that I wonder for some time too. Eg. when posting a patch to some lib from kdelibs, one would at least expect a comment from that lib’s maintainer. Even more, if its the first one for a given library. Now who is it, who is the maintainer of kdeui or kdecore? I’m with KDE for some years now, but I still don’t know who and how decision are made about which patch goes in and which aren’t. Sometimes I have the impression that just applying them and ignoring any comments on kde-commits, one can commit anything one wants. I hope I’m wrong of course, but some transparency about this would be nice. Should KDE introduce elections for this?

Leo S says:

James, while I agree that too many options leads to increased difficulty in software testing and maintenance, you don’t understand why people don’t like Gnome or why people like KDE.

I prefer KDE because it is configurable. For example, in KDE I can rearange the order of the buttons (minimize, close, maximize) in the window titlebar (in kcontrol, Apearance & Themes -> Window Decorations -> Buttons). Now this is definitely an example of extreme configurability, but I love it. I can put the close button in the top right corner and the minimize in the top left. Thus, the two window operations I use most often can be accessed with a flick of the mouse on a maximized window.

Now lets take a feature as an example, say single click versus double click. On Microsoft Windows, the standard is double click and you can’t change it. Now say there was a usability study that determined that double click is far more usable and less confusing for new users (just as an example). Should KDE then remove the option for single click? After all, the usability study showed that single click mode causes confusion for new users..
This kind of thinking is what is shaping the Gnome desktop, and it is not what I want to see KDE moving towards. Yes, this kind of feature cutting will result in a simpler and easier to understand environment for newbies. There is no doubt about that. But we have to consider that there are also a lot of technical users out there. For them, single click (or whatever feature you may describe as “useless”) is not confusing at all, and in fact saves them time. They should not be ignored.

Personally, I don’t have a problem with KDE being targetted to more technical users. Yes we need to work on usability, but cutting configurability is not the way to do it. If it was, KDE would have no users left and everyone would be using Gnome or OS X.

Eric Laffoon says:

I would agree that there are downsides to embedding and reusable objects. I’d say we’ve used them not just because we could, but because they have extended our reach. Writing all these applications requires resources and components also promote consistency. The problem in my opinion is not having adequate management mechanisms. We’ve encountered this with Quanta where we use lots of applications as KParts. We’ve had to write management routines for the interface and largely accept that configuration dialogs are going to multiply and help information for the plugin isn’t there.

One of the things we’re looking to do is create a mechanism for not just a profile, but a personality. This could incorporate rules not only for adding menus, but what is hidden, how it’s arranged and other details for what is shown to the user. Personalities are all the more useful when you look at the possibilities. Aaron Siego rightly says that a bi-level interface for beginner and advance users doesn’t work. However being able to create interface personalities related to tasks or that could be distributed and shared would be a different matter. What Mathias is saying is that a sharper focus to a given task has merit, but why throw the baby out with the bathwater?

We will do the personalities for Quanta and I think KDevelop has an interest too. Let’s face it though, other applications and KDE as a whole could benefit from this idea. Most importantly, rather than the developer deciding what the user needs it would allow users to create personalities for tasks which could be shared via email, download or KNewStuff. By defining a general framework we could continue to have the benefits of components while enjoying effectively more diverse solutions and interfaces.

Tim Whitbeck says:

I’ve devised a solution to the window problem. At least, I hope I have. I started a kollaboration at about it. So far it hasn’t really seen any discussion yet. I’d love for you to check it out, and perhaps let me know what you think.,com_smf/Itemid,48/expv,0/topic,301.0

Datschge says:

I’m concerned about your kind of input you give KDE these days. Especially this blog entry gives me the impression of a populist (as in ‘popular to those who have no idea about what you talk’) superficial rant by someone who seems to have lost touch with the pace of KDE community’s development. This is particularly sad considering this is not coming from a newbie who doesn’t know better but by the project’s found who supposedly should know better and still apparently thinks it’s for the better to do a embarassingly improper sweeping blow to the project’s community. And this at a time which due to the on-going code porting appears to be one where deep changes can be pushed through by a vocal minority. In your rant you basically devalued the current meritocratic rule “those who write the code, decide” after all.

Quotes from the blog: “I don’t want to read a manual before using a mobile phone, or operating a DVD player, and I don’t have to. Sending an email, receiving some files, translating them and sending them back is in the same league of complexity, and it shouldn’t require extensive training, or the need to understand how the system works under the hood.” You start trying please. You just can’t consider it feasible to dumb down one of the most complex technical tool available to people today to be as easy to use as a phone. And I mean a classic dial phone, for mobile phones you should actually do the crosscheck with someone who isn’t familiar with mobile phone what of the phone he is able to use right away. Big fat chances are that he will barely figure out that he’ll need to press a button with a green telephone receiver symbol to actually start the call after dialing. Everything else will most likely throw him completely off if he doesn’t bother consulting the manual beforehand.

In the first part of the ‘Window management’ topic you raise the issue that nobody likes to do (manual) window managing, and the suggested solution is that “the ideal window manager is the one that you don’t use at all. This needs support from the applications.” I would never have thought of that… How about supporting the fit-the-window-size-to-the content-shown ‘maximized’ window mode of the pre-Mac OS X days natively within Qt and suggesting to let the window manager enforce that particular mode by default when showing new windows? Coupled with already existing KDE’s smart placing of new windows this basically resolved your issues, no?

You go on with secondary windows aka dialogs, failing to mention that within KDE those are avoided whenever possible, and are normally never modal, so I fail to see your actual issue with KDE here unless you just want to badmouth it. Then you go on (rightfully so) praising Apple’s sheets and drawers as a solution and I have to wonder if you offer such in Qt and should suggest KDE to use that (after all they ought to be supported in Qt/Mac I would think, so why not in other Qt platforms?).

Then you go to everyone’s favorite configuration topic, right away starting with a ridiculous simplification of human preferences: “Sometimes people like to customize something, but that’s different. Customization means giving the computer a personal touch, with colors, images, or sounds. Changing the computer’s behaviour is entirely different.” Yea right. Make me a slave of my computer’s unchangeable behavior but give me plenty of sweetmeat by pleasing we with my favorite colors, pictures and sounds in return. Way to go! It does not matter that even non-coding people could easily turn KDE into highly customized installations, perfectly adapted to the particularly needed workflow thanks to XMLGUI and Kiosk. No, instead encouraging wider use of these (which would invalidate your “even a community that diverse and large as KDE’s cannot ensure basic functionality working in all modes”) you instead suggest “KDE 4 is our chance to get rid of all that clutter”. And then it goes against the very core what makes KDE a rather special, unique project: “‘Those who write the code, decide’ is a good rule, but it only works if there is a strong maintainer or group of maintainers with a clear and aligned vision. Many parts of the KDE had this, others didn’t. And that’s where we see the issues today.” And let someone lead us to remove the features people with a vision once implemented remove again, and newspeakingly call that ‘vision’ as well. Great vision for KDE indeed.

Then you go on with the file manager topic. As apparently everyone does, I will also agree with you here that the different use cases for Konqueror need to be more clear cut and be reflected in the GUI. And you know what? Konqueror exactly empowers you to do exactly that, with profiles and the ability to link independent XMLGUI files to those profile. Nobody hinders anyone to create highly dedicated profiles limiting Konqueror to just what is needed in a dedicated web browser, just what is needed in a dedicated file manager, just what is needed in a dedicated FTP client, just what is needed in a dedicated Commader clone, just like what is needed in whatever use case one can come up. But no, like with the above configuration topic we rather split everything up, say the issue is with the code while the actual issue is that everyone even including of the project itself (by the way of defaults) fail to make good use of the configuration possibilities.

Regarding those I consider the call for “hot coding sessions” to be misled and just resulting in unnecessarily redundant coding which could be better used within KDE where the actual shortcommings are. I call for better documentation, for better user, sysadmin and distro creator documentation as well as better dox within KDE code.

Thanks for reading till here.

ggabriel says:

The greatest annoyance I’ve experienced with Konqueror, no, with KDE, in the state it is now is derived from the fact that we, as users don’t get the chance to associate a File Type to an application or to behave in a certain way, when it is a local/LAN file or an Internet one. I know all the “network trnasparency” argument, but it’s not how I’d expect it to work.
For example, if I decide to associate a certain image file type to an external application, such as Kwickshow, then when I’m navigating and a link points to an image of that type, I click on the link and Konqueror tries to load the image into Kwickshow, instead of showing it embedded, which is what any user would expect!
This is wrong, but perhaps to solve it would just require to allow for a discrimination of the local/remote procedence of the file and to allow two sets of choices for viewing application and behaviour.

I’d also like to state that in general I agree with the way Ettrich sees things, except perhaps that I’d like to preserve Konqueror as both a file manager and Internet browser, but certainly with a clearer and deeper contextual interface changes when in file manager or browser modes.
With respect to modal dialogs, I’d like to see changes in the lines of Firefox search function (the one that opens a temporary input field in the bottom of the window) that do not occludes the user vision.

