Christian Paratschek is a long term Gnome user who has has not looked at KDE for over two years. He has written about his experiences of testing out SimplyMepis for a week, comparing the two desktop environments and their applications and finding the areas in which KDE can tempt even a dyed-in-the-wool Gnome fan.
Dot Categories:
Comments
... as the writer points out Gnome's UI is considerably more consistent, simple and user friendly ... and this using C and GTK!! GTK/Gnome has great designers UI policy thouhg so it makes up for the horror under the hood.
KDE has for too long focussed on using the "hammer" of C++ OO tech, the Qt toolkit and the ever more featureful KDE-libs to make the KDE UI "powerful" ... but the situation is out of control. The number of distracting buttons and widgets with poorly aliased and poorly scaled inconsistent icons bouncing and flashing gizmos is truly silly. It is *not* "powerful" anymore. Really it *needs a major cleanup*. It's unfortunate that stating the obvious is taken as flaming by KDE fans ... it makes me think that KDE UI "designers" have completely different brains ...
quoting:
"The main advantage of Gnome over KDE is definitely the better menu structure and the rigid rules on interface design that the Gnome Human Interface Guidelines impose. KDE's menu structure is a big mess compared to Gnome's. ... The KDE team has to do something about this because it effectively worsens the usability of KDE. Take JuK or Kuickshow as a positive example and throw some more bloat out of the other applications. The KDE Control Center, for example, is a nightmare."
So so true ...
Umm apart from the glowing 3D buttons (no accounting for taste ... these can be changed by changing themes), the klunky clock (it's supposed to be like that) and a few blurry icons this screenie shows that KDE is pretty damn polished:
http://www.kde.org/screenshots/images/3.3/snapshot11.png
Note the control-center is not in the picture :-D
Hmm, I never understood what the problem with control-center is. I find it easy to navigate - and it does the job for me. (actually an important reason that I dont use GNOME is simply that I never managed to find the keyboard layout switcher in GNOME)
My largest gripe right now is Konqueror. I whish I knew to edit the context menu (the one that pops up when eg. right-clicking a folder) and I REALLY think that Konq would benefit from being split into two apps: One for web browsing and one for filebrowsing.
This would enable the preferences in "Settings->Configure Konqueror" to be split into manageable parts.
Generally, though, I find that KDE rocks incredibly and I count the days to 3.4 :o)
Comparing the GNOME-config-tool (gconf-editor) and KDE-config-tool (kcontrol), the GNOME version clearly an UI horror! The options that are reachable by a real UI are way to few for advanced users. Abd beginners barely use any option at all...
About the context menu - what anoys me most is, that there is a "delete" and "move to trash". I only know people using one of these. So why not make _one_ delete item and make it's behviour configurable. By pressing SHIFT while clicking on delete would use the other "version".
I don't think splitting up konqueror would be a really big benefit. But why not have konqueror show only the relevant configurations depending on the current "mode"? And maybe an additional button in the configuration dialog to show all options of konqueror.
> what anoys me most is, that there is a "delete" and "move to trash". I only know people using one of these.
Actually use KDE 3.3 and turn off the "Delete" entry.
> By pressing SHIFT while clicking on delete would use the other "version".
SHIFT+Trash=Delete is working in KDE 3.4
So what's the reason to keep "Delete" in the menu? Or will it be removed for 3.4?
Anyway thanks for the hint - I'll have a look for it (have to find out how to disable "Delete").
> So what's the reason to keep "Delete" in the menu? Or will it be removed for 3.4?
KDE 3.4 doesn't display "Delete" by default.
> I'll have a look for it (have to find out how to disable "Delete").
File Manager/Behavior: [ ] Show 'Delete' menu entries which bypass the trashcan
"KDE 3.4 doesn't display "Delete" by default."
Wow - once again KDE is "faster" than it's users. I love it :-)
Having superflous menu entries like this is why people knock KDE so much. Consider...