Seth Quarrier says:

I find this discussion very disturbing, specifically because I think that you have argued against yourself so elequentally. I HATE managing windows when I don’t have to just the same as you do, and thus the first thing I do is open Konqi and leave it maximized taking up a full desktop. I then can switch between my file system and the web via tabs, which I think are far less irritating. Kparts then lets me view a pdf or kword without having to go to the trouble of figuring out where on my desktop(s) I want to view it. If I actually want to sit down and take some time with it, I open it in the independent version of the same app which has a custom menu system and work on it, but for quick previewing there is no way that I want to leave the browser metiphor. How nice this is hammers itself home to me whenever I browse a text file on the web and KWrite is opened in its own window as default. Not only do I have to deal with a new window now, I have to wait for KWrite to load just to view a README. This breach in the otherwise sane defaults illistrates the great strength of the kparts.

Kparts is also great, IMHO, for increasing the intuitiveness of the computing as it eliminates the distinctions between different formats, why should I care if the website I am looking at is a genuine html file or a pdf or a .doc, it shouldn’t matter, after all what it really is is a document that I want to read and with Konqi I don’t have to make this differentiation if I don’t want to, while if we had files open in standalone mode per default I would. In short, to me it seems like the sane setting is to leave the Konqi/standalone relationship like it is, let me preview files in my browser and open a standalone form of the same app to work with them. I don’t want to code in Konqi, but I don’t want to have to open a seperate window to read a README.

I think that the opinions on either side of this argument are pretty much equally armed. The question then becomes: How do you give the power-users what they want, while keeping things simple and intuitive for your everyday Workplace Joe? The answer lies in a few places, some of which have been covered here, and at least one that (surprisingly) has not. I’ll try to address each of them in turn as I see them.

1) Fix what’s broken. If it’s a bug, fix it! Don’t get rid of the feature it comes from. That’s like cutting off an entire branch because you found a worm in one apple.

2) Implement distinctly different UI’s for the different kparts. I am in love with the way konqueror works, but I would like to be able to more easily identify which kpart I am currently working with.

2a) Build seperate front-ends to individual kparts. Give each kpart a front-end that it can work independantly in. For example, if I have Konqueror open as a file browser in $HOME, and I go down to my run box and type in smb:/, a seperate copy of konqueror should open with menus and toolbars specific to network filesystems. If I’m in konqueror browsing $HOME, and I open a new tab and type smb:/ the toolbar and menus should change accordingly when I switch back and forth between tabs.

3) Streamline user input situations. I really liked the ideas concerning modal dialogs and the likes. providing these things inline with the current application will cut down on a lot of confusion and increase productivity a LOT.

4) Redesign or modify kpart containers (like konqueror) to provide sensible and APPROPRIATE context menus and toolbars for the active kpart. This is really just a re-emphasis on the points made in 2a.

(This is the part that I would have sworn someone else would touch on, but nobody has)

5) Provide experience level settings. Default KDE to a ‘simple mode’, where everything is pretty much straightforward (even kcontrol could be setup this way, offering different levels of configuration based on an experience setting), and allow people to switch to ‘intermediate’ or ‘advanced’ levels at their discretion. This would involve a massive revamping of the system and the help files, but isn’t KDE4 supposed to be the ‘next generation’ of desktops? Aren’t we striving for a product that satisfies the productivity aspects of Workplace Joe, but at the same time has the flexibility and configurability demanded by our power users?

Now, I know my wife would never leave the ‘simple mode’ setting, my brother would probably not dare to venture beyond ‘intermediate’, and I, personally, would only see simple mode long enough to switch to advanced mode; at each level, however, the needs and capabilities of the user are met and matched by the desktop. Nobody is confused, and nobody lost features.

These things just make simple sense to me, even though they would require a world of work to implement. I have not used a window manager other than KDE since I started using KDE. I switched from Ximian because I too felt a sense of ‘they know whats good for me’, and I am very particular when it comes to how I work. My wife switched from Windows XP to Debian/KDE because of the overall stability of linux as compared to windows, and my brother uses KDE because he says it’s more ‘workplace’ oriented for people who are actually trying to be productive. He hates Gnome with a passion, and chose windows over linux until I introduced him to debian+KDE.

From what I’ve seen on this thread, too many of you are looking at things in terms of “I’m right and you’re wrong” when the whole point is nobody has to be wrong if the environment supports everyones idea of right (did that make sense? It did to me).

Flames & Accolades welcome,

Travis says:

“you go to control center and just change it to what you want, you get another ‘damn. i took the right desktop’ feeling.”

Acutally, I usually scream, then spend a couple minutes finding the right thing to change. :/

Janne says:

I already made similar comments on kdedevelopers, but they have a place here as well :).

There is one area where KDE fails today: it’s aimed at power-users. There, I said it. And it’s true. How can you tell it’s aimed at power-users? It offers gazillion configuration-options, features and apps. Power-users use them. They have the knowledge to choose from all those apps the one they want to use. they have the knowledge to tweak their system to their liking. They are not bothered by zillion toolbar-buttons or dozens of menus with several submenus. Those things do not bother them, and they welcome some of them, because it gives them “power” and “choice”.

Regural users and newbies are at a loss with them.

People want to use their computers. They use them to do certain tasks. While advanced users might need lots of similar yet separate apps and lots of configuration-options and features, regural users do not. They just want to get their job done, with minimum of fuss.

KDE should be aimed at regural users and newbies. And yes, that means cutting back on clutter, UI-elements, multitude of apps and configuration-options. All those things merely confuse regural users. I have seen it with my own eyes. Those things bury the relevant information underneath a layer of clutter.

I can already hear the cries from the power-users, demanding that their precious configuration-options must be spared. Well, they would be. They just wouldn’t be “in your face”. If some power-user wants to tweak some things, he’s advanced enough to do it, even if the option is hidden somewhere. But it’s simply too much to expect regural users to go through multitude of options and apps just so they could get their tasks done.

Example: Regural users are not interested in changing the location of the button in the windecos. Yet that option is prominently displayed in the Control Center. Stuff like that clutter the system, and they make it harder to change the things that are truly relevant. If some power-user wants to change the location of the buttons, he could still do it. But does that option have to have a whole section dedicated to it in the Control Center?

Instead of expecting power-users to tweak the desktop to their liking, KDE is catered to power-users by default, and regural-users are required to go through the settings and apps in order to do something productive with KDE. And that is completely the opposite to the way things should be.

And why do people tweak the options of their desktop? They do so because the default settings do not satisfy their needs. So, the moment the user tweaks the system, the usability of the system has failed. Tweaking is a sign of a shortcoming in the default settings. If the system worked the way the user wants it to work, no tweaking would be required.

Of course, KDE can’t satisfy everyone’s desires. But it should be configured in such way that newbies and regural users are comfortable with it. Those users are NOT prepared to go through configuration-options, whereas power-user can do so just fine. Therefore KDE should be configured and designed with regural users and newbies in mind. Configuring it with power-users in mind means that KDE tries to force newbies to do something they cannot do, nor do they wish to do it. In reality, no-one wants to configure anything, they just want it to work.

My wife finds KDE to be confusing. The UI is cluttered, toolbars are full of stuff, there are too many menu’s, there are several apps that do the exact same thing as other apps do…. It’s all very confusing. I told her that she can configure KDE to her liking. But she’s not interested! For starters, she finds KDE’s configarion-options too confusing and scary. There’s just too many of them! And second: she’s not interested in configuring the system. She does not use KDE so that she could configure it. She uses it to carry out certain tasks. Expecting her to configure KDE to her liking is an extra burden she’s not willing to carry.

Example of all this fluff in KDE that confuses my wife: She wanted to edit some text. Imagine her surprise when she found out that she’s expected to choose from THREE similar yet separate apps! Yes yes, there are reasons for having three editors, but those reasons are bogus. The user simply wants to edit text, expecting the user to choose from three separate apps just to carry out a simple task like that, is simply absurd! The user does not want to make choices like that, he just wants to do his job. Power-users might appreciate the “choice” they are offered, but regural users are simply confused. No, passing the puck to distrobutors is not the answer, when things like this should be fixed at the source: KDE itself.

Jonathan Dietrich says:

1) Something that I have seen for OSX that I would love to see for KDE is
“a universal canvas where you interact naturally with singular, visual representations of the people, places, products, etc that define your life!” (Marketing speech, sorry) see for info.

You manipulate images that represent people, places, things and when you click on them you get a set of available actions which can be expanded using XML. Even better is when you drag and drop from one to another you get a list of actions that make sense for these two objects.

Very cool stuff.

2) “OK” informational only dialog boxes suck. If all you are going to do is give me some info, and not offer me a choice, then don’t steal focus and demand interaction at all.

An idea I have had is implementing something like newsticker, but that would be a universal message ticker. Knotify could shunt these messages to the ticker.

Individual messages could have the following properties:
an action to execute when clicked
a set of tags defining its classifications (might include app)
an icon

The ticker could be set to expire messages based on classifications after a set period of time, or only after manually removing them.

A dcop (or equivilent) interface would allow other processes, like cron jobs etc., to directly add items to the ticker.

Heck you could even set up multiple rows on the ticker and have the configuration send messages of a given classificaiton to a given line. Allowing higher priority messages to appear on a separte line.

Janne says:

“In your rant you basically devalued the current meritocratic rule “those who write the code, decide” after all.”

While meritocracy is good, there is one problem here. Those who code are power-users. Newbies and regural-users are not coders. Therefore their voice is not heard. When the coders decide what KDE should be like, the end-result will look like something that is designed by power-users for themselves, because that’s what it would be. By power-users, for power-users. Regural users and newbies would be at a loss with such system. And they already are with today’s KDE.

AdamW says:

“I don’t want to read a manual before using a mobile phone, or operating a DVD player, and I don’t have to. Sending an email, receiving some files, translating them and sending them back is in the same league of complexity, and it shouldn’t require extensive training, or the need to understand how the system works under the hood.”

There’s a fundamental fallacy in this often-made comparison. Cell phones make phone calls (they can do much more complicated stuff as well, but to do that you usually do need to read the manual, or experiment as much as you do on a computer). DVD players play DVDs. _That’s all they do_. It’s therefore easy to design them in such a way that making calls / playing DVDs is both easy and intuitive, since your interface doesn’t have to do anything else (or, with cellphones, it does them in ways as arcane and complex as a desktop computer). A desktop computer is capable of sending emails, true. It’s also capable of making phone calls, playing DVDs, rendering a movie, editing photos or music, writing a term paper and many thousands of other things. Most people who buy a computer want to do more with it than send emails. It is therefore not feasible to design the entire interface around the task of sending emails. With a general purpose interface you’re more or less doomed to the fact that doing any _one_ task will be more complex than it would on a machine dedicated simply to doing that one task.

A much fairer comparison for the task would have been between the cellphone, the DVD and the Amstrad e-mailer. . But obviously that’s not what you want to talk about, you want to talk about a general purpose interface; in that case you should be aware of the inherent complexity and the falsity of your initial comparison.

Emmanuel Blindauer says:

don’t forget to clean all menu entries in konqueror. it was discussed for kde3 but nothing was modified. Same for the 4 entries for configuration, there are at least 4 entries in menu. which person has used them all 4?

ggabriel says:

Today, when commenting about this blog in another forum I talked about something that I think could also be great seeing in KDE4.
The idea is that it’d be great to see a new kind of kicker applets that show simpler views of the information usually managed by greater, more complete programs.
I envision this as new applets as thin horizontal lines at the top-right part of kicker, from where the user could click to unfold (upwards) the most common used information needed for daily use. This unfolded applets sould make available basic controls for, say, Kopete’s or Kontact’s most used features. For example, in the case of Kopete, this could be the list of users and some basic controls of the conections state of the user; in the case of Kontact, this could be the folders that contain new mails, links to the currently relevant notes or todos, the birthdays at the current date, etc.
All presented in the smallest and concise way possible (i.e. using a small but clear font, tiny icons where appropriate, etc.).

I don’t have much time today, but a mockup of this idea would be far clearer that a thousand words about it, of course.
I think it’d be wonderful if we could have something like this included in Plasma for KDE4.

kajaman says:

While I agree that watching ordinary computer users can be very helpful while designing and improving software, I disagree that complex tools such as Konqueror are bad idea. Yes, Konqueror have to work better with many different “protocols” (problems with “trash://” or “media://”) but it gives users of computers quite high level of _abstraction_, making using it to do advanced tasks very easy.
Konqueror is IMHO the most powerful file manager out there and please, _please_, don’t turn it into nautilus. The true is that using Konqueror is easy, and there is one more thing that is very important in corporate environment – it is easy to make “templates” with it. Imagine that situation, where users have to update data on the some web site. They don’t know anything about FTP etc. With Konqueror we can simply put shortcut to URL (ftp://companys.web.server), and they manage files stored on the FTP server in the same way they do it with local folders.

Konqueror is powerful, while it is also easy to use and configure, please don’t kill this functionality. We have one nautilus. One is enough!

Angel Blue01 says:

Actually you are wrong. I love beign able to single-click a file to start it. I always set it Windows. Double-clicking is wrong, its never done elsewhere, and we have people like my mom who can’t double-click.

I love having the Web browser be the file manager. That’s the main reason I use IE (in Windows). I want app convergence.

I would like a more intutitive control panel.

* Me is known as BobbyOe says:

Your point of view is right. I agree with It, totally.

The main point is, I use Konqueror for just a while now, and I’m not loving it, just using it. Means it works well, and that’s sufficient.

To do good apps, just do something that works, don’t crash, and is flexible (just a little bit needed).

That’s not true of course. I loved Firefox, and others apps that just did what they are supposed to do. But I much prefer to always use same apps, see same buttons, and be “relaxed” and not frustrated by a new button that I don’t know what’s doing.

The same reasoning say “If you spend 10 years to know an OS, why you would change ?”.

And that’s why apps should be more like what you said. I wish to see some of the changes in near future.

And I hope that it will contribute to the KDE success

Klaus Blindert says:

I think stripping down the functionality to a bare minimum
would be a mistake. On the one hand power-users might be
the only ones who – for example – really want to change their browsers personality. So there shall not be a toolbar button for that by default just because someone implemented that and found it a good idea.

But I think average users are still introduced to new software and technology by the power users. Thus cutting them of from KDE by stripping things out will make them introduce users to other DE’s or just let them say, “Stay with Windows, you’ve got nothing new to see over here (and it’s easier/more familiar)”

Then average users being confronted with KControl or too many choices ( for example if I want to preview a plain text file I get: Embedded Editor,Embedded Vim (!), KXsldebug(!), KWord,KLinkstatus (!) ) will get scared.

So my proposal would be:
Add more customization, more options!
Yes, but then give them in bundles, like the “Office”, “Web Development”, “System Administrator” where a wizard takes a few well thought out options ( Do you need a spreadsheet? , … ) and then automatically configure the system alike.

I’m not sure wether this would be a nightmare to implement, but I guess even power users would like it.

Concerning konqueror as filemanager … I’ve infrequently tryed it out and only half a year back I really started using(and liking)
it. So whatever you’ll be doing to it, be careful πŸ™‚

smileaf says:

What I’d almost kinda like to see is modules used like how java uses classes.
create a module for file, ftp, etc just like kioslaves are now.
but also have them mold konq into a new beast to fit it’s task.
Using this approach however I’d like to see like tasks group in 1 window.
If it’s a file based view that’s 1 group, 1 window.
if it’s html that’s another.

If a module builds upon another one the higher up ones only get more powerful.
For example:
low level we have local file browsing,
next level we have remote file browsing. I’d personally expect to be able to do everything I can do locally. currently with the ftp kio slave moving files across partitions cannot be done. to the ftp user this is not obvious.

I would have to disagree with that we should split konq into multiple apps. that I would think would create 3 times more coding hassle.
in another website about having 1 viewer for every file type i would also have to disagree with.
Why? lets continue to use Konq as an example.
It’s a browser. when your using the web your browsing. when your finding a file either remotely or locally your browsing. it’s a like task.

Lets also go back to my idea of layering.
in another website the author made a really good point.
Look at kaddressbook. Why can’t we have that functionality by default and have every kde application that would use a piece of information stored in that get that info from that.
such as why do we have a place for email in kcontrol.. _and_ kmail?
if I add an IM contact in kaddressbook .. why isn’t it added to kopete?
Why can’t I set a contact for a contact in kopete to use after it’s added?
If I add a rss feed in kaddressbook why can’t akregator pick it up?

If it’s something that multiple applications will use it should be lower down in the layer.
higher up layers use those lower layers to add more functionality.
This to me would be an excellent usability improvement. As well as a good model for future expansion.

Another unutilized idea is the sidebar.
File preview, Numerous information presented in a reader friendly form.

alucinor says:

I think the best way for Konqueror to suit both the needs of powerusers and average users is to use an approach similar to Firefox, where the base package is very lean and simple, but it is infinitely customizable in the form of extensions. Being a self-proclaimed poweruser myself, I have my Firefox totally tricked out with several toolbars, as well as the Opera-like sidebar extension. I really like in Firefox how you can drag and drop all your toolbar icons to essentially tweak your interface in whatever way you see fit.

Yet my grandma also uses Firefox and loves it because she understands it, as she keeps her interface at the default.