- Deleting directly by bypassing the trash can is not something you want novice users to do
- Advanced users can easily learn keyboard shortcuts like Shift+Delete - that is if they don't know them already (this has been the shortcut in KDE for this for many moons)
So why even have the option to put it in that menu?
Many moons? Shift+Delete has been working because there has always been a menu entry "Delete" with that shortcut. Try in KDE 3.3: Hide the "Delete" command, Shift+Delete will not delete anymore but trash. It's better in KDE 3.4 - and not "for many moons".
As far as "For many moons" - Shift+Delete has bypassed the trash since KDE 3.0 at least, and I suspect even 2.2 and before, but I can't remember that far back. It's irrelevant *why* the shortcut worked. It is fairly simple to make Shift+Delete work without the menu entry (and it seems that has even already been done for 3.4). So the menu entry is certainly not required, for anyone.
"So why even have the option to put it in that menu?"
Because it's used with exceeding frequency. In my case it's the single most used item on that context menu. I want a delete function with usability. "shift+delete" isn't as usable as a context menu item. The primary usability problem is that I have yet another keystroke to memorize, and one which may conflict with a keystroke in the focused application. A lesser usability problem is that I now have to remove my hand from the mouse/trackball, move it to the keyboard, type the keystroke, and then move my hand back to the mouse/trackball.
Keep the keystroke combination, by all means. It sounds damned useful for people who like to work that way. But keep the that delete option in the context menu. Yes, yes, yes, I hear your whining that newbies should not delete stuff (like my mom who though her 20Gb harddrive was full because she never emptied her trash), so make "move to trash" the default configuration, while granting intermediate and advanced users the option to change it.
> I REALLY think that Konq would benefit from being split into two apps: One
> for web browsing and one for filebrowsing.
People keep saying this and I ask one simple question: WHY????
Konqueror is a container and it already has two plugins one for HTML browsing and one for File browsing (more actually when you consider different formats). I think that some of what needs to be cleaned up in the KCMs for Konqueror are issues caused by a false division between File manager and Web Browser.
Perhaps some of this confusion is based on a false assumption that Firefox is ONLY a web browser. Last time I checked, it will browse your local domain just as well as it will browse the web. So perhaps another question is to ask that you please explain exactly what you mean.
I do not like the false distinction with the two View Profiles. I have tried to remove it from my desktop but have been unsuccessful. There is, however, something to be said for different Konqueror configurations based on the protocol (http, ftp, file, & etc) being used and/or the plugin being used. I would find this much preferred over the View Profiles.
I like the fact that Konqueoror can be made to have "multiple personalities". In my nebulous mind, I lump "View profiles" with "protocol", and it generally works the way I expect it to.
Does anyone know if Konqueror can have different "tab behaviors" for "View Profiles" or are "tab behaviors" global through out konqueror? The latest i've tried is 3.3, and this does not seem to be the case. Thanks.
DON'T SPLIT Konqueror, it's a fabulous application just like multiview was on the Amiga.
I like very much the idea that a brouwser can browse any type of information whatever it is (a zip, a directory, a database videos, musics or web stuffs).
At the end, it is data that is stored in different manners. Why having multiple browsers for multiple way of storage? this is useless. Having one APP for that is realy great!
At home, do you have one table for each type of job? Is your kitchen table only able to handle plates and nothing more?
This is really a stupid complaint to constantly have. Even Evolution has basic support for HTML rendering. If we renamed the darn konqueror web browser link to KWebBrowser and made a symbolic link from that to the "kfmclient openProfile webbrowsing" they would NEVER KNOW THE DIFFERENCE. Heck, you would probably have a dozen KDE reviews that would talk about "the brand new KDE web browser, now entirely seperate from the file manager."
Next time I hear this damn comment I am gonna yell "type /home into the location bar in foxfire and shut-the-hell-up!"
Bobby
Well, that screenshot show the busy Kicker. I mean, why have Kmail, Kword and Help there? Help can be accessed through the Kmenu already, as can Kmail and Kword. Why are Kmail and Kword important enough to have a separate button in Kicker, whereas some other app does not have a button? And why do we have "Home" there, when we already have it in the desktop and you can easily access it through the Konqueror as well?