Heck, if someone ever creates a filemanager extension for Firefox, it’ll only be a matter of time before Konqueror really has no point, unless all us KDE peeps can keep up with whatever crazy XUL concepts they’re dreaming up at the Googleplex and Mozilla Corp.

rod says:

one thing i see with kparts it’s they integrate their buttons and menus with the main window menu, and change the positions of other buttons. The good thing is they help to avoid filling the desktop with windows that have to be closed by hand. Before using KDE i tried the rox filer and i liked the mouse-right-button menu, so why not remove (optionally, windows look prettier without a menu on top) the whole main menu, and put it in the right button, a la NeXT.

this will need to rethink all the mouse handling of applications (like having a one button mouse), and has the problem that no-kde apps don’t fit.

P. S.:

1.- i see that removing options and make very simply interfaces it’s good for casual users and offices (mainly where gnome is targeting) but it tends to dissapoint people that uses computers for more long time, like home users, that like customizing his desktop at will.

2.- it would be nice an option to save my KDE Destop configs in a single file, to load if i uninstall and reinstall.

I find your points to very sound and very in tune with what I myself find to be true.
I use the ion3 window manager for the exact reason that you mention: I don’t want to spend time managing windows. I did it once, and now all my windows always come up where and how I like them.

I am myself very interested in usability – in particular the thinking about interfaces in ways that don’t necessarily adhere to the “has to be entirely compatible with what people are already used to” school.

I would like to point out one thing though;
You mention watching inexperienced users and learning from it. This is a very good idea. But you should also take the time to observe som truly proficient users. Often, this will give an entirely different set of clues as to how you can design your interfaces.
Newbie-friendly is certainly not the holy grail. It’s an important thing, especially for a platform that is trying to lure users away from something well known – but don’t forget the power users. Just because it takes a while to learn, it might still be an excellent interface with a very optimal workflow.

abdulhaq says:

Reading the above posts I would like to propose a solution to some of the problems people are facing:

1 – Files/links are ALWAYS previewed when possible when single-clicked and always in read only mode.

2 – When (1) is performed and the user is viewing the file, there should be an easy and consistent way to open the file for editing – a large button in the toolbar, hot-key etc.

3 – If the user knows that they want to edit the file/link then they can dispense with the viewing phase and open the correct app in its own window by choosing an easy and consistent option such as described in (2).

The above idea solves many problems including my own bug-bear where I close the viewing app after reading the doc only to realize that I’ve just closed konqueror by accident.

By the way, the designer of this website could increase usability by increasing the size of this silly little box I have to type in to make this comment,


Zeljko Vukman says:

If you guys want to do something really useful then go and create KE (K Environment) and abandon KDE (K Desktop Environment). Kill the desktop paradigm and kill the WIM(P) (Windows, Icons, Menus and Pointing device) user interface.

As long as we have windows, windows managers and desktop around we will never have real usability.

MacOSX, WindowsXP, Linux, BSD etc. are all in deep trouble. All of them follow the same paradigm. Pixel here and pixel there doesn’t make MacOSX so much more usable, because you still have to: launch, close, resize, move, minimise, maximise etc. your applications. Billions of times every second the world population do stupid things like managing windowed applications. What a waste of time.

A couple of days ago an old man called me and told me that he had lost “his Internet”. I went to his place and discovered that he had minimised browser window. It was there, just minimised, but he couldn’t find it. It sounds funny, but it isn’t. Windowed applications and UI based on desktop and windows IS a big trouble. 99 % of people who use computers are not like you and me, they are NOT interested in computers. You know it.

So, beat Apple and Microsoft by creating something which is totally new. Start thinking beyond windows & desktop paradigm. Let’s do some brainstorming, and you will see that it is possible.

It IS possible, because you have technology: embedding is your savior. Not embedding as it is implemented right now with crippled applications in Konqeror’s windows, but universal embedding in a workspace area. Imagine that you have a tabbed panel from which you just drag and drop applications to a workspace. Applications automagically embed and autosize into the workspace. You never close them, just drag and drop them back to the tabbed panel. One application on the workspace always autoresize to a full-screen state. If you just use one application you don’t need it as a small window. You drag and drop another application to the workspace: they both autoresize to half/half of the screenspace, etc. If you want to make one of them bigger, another one becomes automatically smaller. That kind of behaviour you ALREADY have inside many applications. Why not implement this globaly?

And panel, imagine that tabs in panel represent virtual workspaces (as now virtual desktops) each of them having one application which is already opened. Clik on the tab and you have your application in full screen state. Drag & drop another application from another tab and you have two applications autoresized on that workspace.

Imagine having one tab for Settings. When you click on it you get workspace with global computer settings. When you drag and drop it to the workspace where you have one application you automagically get settings on a half of the screen area for precisely that application.

With that kind of computer interface you will still have GUI, but millions of times more intuitive GUI than it is right now with desktop, billions of icons all around desktop, and what is most important you will get rid off minimising, maximising, closing, resizing etc.

And for the freaks who use desktop to watch nice pictures here is solution: image viewer on a image viewer workspace with all their clocks, weather etc. Karambas.


P.S. I’m going to make some mockups and post them on my blog.

I agree with many others in this blog.
DE’s need profiles for diferent users. In the same way we choose for the first time the desktop appearance, it should be available three profiles ‘a la nautilus’. Advanced 100% of functionalities and configurability, Medium 50% and Simple 10%. That choice should be available too at Kuser and a proposal for interface and profile choosing could be something like Kde-kiosk.
Of course, everything should be polished more than just mixing everything up.
Kde-kiosk profiles could be shared at kde-look and so users can download w3.11 profile, Kde-lite profile, Kde developer profile, kde designer profile, etc…
Through a download like the one at desktop backgrounds, users can choose what they prefer most. But anyway, default profiles should be enough in a first place.

mic says:

“Keep It Simple Stupid”. It’s works!

madman says:

About the ideas behind an image viewer. I’ve seen windows’ image viewer and it’s a crap. The real INTUITIVE image viewer with slideshow function is GQView. You can pass a directory path to it. Then press ‘F’ and you have a full screen viewer with easy keyboard/mouse navigation.

You can browse photos with the mouse scroll, left/middle click, or keys: space/backspace. It doesn’t have any toolbar, which would take up a lot of space which can be used for displaying photos

Ian says:

I think KDE needs to balance making something intuitive for the novice and adaptable for the power user. OSX in my experience leans far too much towards the former, KDE perhaps leans too much towards the latter.
Personally, I find OSX very frustrating because I can’t configure it to work the way I want. KDE, on the other hand, I can pretty much adapt to work just the way I like & that’s why I use it!

Sure you could change the default toolbars in KDE to be much simpler, but to make them less configurable would be a very big mistake in my view.
As an example, in kmail, there are 2 buttons on the default toolbar that I never use, the rest I use all the time. I also add about seven extra buttons, all of which I use daily. Try doing that in an OSX email program – I doubt you could customise the toolbar at all.
I would never want big icons and text in my toolbars, that might be useful for beginners, but once you know what the picture means, you don’t need text and they only need to be big enough to recognise easily.

On the other hand, kword has 3 rows of toolbars across the top by default and one down the side, plus the Document Structure panel.
Compare this to MSWord97, which has 2 rows of toolbars and none on the side. You end up with far more useable space for your document in Word and have pretty much the same amount of functionality in the toolbars.
It may not matter on my 1200×1600 screen, but try using kword on an 800×600 display – it’s horrible. Word, on the other hand, is perfectly usable at that resolution. That comes down to the default toolbar icon size, I guess. Still, at least in KDE I can change that, Windows doesn’t give you the option.

There are some usability features KDE could adopt from OSX – drag & drop for instance. MacOSX uses drag & drop far more extensively than any other current OS/DE. Then again, because I’m not used to dnd working in some situations, I don’t think to try it, so for me it’s not intuitive.

As for splitting up Konqueror, like a lot of others, I really like being able to use one app to do so many things in one window – nice and tidy. And it does that without the highly unstable ‘kitchen sink’ approach Microsoft use – in KDE you still have small components that do one job & do it well, they just fit together to form ‘swiss army knives’ if that’s how you choose to use them. If you don’t, you can still use each tool as a standalone app.

The trash example is quite right in my view, but then I have desktop icons turned off and use the trash applet in kicker, which gives you the empty trash option when you left click on it, just like a Mac!

So to sum up, I think KDE should stay just as configurable as it is now, but provide a simpler interface by default.
Or perhaps there could be two defaults settings – beginner (fewer toolbars, less options) and expert (all the options turned on and more toolbars, all fully customisable).
The hard part would be getting a room full of people to decide which options to leave in the simple default πŸ™‚

Perhaps a settings wizard could be used to allow the power user to easily transfer their carefully customised setup to another machine.

Eric B. says:

I’ve been a kde user since some time now. I’m not a kde zealot and I think kde is somehow “overloaded”. Been thinking that way a very long time but never couldn’t go on Gnome which was lacking some features I needed.

I’m very happy to see kde developpers are looking elsewhere to learn, question themselves and maybe improve our fav desktop.

I’m a lil bit scared when I read some kde users saying “please do not make kde 4 too much like this or like that, we love our toolbars, icons, our stuff, we like it this way, we don’t want to look like Gnome, we are different, etc.”. Kind of scary IMHO. It’s not surprising people get attached to their environment and GUI. It may be when they want to “freeze” it and see their interface like an “anti-this-or-that” so after a while you got an army of zealots keeper of the truth telling you : this is the law, here you did okay, there you did wrong.

I’m ready to change my way to do things on the kde desktop if it changes enough for that (at least with default settings). Do not hesitate to evolve, do not hesitate to start some “user revolution” if necessary. I want the kde developers to know that and I hope I’m not the only one to think that way lol

Dasher42 says:

I am a very happy KDE user, and prefer it for functionality to most anything else out there. My thanks to all the developers behind it. I want careful thought put into any changes made to it, because while changes are needed, let’s not break a good thing!

There was one aspect of KDE that made me make the switch – extensive keyboard configurability. Minimizing and maximizing windows, most of all, should have single keystrokes bound to them in all applications. An easier interface to minimizing and maximizing make window management less of a pain. I respect Wirth’s careful thought, but in those rare moments I need tileable windows, I really do need them.

That said, I knew my preferences, and what I would have liked would have been the chance to make them once. Like many KDE users, I spent a chunk of time configuring it, and I trimmed as many little-used toolbars from my preferences as possible. I think KDE has been thinking too much in the Windows 95 box of desktop use. Point is, there are a lot of common keystrokes between similar applications. Rather than navigate through a lot of separate configuration programs, let’s tackle a basic UI issue that OS/2 got right: preferences, including things like keyboard shortcuts, should be object-oriented. We shouldn’t have to redefine the same set of keystrokes for KDE, then Konqueror, then others. Oh, and drag-and-drop application of font sizes and GUI styles to windows was a good idea, too. KDE has drag and drop working in ways that I really appreciate, and I’d like to see more.

The SimpleKDE fork should not remain a fork for long. KDE needs simple, clean layouts by default, just like that. I think the default KDE desktop package should also be trimmed down; let users grab games or education packages if they want them, rather than install by default. And let’s not repeat having three separate text editors by default!

KDE has some key things right, and I think with some clear UI design in KDE 4 it will have nothing to worry about next to other desktops.

dr01d says:

One of the draws to Linux is the ability to have choices. KDE does have a lot of clutter. But, maybe it’s a necessary thing. No matter how you slice it, Linux users aren’t going to like the way some things are presented to them, and they are going to want to tweak things, change settings, etc. That’s what’s so great about the platform! My point is this: there have been some good points raised about usability and simplifying KDE to be better. I actually agree with a lot of them. However, as opposed to ripping everything out, making dramatic changes, and streamlining everything… why not have different “modes” (preconfigurations) of KDE? I (like everyone else I know) has to spend a lot of time configuring things how I want them after doing a fresh install. Why not have a wizard that does more than allow you to choose themes? Why can’t the wizard offer some preconfigurations? Then I could choose the preconfig closest to my work setup and minimize the work I have to do. There could be a developer config, newbie config, power user config, artist config, minimalistic config (maybe like xfce4), admin config, etc. Choices. That’s what it’s all about. Oh, and the ability to get in there and change everything too πŸ™‚

Dusty says:

Simplicity.. Linux is overly convoluted.. Just because we are geeks and we want power doesn’t mean it should be right there in front of everyones face.. make things simple and elegant, keep the complex workings out of the face of those who don’t know and don’t need to know.. KDE is great.. and I mean that.. but where is the innovation.. we are still putzing around with 30 year old concepts here.. how about real innovation for a change.. we need intelligent interfaces that conform to the user, not vice versa.. The way I see it is this.. KDE is the gateway to the OS.. So if KDE isn’t as easy to use as say OSx people (newbies and nongeeks) assume that the OS isn’t as good as OSX (or even Winblowz).. We need simplicity, visual impact and work flow improvements..

my two pennies..

Noam Samuel says:

>But we (the KDE users) can customize and configure quite many aspects, and thats just what we (or I) want.

As a GNOME person, i’d want you to note the way GNOME solves this: the options are minimal, with simple and usable interface, but the desktop is configurable via GConf.

the idea: minimizing normal conf but still letting power users edit “Advanced preferences” without cluttering the control center.

Ofer says:

For me Konqueror is an excellent utility in all of its aspects (but still, I think, a bit immature as a web broaser). I find the view profiles idea amazing. It would be even better if we could also launch apps and more as app-profiles, for example, launching kmail, oo, vlc etc….
Great preview capabilities, yes! embedding, yes as well! I often make use of the embedding capability. In particular, in parallel with a terminal which I open within Konqueror (I latex tex-files and see the kdvi or kghostview gets refreshed interactively).
Another idea for improvment: a special toolbar for available profiles, with decorated symbols for each profile.
Lots of thanks to the KDE team!

tbscope says:

“Konqueror is not a document manager”

In my opinion it is.
It’s a network transparent document manager understanding lots of protocols to manage documents.

The webbrowser is an embedded html viewer of this document manager.

Why not develop a dedicated webbrowser for KDE instead? And leave konqueror like it is, maybe cut out some embedding.

If you want to solve window management, embedding is a good way to do it I guess.

As for configuring the desktop and all programs?
Why not use webpages for this?

pblewis says:

On window management.

May I add my ha’p’-orth, the behaviour that always places a window at the top left of an empty desktop is infuriating. Why does it still do this as default? Certainly it is possible to place it central and then configure that as the remembered position but who always works at the top left for preference?

On the plus side, now, after many iterations, the dialog that appears permanently on the top of all windows on all desktops but does not grab the focus, works well. So please don’t jeopardise this success.

On Konqueror:

Right clicking a file/folder has na “actions” menu item, which for a folder has a “browse with gwenview” entry which is one clock away from your full screen browsing. It would be great if these actions and the “open with…” actions were presented in a floating or fixed toolbar which adjusts its contents to the selected file/folder.

Thanks for thinking.

Trueneo says:

Hi Mat,
I’m agree with you. Desktops today, and even more yesterday, needs a knowledge that the newbye does not have. All of us, linux users, are smarter enough to configure, change, operate, customize our desktop and all the other components of our linux box. We do not need a system simpler because we know how to work with it.Instead we want a more complex KDE to show at the MSWindows/Gnome users saying “Hey! This is a DESKTOP not yours”.
The question is: “my mother will get comfortable with her linux desktop?” the answer is “NOT!”.
An example:
my mother wants her ADSL connection, the ISP give her, as usual in Italy, a login a password and an e-mail address with all the data like POP3 and/or IMAP, popmail and sendmail ecc.
My mother turn on her linux box and she freezes.
How can I connect to internet?
The simplest way could be an icon on the desktop called “INTERNET”.
My mother click on it, nothing happen, she do not know the double click (the kde default is one click but her son change it in double because is difficult to select,move, copy and so on with one click). My mother try to press the LMB for a while but nothing appen. STOP! Maybe something we learn, why do not popup a menu or open the application if we press LMB on the icon without move it? My mom get confused with the three button of her mouse she looks at it to find the second button or the RMB! Her Double click is too slow. This does not mean that we eliminate double click or the popup menu with the RMB. The desktop needs to be smarter, if she (my mom) clicks on a icon and keep the pression means that she wants a response from that icon. If nothing happens she wants to know why, what’s wrong with this icon?
Go on.
My mom finally open the INTERNET prog.
The Wizard start to ask a lot of questions! My mom freezes! Why the conf wizards are so confused? The ISP gave three data to my mom: login, password and an ip. I think that an Internet wizard would asks as FIRST, in only one window the login the password and the ip of the name resolver and a telephone number.
But how the sistem will know what device to use? Trying? NO, if my mom insert a telephone number, try to use the modem (if there is one, if it fails go on with the DSL device.
My mom is surfing the internet, she does not insert URL address, she googles some words and load the page of choice.
Now she wants to write an e-mail.
She open Kmail, she freezes once again!
The ISP gave her the following:
login: mom
password: love
pop from:
send to:
The first windows kmail have to display is
your name:
your e-mail:
pop from:
send to:

My mom does not care if she will use pop3 or imap! The password have to be saved automatically! She does not want to insert it every time. Don’t bother her with KWallet! Don’t bother her with PGP or signature, she wants to read html mail and does not know how to enable it and does not want to learn how to enable it!

Everything must have the base config and the advanced config, the base config is suitable for my mom, the advanced config is for me.

I think i could write a book on this, maybe all the basis of the desktop have to be rewritten. All the things that we consider stupid are difficult to persons like my mother, try to leave your granfather in front of your kde and tell him: “here, there is is your favorite far west movie try to watch it by yourself”. On the desktop there is a icon with the name of the movie in DivX.

With respect.

momesana says:

Great Ideas!
I have been working with KDE for the last 4 years and I am aware of the problems you mentioned. KDE-4 needs to be different from what we have now. We don’t need more rarely used features, but we need a usable desktop that people can comprehend easily. Kde as it is now has been valuable for the last years but what we need now is a redesign based on the behaviour of users. Intuitive userinterfaces are very important. Watch the users closely and you will discover the shortcomings of todays desktops. Hopefully we will handle the problems in KDE4.