The answer to that problem is to by default have only absolutely needed buttons in Kicker, and nothing else. And in my book that would mean Kmenu, Show Desktop and Konqueror (Konqueror is a separate app and it could be dropped from the Kicker. But, IMO it is a central app to KDE, so it could be there).
That screenshot also covers the background-image which makes the desktop look busier than it should be. A solid color or a nice gradient would do nicely. If the user wants more flashy background he could add it himself.
I mentioned elsewhere that I'm writing a document about KDE and it's UI. I have few simple guidelines I apply everywhere:
1. The UI should offer only the basic features by default. If the user is advanced enought to miss some of the more advanced features, he's advanced enough to add the functionality himself in the toolbars. Instead of offering everything at once and the user (newbies and advanced users alike) then needs to trim down the UI to make it seem less busy, KDE should offer the basic features by default (that still get the work done effectively), and the UI could then grow with the user as the user becomes more advanced with the system.
2. Don't show things that are not needed. Example: Why do we have bookmarks in the desktops context-menu? Bookmarks should be where they are used: in Konqueror. If we have bookmarks in the context-menu, why don't we have addressbook there as well? Fact is that bookmarks don't relate to desktop (that is used to house icons and acts as a background for the apps to run on) at all. They relate to web-browsing primarily. And even if some user does find the bookmarks in the desktop useful, 99% of the users have no need for it. Why clutter the UI just to please 1% of the users?
> Why are Kmail and Kword important enough to have a separate button in Kicker
You lost me reading here. You have no clue for what most users use their computer.
"You lost me reading here. You have no clue for what most users use their computer."
And you do? How do you know most user want to use Kmail or Kword? Why not some other apps instead?
My point was (which you apparently missed completely) is that KDE should not try to guess what the user wants to do. Right now it guesses that the user wants to use Kmail and Kword. An assumption that may or may not be true. But instead of trying to guess what the user wants to do (and clutter the Kicker while trying to guess the user) let the user tell KDE what he wants to do. And, in this case that would mean that the user adds the buttons to the Kicker that he wants to have there! I do NOT suggest that we do not allow user to add buttons or features to the Kicker, far far from it! What I'm suggesting is that KDE gives a basic, clean Kicker by default, and the user can then add features to it as he sees fit. Instead of trying to guess what the user wants, let the user tell KDE what he wants. Instead of forcing the user to trim down the Kicker, let the Kicker grow with the user and his capabilities.
Yes, many users will want to use Kmail and Kword. But many users also want to use Konsole, Quanta, Juk, Kate, Knode, Kstars etc. etc. Why don't they have buttons in the Kicker as well? Just where do you draw the line as to what should be in the Kicker and what should not?
Kicker also has a button for Help. And while Help is useful, the user does not use it all the time, therefore we do not really need quick-access button for it. Help is already available in the Kmenu.
"But instead of trying to guess what the user wants to do let the user tell KDE what he wants to do."
Isn't it already possible? A user can customize KDE at a hight level.
"... that would mean that the user adds the buttons to the Kicker that he wants to have there"
Which leads to an empty kicker and at the end would at the end be a "linux from scratch" for every user?!? No - proper defaults are a very good thing! Finding those "defaults" is the hard task. Curently KDE may offer to much as default. But an empty kicker is for sure no good default either!
"Instead of forcing the user to trim down the Kicker, let the Kicker grow with the user and his capabilities."
Isn't it easier to trim down, than to "bild up" for the user?
"Just where do you draw the line as to what should be in the Kicker and what should not?"
Welcome at the hard task of UI design...
"Help is already available in the Kmenu"
But that needs way more mouse movement ans clicks. Getting in "contact" with a new DE it's handy to have help "at your fingertips".
"Isn't it already possible? A user can customize KDE at a hight level."
Yes it is. But even with customization, KDE tries to "pre-guess" what the user wants. User then has to spend time correcting the wrong guesses made by KDE.