Roberto G. says:

There is ONE big thing i’m sure that should be done. At the very first start of kde4 a dialog box must ask if you are or not an experienced user. Depending of that, the behaviour of the system changes. The way of these changes could be discussed, but not the presence of this approach.

Marco Menardi says:

Yes, I DO HATE single click also! It hurts me a lot more than it helps (does it help? Is double click too difficoult for someone?).
And for UI design, I would suggest Alan Cooper’s books:
and Joel Spolsky book and essays:
Usability experts, or people expert in the task to acomplish, should tell the developers what is needed, not let the developers do what they want, and then try to correct with some “UI cosmetic”. Often the entire workflow is badly designed, something you can’t correct just moving some widget around.
Btw, also select a range of files in the File Open KDE dialog with multicoloumn view is a pain, if you select the first of the first column, and the 3

l.guégan says:

First, I am not a KDE user, nor a Konqueror user.
But this my (little participation) : Don’t try to be attractive.
We will all become power users because of the benefits to be a power user! I mean what made the success of great softwares? – The freedom power user get from it. Don’t follow the competition. I wish you a lot of succes, but please follow power users, they know beceause they practice and they enjoy… This is the only rule, every thing else is philosophy.

wouter says:

You are so right!

Juan Liska says:

We’ve been down configurability road and it got us lost in the jungle.

I agree with your comments and hope that you will be able to convince all those coders and users who say they are in love with endless complexity.

Gnome has the right idea but they sometimes make the system unusable through excesive limitation. (pet peeve: losing contents of clipboard when you close the ap). I think the key would be to allow configurability in cases where it makes sense and for the configurations to be done uniformly.

I would like to see all advanced rarely-used configuration options in an xml or “registry” file. At least you’ll know where to find them in the rare event that you should need them. Also, this way one could share configurations with other ubernerds via email w/o requiring screenshots.

Dan Michael says:

I would love to see an option to chose between single-click and double-click. I haven’t been able to get used to single-click yet (I also keep opening files and applications by accident).

As always, I agree only partially with the article. One of the reasons why users come to KDE is specifically because it is so configurable. For a good example, about a quarter of the above posts explicitly state it. Yet another quarter decides to debate either with you or themselves about what they like best. In other words, even within the community itself, there is enough dissent to show that limiting what you can do with KDE would not be a smart decision.

However, KDE does need an overhaul. The overhaul shouldn’t be to remove anything. Not only will that annoy users who depend on or appreciate the application, but it will also demoralize the people who have put so much effort into their application.

I have stated this several times before, and I will continue to state it: The problem is presentation, not configurability. The problem is presentation, not conglomeration.

One problem with KDE is that configuration has been outstripped by configuration. There is a specific granularity which should be presented to users. KDE does not present the necessary ganularity. It is true that normal users will not want to tweak many of the settings that KDE affords us. However, the object should not be to remove those features. Put them in a separate level of granularity. Place them in a configuration dialog, or in an “Advanced…” or “Details…” tab. Don’t throw everything up on 1 page.

Another problem concerning configuration is that many times, the configuration itself is plainly presented wrong. Most users find the plethora of check boxes, text areas, buttons, and lists utterly confusing. But don’t remove them. Find a better way to manage them. Take most of them and throw them in an advanced pane, then provide an easy way to manage the basics. And make that top-level management work. If I decide that I want to use the “embedded text editor” for all of my edits, then have it change all the parts of KDE that it needs to.

A real-life example of this is the network configuration in KWifiManager. Tabs are an inheirently wrong way to manage the information. Instead of using them, find a better way to organize the information. Use a list, or a combo box. And when you change something like that, it should remain consistent across configuration. If panel configuration uses a list, then virtual desktop management should also use a list, as should KWifiManager.

The congomeration of ioslaves and kparts that has become Konqueror likewise needs to be cleaned up. However, again, the problem is with presentation. If Konqueror is going to be a swiss-army tool, it should self-deploy. You shouldn’t have to select the tool you are going to use, or sift through 800 other tools to get to the one you need. It should seamlessly shift between profiles. It should provide context, and it should be consistent.

One of the best ways to to this, IMHO, is to have some form of side panel (yes, akin to XP) that changes with the profile and protocol. The menus should behave similarly. If I am in file-management mode, then web-specific extensions shouldn’t be shown. If I am using it to read a DjVu file, then the Okular menus should appear… etc, etc.

You say throw the application out because it is complex. I say make the application more intelligent, so that it’s complexity is hidden from the user.

Concerning the dialog options: I think you were spot-on with them. I prefer the “drawer” boxes, because they are less intrusive. Also, modal dialogs should “stick” to the window that they are attached to, and some form of visual indicator should show the user that the window is not accessable.

Moochman says:

I like a lot of the comments brought up in this blog, particularly the part about e-mail applications and soforth being made easier to use. At the same time, I understand people’s objections to removing customizability from KDE.

It seems to me that a common misconception about user interfaces is made, which is that removing customizability will result in greater ease of use. I couldn’t disagree more: It is only by virtue of customizability that I tolerate most of the horrible defaults the hodgepodge of programs on my computer throw at me. What KDE needs is two things: Better organization for KConfig, and better defaults. I think better defaults are pretty self-explanatory, although certainly they are fairly subjective. As for better organization, let me just provide the example of the double-clicking settings being in Hardware / Mouse settings!!! I never would have found that on my own; it belongs in Desktop customization.

Of course, KConfig is still far easier to navigate than Windows’ settings are and more feature-rich than GNOME, but it could always use improvement. Something like OS X’s presentation that lays out all of the applets in one easy dialog instead of a nested heierarchy would be a good start. Also, when are we going to see more integration of basic hardware settings? The tools already exist, after all–YaST and SaX2 are GPL now, aren’t they?

Regarding Konqueror integration, I don’t think it’s necessarily a bad thing, it’s just that 1) It adds to the already messy complexity of Konqueror and 2) Not enough programs make themselves available as KParts to do the job.

With #1, some minimal effort has to be put into revamping Konqueror’s menu structure, default toolbars and default context menus. I think a happy medium would be a very simple default toolbar (akin to Nautilus or Windows Explorer) at the top of the window, with a context-specific toolbar in the viewer pane on the right that changes depending on the KPart being used/the content being viewed. This would also be in fitting with how people are used to browsing the web: The plug-in (for instance Acrobat Reader) brings up it’s own mini-toolbar below the main one. With KParts being designed specifically for use in Konqueror, this design could be even better-implemented than in the Acrobat example I mentioned.

As far as #2 goes, if more programs allowed themselves to be embedded, then the issue of “taking away mindshare from the best of breed programs” would not be a problem anymore. This could be solved by more application developers deciding to integrate with Konqueror over time, and/or by making it easier for non-KParts applications to embed themselves (like browser plug-ins), or for third parties to build bridges to such programs.

However, despite the fact that integration issues could be made a lot better than they are now, the fact remains that many people will prefer the “one program per function” approach. So, for those people, KDE needs to make it easier to choose whether or not Konquerer handles each filetype. An easy implementation would be something like the Windows approach of Right-clicking, choosing “Open with…” then choosing the program and clicking “Always use this program.” Konqueror could be on that list, so if the user wanted to switch back to an integrated approach they could do so just by repeating the process and choosing “Konqueror.”

A few more thoughts on Konqueror: Those icons need to be clearer, less redundant, and more space-efficient. Why is it that most of the icons are indistinguishable if I squint my eyes, that the width of the icons varies, that there are so many icons I sometimes need to maximize the window to see them all, and that if I choose to view text under my icons it makes a few of them expand to fill up the whole screen and obscures the rest? These are all issues that could be fixed with a little much-needed UI-tweaking.

Finally, the Konquerer Preferences are a mess. I don’t want to see web-browsing stuff mixed in with behavior and interface stuff! It just adds to the already confusing myriad options available to me! As long as Konquerer the web browser is integrated with Konquerer the file manager, at least separate the two preference dialogs as is done for every other KPart!

Well, that about sums up my two cents on KDE: The Next Generation. I heartily salute all those that are making it possible, and congratulate all of you on making the most configurable, most powerful and best-looking DE out there. I’m confident that with this next iteration, “easiest-to-use” will be another descriptor I’ll be able to add to that list.

Massimo Fierro says:

I’m a great KDE fan (hopefully developer in the future) and I’d like to give my humble opinion:
KDE is great like it is, with all it’s features and configurability…. BUT!
But this is the opinion of a persion who uses computers since he was 10.
If I’d put my father in front of KDE he’d get lost in 3 seconds: KDE needs too keep it’s full set of features for advanced users but should provied a more polished and uncluttered interface to the “common” user.
I.E.: you get it simple and, if you want, you make it as complicate as you want.
I also think that Ben Meyer, in this article, has got some points.

Let’s make KDE accessible to everyone! πŸ™‚

Cliff Beach says:

“KDE should be aimed at regural users and newbies.” Oh yeah, all those computer newbies who use Linux. There are no regular users on linux. Of course! you say, we want to dumb down KDE so there are more of them! Until Linux itself is more friendly this is a useless goal. You can dress up the DE but when something happens underneath, newbs are helpless, and they just aren’t going to be using Linux anytime soon.

Discussions like this make my head spin. No one involved in them is anything but a power user. How these power users can ask the devs (also power users) not to focus on power users is mind boggling to me.

I don’t want a DE that will be useful for some nonexistant computer newbie linux user, I want something that I can use, and that you can use. I don’t really care if newbs would choose Gnome instead (here’s a hint, they won’t. That would mean choosing linux).

I do want to see new users using Linux, and I used to encourage it. But I was simply defeated by most peoples’ vehement resistance to learning anything about how a computer works, by their resistance to anything new at all. Most of these people *try* not to know what is going on, and an attitude like that will never let someone use Linux no matter how dumb the DE becomes, so forget them. Try focusing on the people who actually use your software.

Janne says:

“Oh yeah, all those computer newbies who use Linux. There are no regular users on linux.”

And with that attitude, there never will. But, in reality, there are lots and lots of “regular users” using Linux. Lots of them. I have seen them. And I have seen several newbies wanting to use Linux. But currently KDE does not cater to them at all.

“you say, we want to dumb down KDE so there are more of them!”

Making things easier to use is not same as “dumbing down”. And lete me repeat: I’M NOT ADVOCATING REMOVAL OF FEATURES! I’M NOT ADVOCATING REMOVAL OF CONFIGURABILITY!

“Until Linux itself is more friendly this is a useless goal.”

KDE (in this case) is part of that “Linux” you refer to. By making KDE easier to use, you make Linux easier to use.

“Discussions like this make my head spin. No one involved in them is anything but a power user. How these power users can ask the devs (also power users) not to focus on power users is mind boggling to me.”

Because some of us actually see regular users using KDE? Because some of us can see the problems with KDE? Because some of us would want KDE to be easier to use even to us? Because some of us want clean and smooth UI?

“I don’t want a DE that will be useful for some nonexistant computer newbie linux user, I want something that I can use, and that you can use.”

Being easy to use for regular users does not mean that it will be un-usable for power-users. And even if it were un-usable to some power-user by default, that power-user has the required knowledge to make it usable for him. But regular users do NOT have the required knowledge to make hard to use KDE easy to use for them.

“I do want to see new users using Linux, and I used to encourage it. But I was simply defeated by most peoples’ vehement resistance to learning anything about how a computer works, by their resistance to anything new at all.”

Isn’t that a nice excuse to keep KDE confusing and complex? Hint: People might be hesitant to learn anything new. But that doesn’t mean that we should therefore make/keep things hard to use.

“Most of these people *try* not to know what is going on, and an attitude like that will never let someone use Linux no matter how dumb the DE becomes, so forget them.”

You call easy to use UI “dumb”? That tells quite a bit about your attitude on this matter. Hint: It’s not dumb, it’s merely easy to use. And that is what I have been talking about. And it seems to me that you oppose making things easy to use. Things should be easy to use. They should be as easy as possible. KDE is not easy as possible to use.

Seriously, with attitude like that, Linux can forget about ever challenging Windows or Mac on the desktop. Or maybe Linux will challenge those, but it wont be done by KDE.

MrCopilot says:

Ok, You make some valid points. The one that causes me to write is Konqueror.

Web Browser is its future? Come on man. File Manager is it’s present (at least on every desktop KDE I have seen/worked on. Parts are Konfusing, But the absolute best thing about Konq is the ability to open a console. Windows should have had this years ago. I Have to use both XP and KDE at work, I HATE having to open a command prompt, navigate to the directory I want, through dos commands with the crappiest autocomplete I have ever seen, just to type qmake or uic. Contrast that with Konq click Shell Btn (Why isn’t in the toolbar by default.) navigate using filemanager (The fact that all the linux commands are displayed in the shell are brilliant teaching tools for my kids. A Simple clicked folder= display “cd Foldername” in shell is a very powerful meme.)

I almost never use Konq as a web browser. Why compete for third or fourth place. I’m not saying it isn’t a great browser, just not my cup of tea. Please don’t take away the best parts for simplicity sake alone. There is no reason you can’t simplify the GUI for newbs ala xine. (Better than xine please! Bad example.)
Giving more config options is a sensible approach, I could understand if KDE was for windows, Linux Users are used to configurability (Don’t force us to by default, ALLOW us to configure as much or as little as we need. The Linux Way, no?)

I am comfortable in KDE, hate working in Gnome.(Sorry, No offense, As a Programmer & PC Tech I felt looked down on.) Learning curve from windows was not that steep to become semi-productive.

Keep up the great work. I feel very confident in your and your collegues hands if discussions like this are as part of the norm as they seem.

Kan we get a better name than K-menu (Very lame) Katapult (

IRIX says:

first, sorry for my english xD

now, many people says “kde is not osX” and “kde not like gnome”, but i say too “kde not like windows”, in the last relases of kde, i see a icon that hate, “sistem:/”, like the “my pc” in windows, or the “media:/dev/hdax”, again like “c:/”, i like when the people use kde think “i never use thtat but is easy to use” and not “is easy because is like windows”, what i the next, a blue screen when a apps fail, to the users don`t miss windows, i like a kde4 original, not a copy of osX o gnome o windows, get out the icon “system:/”, the “refresh desktop”, the side image in the start menu, and a birth of new ideas, so the users con says “kde is like kde”.

opera says:

Does people read this far? Anyway i gotta answer some replies;

Most people in this world DO agree with this blog entry. However most people reading this blog might be KDE people, used to KDE. Changing what you’re used to is scary.

I don’t know anyone recently starting to use a Linux distro, that selects KDE. Everyone (which is kind of funny, since I totally agree) thinks it’s too bloated.

KDE is written by some people for themselves. Gnome 2.x is written by some people for the rest of us. Neither is of any good. That’s why I use xfce. This step to actually once and for all give it a try, let real people have a saying is great. Writing something for yourself (in the case with KDE) and writing code simply because you can, not because it benefits the project, might end up you liking your own masterpiece, but noone else does. Writing code that “you think other’s will like” (like gnome) and harmlessly removing options is probably as worse. Then no sane person could ever like it.

Some of the replies here comes like this: “I like Konqueror the way it is, don’t change it.” Well, someone not using KDE could never ever start liking konqueror, trust me. If KDE is meant to be used by KDE developers alone, then for god sake, keep up the ‘good’ work. Or open your eyes and let other’s in.

Even tho Microsoft is a major corporation with lots of money and stuff, they would never ever in the history of mankind succeed with this attitude; “We write what we like, and if you don’t like it, don’t use it. We’re leet, you’re stupid. We like dozens of options and we like to be able to configure things the way we like”.
I don’t believe m$ write a single line of code without asking real people first. Why do they succeed? Don’t give me the bs “they got the business hooked”, that’s just half the truth.

I’ve heard some really good things about KDE 4, but the attitude of the current KDE people gives me the feeling that KDE is as dead as gnome. Some new wonderful project will evolve that kicks ass, and then KDE people can continue to use their super-configurable system. I bet there are people who just can’t stand X. Terminal is ok. Terminal is configurable. X is scary. Get out of the closet will ya!

Mike A says:

Any fool can make things bigger, more complex, and more violent. It takes a touch of genius – and a lot of courage – to move in the opposite direction.

— Albert Einstein

Carl says:

I must object heavily to getting rid of embedding in Konqueror. The problem is not that opening certain files embedded in Konqueror aren’t doing everything that they can, it’s that the embedding applications aren’t written very well. Think of this, my grandmother is using Konqueror to browse the web. She has no knowledge of what a PDF file is or OOO documents or even what a web page truly is. Nor should she have to. However, when she clicks a link that points to PDF file, KPDF will start up opening the file. Why is this bad? Because she didn’t tell it to open and therefor has no idea what it is and where it came from. And in some cases doesn’t know how to get back to where she was. However, if she clicked that same link and KPDF opened embedded in Konqueror, to her it is just another web page.
As far as Konqueror as a file manager, I can see the argument to having it as a seperate app. However, it is nice to be able to switch to a web page from a current folder without starting up yet another app. I think a lot of the solution to embedding can be solved by a good sidebar. I propose KMetabar which is found here:

I’ve been able to contribute a few lines of code to this app so perhaps my opinion is biased. However, I think it is an excellent sidebar that can preview multimedia files and other previews on the side. It even offers an ‘Empty Trash’ option when in trash:/ (Partly thanks to me πŸ™‚ ). However, how do we get from seeing the preview in kmetabar to an embedded app in Konqueror? I think a nice GUI effect can come in handy here. For instance, I highlight a .pdf file and see the preview in KMetabar. I decide to open it embedded in Konqueror. When I tell KMetabar to open the file, the preview image expands and moves over to cover the filemanager view. This gives association between a preview to working with a file without the question, “Where the hell did this come from and what is it?”
Like you said, task management is a problem. So, the less windows, the better. The problem with Konqueror isn’t that it is a swiss army knife, it’s that it doesn’t have a perfect workflow. The use of a good sidebar (KMetabar) would be a great start.