"Which leads to an empty kicker and at the end would at the end be a "linux from scratch" for every user?!?"
If you thought that I'm advocating "empty" KDE, then I apologize for not being clear enough. What I'm advocating is a clean KDE with sensible defaults that doesn't try to offer everything by default. "Empty Kicker" would mean Kicker with just the essentials. And I think "essentials" would be Kmenu, Show Desktop, Taskbar, notification-area (or whatever that thing next to the clock is called) and Clock. Other things that could also be there is virtual desktops and Konqueror. Everything else is just unneeded extra that could easily be added by the user if he so wishes. Of course you could also say that removing unneeded buttons is also easy, but why should the user waste his time on that? "Empty Kicker" would still let the user do his job effectively, whereas "Busy Kicker" would look cluttered and unappealing. And since Kicker is visible all the time, it makes the entire desktop look cluttered. Remember: first-impressions count!
"Isn't it easier to trim down, than to "bild up" for the user?"
IMO, not really. If you display all possible options and features, you make the app and UI busy and confusing. New user would be intimidated by multitude of buttons and features. Of course, there are speciality-apps that have (and need) lots of buttons and features (Blender for example), but we are talking about basic apps for regural Joe here. Such apps do not need multitude of buttons and context-menus that are as long as your arm.
If we expect user to trim down the app, we are faced with a fact that when the user starts the app, he's faced with an app that is cluttered, confusing and frightening. In short: it's unappealing. The user might give up at that point, instead of spending time trimming it down. But if we give him an app that has clean UI and only _really_ relevant features readily apparent, the user could still get work done, and the app would not intimidate him. As time passes, the user would customize the app to fit his exact needs. Difference is that while that customization happens, the user feels like he's operating a sports-car instead of gigantic 18-wheeler. He's using an app that doesn't intimidate him and he knows he can control it, instead of using gigantic app that scares the user.
"But that needs way more mouse movement ans clicks. Getting in "contact" with a new DE it's handy to have help "at your fingertips"."
Not really. Maybe marginally so, but it's hardly relevant. And couldn't we use that explanation everywhere then? Accessing apps through the Kmenu requires too much clicking and mousing around, why not put each of those apps in the desktop itself? Sure, we would have few dozen icons on the desktop, but at least it wouldn't take many clicks or mousing around, now would it?
> KDE tries to "pre-guess" what the user wants. User then has to spend time correcting the wrong guesses made by KDE.
There is no single "the user" whose preferences are to be correctly guessed.
"User then has to spend time correcting the wrong guesses made by KDE."
With "proper" defaults, it's not a lot of work. But it's a lot of work to "build up" everything up, if the default is "empty".
"Of course you could also say that removing unneeded buttons is also easy, but why should the user waste his time on that"
If kicker is "empty" the user has to add buttons to make KDE "usable" at all. It's always some work for the user done to customize a DE for his needs.
With nice defaults, many users (especially beginners) can use it right off.
"first-impressions count"
Well - what impressions do you get from a nearly empty empty UI? I get the impression that is does not offer a lot functionality (=no good tool)
"only _really_ relevant features readily apparent"
Well - that would led to a nearly empty kicker (no taskbar, buttons, clock, dockbar, pager,...). That's because "_really_ relevant" means what _every_ user needs. That would be the KMenu and nothing more I guess.
So it comes down, that there should be a nice default setup. But this setup should be rather simple but not "empty". This lets the user start right off, but still does not clutter to much and has room for improvements (more buttons for kicker).
I'd say KDE's default should be a little bit more "simple". But not as simple as GNOME's. The hard task is to strip the deafult's but leave the needed.
But no matter what the default exactly looks like, most users won't like it because they have to customize it ;-)
"With "proper" defaults, it's not a lot of work. But it's a lot of work to "build up" everything up, if the default is "empty"."
The default is NOT "empty". The default is an UI that highlights the primary purpose and features of the apps, instead of throwing all of them at the user.