Daniel D. says:

I for one support the separation of “Content Viewer/aka browser” from “Content unit management/aka file-manager” when it comes to Konqueror.

My best analogy to it is the symbiosis of Windows Explorer and Mozilla FireFox.

File Manager.
Explorer MANAGES (gives you access to) files, including allowing the access to ftp, smb and local through a hierarchy/tree-like structure.

Content Viewer
FireFox (coupled with pdf extension) is a perfect content viewer example. (Add to that office files support viewer and you get a universal browser – as in READ ONLY)

Content Editor
(A third level of this structure can be “Content EDITOR”. It can be a kpart type of embedded paradigm or a stand alone, specialized editors of today. You could get to this stage by clicking a prominent option to do so in the “Content Viewer”)

My reason for the separation of the tasks is that of the “length of life” issue. The need to navigate to a file is short. You get there, you pass a reference to it (file) to the viewing app (which has long-term focus).

Problems with current implementation:
Viewing a content over/in the same window as the file directory view (like viewing a readme file in kwrite kpart in the same window you just double-clicked the file) stops me from choosing other files in the same/other folder for parallel processing. Current behavior FORCES mono-tasking behavior onto user.

The file management vs. content browsing behavior of the empowers the user to separate TASKS from the WAYS TO GET TO THEM.

Zygo Blaxell says:

I too have observed that people do go a long way to avoid resizing windows, but I have a different suggestion for the root cause: to put it simply, people avoid managing windows because window managers suck. Please don’t blame broken WM UI on applications, and especially don’t try to fix broken WM’s from within applications.

I was an untrained user once. When I first started using GUI systems with modern window managers (including what is now called the classic Mac, DECWindows, Mwm, twm, olwm, olvwm, and several distinct things called Windows), the one thing I just couldn’t understand is why they seemed to all go out of their way to make window management painful. I remember what seemed like weeks where I was being painfully dragged into X environments that seemed to be precisely engineered to do useless and inconvenient things with windows. My first response was to avoid X entirely, and do everything I could with a whole lot of 80×50 VT terminals. It wasn’t until I discovered that I could reconfigure the window manager to behave more sensibly that the pain diminished until I could get work done (although to this day it’s still moderately painful to do some things).

One thing that struck me as odd was the way that simple activities like moving or resizing windows required hitting very small areas with a mouse, or very specific corners of the window (so if you have 3 corners of a window exposed, you have to find the fourth to resize), or they took up square inches of screen real estate with huge, useless borders and rows of buttons. The most mundane of window management operations–manual placement and sizing–was for some reason much more difficult than it needed to be, and the huge borders used by some WM’s filled precious screen real estate with utterly useless pixels. I never got used to this idea–it still feels just as wrong to me today as it did the first day I was exposed to it.

To this day, when I’m forced to use a braindead window manager (e.g under Windows), I mostly maximize everything and switch windows with the taskbar. The user interface for moving and resizing windows is so bad on Windows that I avoid doing these tasks at all.

Good management software, no matter what the resource, should actually *manage* things–that is, they should and implement or enforce some useful policy, and keep as many of the mundane details as possible to themselves. Unfortunately, window managers don’t actually “manage” anything–the best of them provide a clunky, awkward interface for shuffling through windows much the same way you would shuffle through sheets of paper on your desktop…if you had only two fingers, on one hand. If I wanted to be forced to use my hands to manually organize a pile of loosely related thin flat objects, I *have* a desk with a messy pile of paper on it already, I don’t need or want an interactive picture of one on my computer screen.

The only thing that most desktop software does, that my messy desk doesn’t do, is have papers that spontaneously jump to the top of the piles and attach post-it notes to themselves, demanding attention to themselves. I actually do want that feature on my desk, but I don’t tolerate it on my computer… πŸ˜‰

The solution is definitely in the window manager, not the application (although the applications should provide hints to the window manager so that the WM can implement a non-trivial WM policy). If we allow applications to simply grab focus whenever they think the user would want them to have it, or change their stacking order because they think they are the most important application running on the system at the moment, or set their size and position arbitrarily because the application author believes that they know better than the user where the application should start up, eventually the desktop descends into anarchy. Note my choice of the word “allow”–I have on occasion hacked things up so that window-manager applications can deny requests to change the stacking order of WM-owned windows, even with low-level X calls, because I believe that to be the bailiwick of the window manager and the user, not some application that has overestimated its intelligence.

One should be able to open a bunch of windows and have them sensibly arranged by default–position and size. There should be tools in window managers which select a bunch of windows and arrange them in one of several possible arrangements (NxM arrays of tiles, cascaded like folders, etc). Sometimes these arrangements should be “sticky”–once the screen area is divided in some way, windows can only occupy the established areas on the screen. This replaces having to move and resize windows with the task of selecting which windows occupy which regions. If you’ve ever seen the “ratpoison” window manager, imagine something like that on some desktops, while on other desktops a more conventional window management policy is used…and imagine it with mouse control. πŸ˜‰

Done right, a non-sucking window manager running a bunch of small applications should look like one application with a bunch of embedded plugins.

The orthodox file manager design (with the screen divided into three fixed areas with specialized purposes, as in Midnight Commander and others) is the kind of thing that a non-sucking window manager should be able to arrange automatically, given any three windows. With some hinting from the user (or a reasonable set of defaults shipped with the desktop), the non-sucking WM should even be able to figure out which windows are which.

Petr Janda says:

KDE should do away with the Knamingeverything scheme. I personally find it annoying in most cases.

Jamie says:

Personally, I think that KDE should do something about the fonts. They have always been illegible on Linux. Furthermore, something needs to be done to add some polish to the UI. i.e. the default background of the Windows is always this putrid dark grey colour. The KDE 4 team should make a real go of it and exceed OS X’s user interface standard.

WGi says:

The part about konqueror I can not stress enough … I have been preaching this for *years*. I I have watched this _plague_ spread over from windows with tears in my eyes. I have to admit, it was one of the major features that drove me away from KDE to XFCE+Firefox. I just hope so much that this disease will finally go away!

Jonatan says:

In my opinion Konqueror’s versitality is what distinguishes it from other browsers. If you don’t like embedding there is nothing to stop you from right clicking on a file and using the “open with” command. Hell, you can even check the little box saying “remember application ascociation”.

About configurations, I have to disagree. It is a huge feature to be able to change how your desktop environment behaves, and it extends far beyond just altering the colours of a titlebar. Take double vs single clicks as an example. I prefer using the mouse so I have KDE set to use double clicks and single click for selections, my computer scientist friend, on the other hand, does everything with the keyboard and has the mosue set to single clicks. Yes, it is frustrating to have to change the configuration to suit your own tatse, but the solution is not to force a default down the user’s throat. A better aproach is to try to place configuration settings in sensible places so that they are easy to find if you are not happy with them.

About the trash can I think you are probably right. I have still to figure out how to get an icon for it on my desktop.

Danni Coy says:

Ok I am a unashamed power user and I have spent a conciderable amount of time teaching people completely new to computers how to get started.

I seriously hope KDE does not ditch its power user base all together. There isn’t many places for us to go.

Recently I have started having people approach me to start using linux. Many of my friends have gone from “this is too hard” to “this rocks” in the space of a few weeks… My KDE setup is of course very highly tweaked. But here’s a rundown.

Kubuntu is the first KDE desktop that I would even think of putting in front of a mere mortal. Have a look at the default menu structure. That’s a sane amount of stuff to install by default.

Single Click is evil, Double Click is even more Evil. It took my parents quite a few months to develop the dexterity to manage a double click, Single click has a different set of problems (accidentally opening a file is too easy). I tend to favour large Icons (especially Image Previews) and like the idea of having overlays that let you click on a certain part of an Icon to open a file etc. This still has the problem of being unfamiliar.

Only unmaintained or universally unwanted options should go. Options should be split into essentials and tweaks. The UI for a tweak should be confined to one task and each tweak should be able to be turned on independently. Maybe using something along the lines of KGetHotStuff to enable the UI for various tweak (essentially giving a good description of what a tweek does and what you want to do with it).

There are lots of features in KDE that are really really cool but not very discoverable. Work needs to be done on the presentation of these features. Nearly everybody loves the ability to split konqueror into separate panes – its just that it would never occur to most people to do that. Same goes for internet short cuts. Nearly everybody who has seen me gg: has pulled me up to asked what I was doing and if they could do that in other browsers. KIO is also not that obvious.

Commenting closed.

Get started today with Qt Download now