"If kicker is "empty" the user has to add buttons to make KDE "usable" at all."
By "empty kicker" I refer to Kicker that doesn't have buttons for individual applications. It would still be usable. You would still have Kmenu and the works.
"With nice defaults, many users (especially beginners) can use it right off."
And that's exactly what I'm suggesting! And what I suggest is that instead of overwhelming the user with multitude of options and buttons (it does happen, I have seen it myself), they are gently eased in to the application. Really, I'm not suggesting a "KDE from scratch".
"Well - that would led to a nearly empty kicker (no taskbar, buttons, clock, dockbar, pager,...). That's because "_really_ relevant" means what _every_ user needs. That would be the KMenu and nothing more I guess."
Well, you guess wrong.
"I'd say KDE's default should be a little bit more "simple". But not as simple as GNOME's."
I'm definitely NOT suggesting a GNOME-approach here.
"The default is an UI that highlights the primary purpose and features of the apps, instead of throwing all of them at the user."
But it also shouldn't hide it's power...
The question is who are the "target users" of KDE. Advanced users, geeks, beginners, absolute novices?
"Well, you guess wrong."
For example you suggested to remove KPager. But Kicker without KPager is "worthless" for me. And so it is for every advanced user.
"I'm definitely NOT suggesting a GNOME-approach here."
For me it sounds like this. So could you explain me the difference?
"But it also shouldn't hide it's power..."
Like I said elsewhere, I don't have an easy solution as to how the apps should present the more advanced features to the users. But the current system which intimidates the users and makes the UI look cluttered doesn't work either. And even with the current cluttered UI, we STILL have features that are relatively unknown.
"The question is who are the "target users" of KDE. Advanced users, geeks, beginners, absolute novices?"
Answer: all of them. The system should not be intimidating to new users (like the current system is), but it should be flexible enough so it can grow together with the user.
"For example you suggested to remove KPager. But Kicker without KPager is "worthless" for me. And so it is for every advanced user."
I did not suggest getting rid of Pager, I was talking about application-buttons. And all of the stuff I mentioned for removal could be added later by the user if he so wishes.
"For me it sounds like this. So could you explain me the difference?"
What GNOME does is that they remove features because they deem that they are "confusing" to the users. I do not suggest removing one bit of KDE's functionality, I merely suggest that the UI is designed in such way that the primary features of the apps are highlighted, while the less used whiz-bang features are not. Instead of prominently displaying all possible features to the user, the UI should focus on few key features, while keeping the more advanced features available, but not "in your face".
"What GNOME does is that they remove features because they deem that they are "confusing" to the users."
GNOME removed features?!? No - they hide them. That's about what you suggest. GNOME hides them very well. Many features are only available when usinge a gconf-editor.
"Instead of prominently displaying all possible features to the user, the UI should focus on few key features, while keeping the more advanced features available, but not "in your face"."
I do not agree. Reducing the UI to the minimum that is needed is no good UI. It should be reduced to the most used functions - but no more ;-)
Reducing to a few key-features locks out many users from quite a lot of functionality.
>Reducing the UI to the minimum that is needed is no good UI. It should be reduced to the most used functions - but no more ;-)
Exactly.
That's why I proposed to add statistic 'click' counting of every menu items in KDE
to be stored somewhere on disk when the app close,
such that this info can be returned to developers
who could make proper choices and have this info displayed on some stats.usability.kde.org webpage.
Sincerely yours,
Fred.
"GNOME removed features?!? No - they hide them. That's about what you suggest. GNOME hides them very well. Many features are only available when usinge a gconf-editor. "
No, those are configuration-options, not features.
"I do not agree. Reducing the UI to the minimum that is needed is no good UI."
So, you want the UI to have more than what is needed? That is know as "unneeded clutter"
"So, you want the UI to have more than what is needed? That is know as "unneeded clutter""
I want powerful applications. Having every advanced feature hidden reduces power, until I customize the application. But as every user, I'm very lazy in customizing every application. How many application's toolbar have you edited? Most people just customize their desktop-image, but not a lot more. So it would end up in reduced power of KDE.
There is no black and white where black is clutter and white is the UI with only the needed functionality. There's also some grey between where clean and slick UI are.
> How do you know most user want to use Kmail or Kword?
It's not about specific applications but application types/use cases. I do know that many people use computers only as typewriter and for Internet (especially Email).
> Why not some other apps instead?
You don't really expect KDE to put Evolution as standard mail client there, or?
> Yes, many users will want to use Kmail and Kword. But many users also want to use Konsole, Quanta, Juk, Kate, Knode, Kstars
That are different "many" in orders of magnitude.
"It's not about specific applications but application types/use cases."
So where is Juk? Or Kalkulator? Kdevelop?
"You don't really expect KDE to put Evolution as standard mail client there, or?"
No, but if KDE has Kmail and Kword, why not have Juk (for example) there as well? Why does KDE have to have a button for mail-client in the Kicker in the first place?
"That are different "many" in orders of magnitude."
So, what is Kword doing there then? Why not Kate? Or Juk? Or are you saying that people write text so much more often than they listen to music that we absolutely must have Kword there, whereas Juk (for example) is not needed?
Again: it's all about defaults. Defaults any distro can change if it deems that appropriate for it's userbase. Personally, I think the default of having the KDE word processor and email application there is sensible, and removing these is *very* easy. Still, I'm very curious to hear a good argument about what should be there and what not, and *why*.
>So where is Juk? Or Kalkulator? Kdevelop?
>So, what is Kword doing there then? Why not Kate?
Actually there are compelling reasons. Except maybe for Kalkulator. It's mostly based on the type of users and usage patterns for the different kind of applications.
KDevelop are easy, since its a developer tool and not intended for the general users. Same for Kate, it's a advanced tool for editing text files compared to KWord which are a tool for writing documents. Users who need KDevelop and Kate in easy reach are also users with skill to add those applications to Kicker themselves.
Juk on the other hand are not added for reasons of usage pattern. You start it once per session or even let sessionmanagment do it for you, the rest of the time it lives in the system tray. There is no reason to have it in easy reach in Kicer.
"Why does KDE have to have a button for mail-client in the Kicker in the first place?"
Because about 99% of the KDE's users use email. And they use it very often.
KWord is not part of KDE, so it's not part of KDE's default (isn't it?) - maybe in your distribution's default setup.
There's no juk, kdevelop, ... button, because I doubt any of these applications is used by more than 50% of the KDE users.
"KWord is not part of KDE, so it's not part of KDE's default (isn't it?) - maybe in your distribution's default setup."
Well, Kword was visible in the KDE3.3 screenshot in the KDE-website (you know, the screenshot that sparked this discussion?). And, AFAIK, it's automatically added to Kicker if it's installed. My distro is Gentoo, and it offer basically the vanilla KDE with vanilla Koffice, and it has Kword in the Kicker by default.
Gentoo doesn't automatically add kword to the kicker.
Are you thinking of kwrite, perhaps?
"No, but if KDE has Kmail and Kword, why not have Juk (for example) there as well? Why does KDE have to have a button for mail-client in the Kicker in the first place?"
Sorry for flaming you but...are you stupid or what? It's simple: because the MASS uses email and so an email client. I think is quite obvious for everything. The mass using an audioplayer is really cut down when comes to corporate users. While emails keep a very strong point in this scenario.
And how do you know that most KDE-users are "corporate users"? Yes, email is very popular thing to do with a computer, but so are many other things. And besides, when I use email, I usually start it up once, and just let it run. Why do I need a quick-access button for that? I'm not constantly restarting it.
"Sorry for flaming you but...are you stupid or what?"
So, offering constructive criticism make me "stupid"? What a great attitude you have! I have spent quite alot of time trying to figure out ways to make KDE better, and I have spent time discussing them in a constructive manner. And when I do that, some armchair-expert comes along and says "are you stupid or what?". Maybe I should just shut the fuck up, since no criticism is apparently allowed?
I guess all those people complaining that KDE does not have sensible defaults and/or the UI is messy with way too cluttered toolbars and too large context-menus, are all wrong and they just haven't "figured it out"? I guess they are "stupid" too for voicing their opinions? I at least offered ways to improve things, but apparently that's "stupid" as well.
Fuck it.
Please calm down all. I appreciate the thoughts of *all* people on dot.kde.org so please don't stop. We all have a common goal so let's keep focus!
Ciao'
Fab
"And how do you know that most KDE-users are "corporate users"? Yes, email is very popular thing to do with a computer, but so are many other things. And besides, when I use email, I usually start it up once, and just let it run. Why do I need a quick-access button for that? I'm not constantly restarting it."
Ok, so if you're are not stupid you love to play...I mean that email client is a *common base* for every kind of KDE user. The corporate user was only used as example to blame your "we have to put a media player shortcut too, then".
And YOUR single experience is not meaningful at all. I, for example, tend to close very often KMAil or Kontact, and relaunch it when I need it.
So, offering constructive criticism make me "stupid"? What a great attitude you have! I have spent quite alot of time trying to figure out ways to make KDE better, and I have spent time discussing them in a constructive manner. And when I do that, some armchair-expert comes along and says "are you stupid or what?". Maybe I should just shut the fuck up, since no criticism is apparently allowed?
I guess all those people complaining that KDE does not have sensible defaults and/or the UI is messy with way too cluttered toolbars and too large context-menus, are all wrong and they just haven't "figured it out"? I guess they are "stupid" too for voicing their opinions? I at least offered ways to improve things, but apparently that's "stupid" as well.
The "stupid" thing was related to the fact you were (and are) still answering the same exact things when tons of other people point you out you're talking about a non-issue, or if you prefer, a special case that only match your personal tastes.
Good defaults are *definitely* the way to go (in KDE and in usability issues in general). And KMail on the kicker IS a good default, as it was a bad default having Konsole on it (and infact it was removed, thus with some criticism from CLI aficionados)
"And YOUR single experience is not meaningful at all. I, for example, tend to close very often KMAil or Kontact, and relaunch it when I need it."
I did not intent to say that since _I_ don't restart Kmail all the time, we shouldn't have the icon there. We have as many different ways of using KDE as we have KDE-users. Some like to launch Kmail often (like you) some only once (like me), some not at all. What my real point about Kicker and the buttons was that they make the UI look cluttered. And they do. And using Kicker by default as a tool to quickly launch applications is difficult since you can't really know what apps should be there and what should not.
Email is not IMO a "common base". Of course it's very popular, but so is media-playback. What metrics do you use to determine popularity of some certain function? How popular should it be in order to have a presence in the Kicker?
"The "stupid" thing was related to the fact you were (and are) still answering the same exact things when tons of other people point you out you're talking about a non-issue, or if you prefer, a special case that only match your personal tastes."
Clutter in KDE is a "non-issue"? To me that seems to be one of the biggest things in KDE people critique KDE for! No, I'm not suggesting removing Kmail (and Kword and Help) from the Kicker because it fits my personal usage-pattern. I'm suggesting it because we don't really know what apps should be there and what should not. removing them from the Kicker reduces the clutter, while not reducing functionality. The apps are still available just fine, they are not just thrown at your face when you log in.
Isnt this a problem with distributions? If some distributor thinks that KDE is bloated, mainly the menus, control-center and kicker, they can customize and deliver a better product.
An example: conectiva 10 brings mozilla, instead of konqueror, in kicker. Why? Because the users, in the beta period, wanted that way. I'm sure other distributions do the same.
And KDE is not far, by default, from your expectations, just take off kword and kmail. But i think that one navigator, one mail client and one editor is a good choice. As KDE cannot expect that the user has installed evolution, openoffice and mozilla, because of the dependencies, so they choose KDE programs.
You see, you say that multiple workspaces are important: for me they are not, because kicker works so nicely that this option becomes irrelevant.
My only flaming statement is: who took off the new tab button from the toolbar in konqueror ? Certainly some Windows ore Gnome advocate that says that KDE is bloated. I hate you :-)
"And KDE is not far, by default, from your expectations, just take off kword and kmail."
If you think removeing those two fixes the issues with KDE, you are far, far off the mark.
"If you think removeing those two fixes the issues with KDE, you are far, far off the mark."
No, for me its already nice. With kword and kmail in kicker. But you got the point, i'm sure about it.
E-mail is something 99% of people use on computers. For some people, it's one of 3 or so reasons they have a computer. Media playback is less common than e-mail, especially when you consider that a significant number of people probably don't download/rip their music, so they couldn't use juk for their audio anyway (they'd need kscd).
Also, juk runs in the system tray. This means it's likely to be started once, and then never again (assuming it doesn't crash, and you enable session management). Kmail doesn't run in the system tray (unless you get some kind of a tweak), so the average user is much more likely to be opening and closing it repeatedly over the course of their session.
"Clutter in KDE is a 'non-issue'?"
No. I think what he meant was that you're complaining about 'sensible defaults' and the fact that kmail is in the kicker, but that _is_ a sensible default, because _most_ people use e-mail. The fact that you don't understand that is what baffles most people here.
You're also not being consistent with your arguments. You say in one place that you're not arguing that kicker should be devoid of all shortcuts, but here you're saying that there's no way to know what programs are sensible, so we should leave it blank. Having no functionality other than a menu isn't very sensible, in my opinion.
Also, it _is_ possible to determine what programs are sensible by default. Look mostly at new users:
1. Most people use e-mail
2. Most new users won't be using 8 virtual desktops, so they won't want kmail open all the time, so they'll be opening it and closing it often, so it's good to have direclty at hand
-- So put it in the kicker
1. Not everyone plays mp3s
2. You only have to open juk once and it uses a miniscule amount of screen real estate to stay open, so people won't be opening it and closing it all the time
-- So it's not necessary to put in the kicker
As for help: 'This is my first time in KDE. Look at all these icons without explanations. Let me move my mouse over them... Oh, this one says help. I'll click it to learn some stuff.' It can be useful for new users to have a 'press here for help' button right away.
If you're an expert and know your usage patterns exactly, you can spend the 5 minutes it takes to set up all your own shortcuts. You only have to do it once.
"The UI should offer only the basic features by default. If the user is advanced enought to miss some of the more advanced features, he's advanced enough to add the functionality himself in the toolbars."
There's one problem. If functionality is "hidden", even advanced users don't see them. If thy don't see them, they don't know that it exists. If they don't know of it's existance, they won't search for it. If they don't "search" for it, they'll never add it. This even applies for advanced users. Only geeks or poweruser would look around if there is any feature that they may have missed...
Just have a look at GNOME. It's nice for absolute beginners and geeks. But advanced users complain about all it's missing features - although most of them are only hidden.
What you say is true. And I do not suggest that we reduce KDE's (or it's apps) functionality one bit. So we then have a problem as to how to tell the user about the more advanced features that are at his disposal. I do not have a quick and easy answer for that. But I do think that just about anything is better than showing them all at once. That results in cluttered UI that is difficult to navigate and gets in your way when you try to do the everyday stuff. KDE should focus on letting the users do the everyday stuff with zero distractions and with clean UI. The advanced features should come only as the user sees fit.
But, like I said, I have no easy solution as to how the user should be made aware of the more advanced features. Of course there's documentation and Kandalf's tips, but does anyone read those ;)?
"Of course there's documentation and Kandalf's tips, but does anyone read those ;)?"
Maybe I would read the tips if they showed up when a new version of a program is installed and show information about changes and new features.
"There's one problem. If functionality is "hidden", even advanced users don't see them."
One reason they won't find them is because they won't be in the documentation. As wonderful as the KDE documentation team is, they have a difficult time keeping up with the pace of KDE development. If the documentor isn't tracking CVS commits, most featurettes are going to be overlooked.