DEC
15
2003

OSNews.com: KDE 3.2-beta2, Towards a Better KDE

KDE 3.2 Beta 2 was released last week for general testing and OSNews offers a preview of what is to expect from it early next year upon its release. The article mentions KDE's new features (faster loading times, Konqueror's Service Menus, Kontact, KPDF, Plastik theme etc), the problems that still plague it (cluttered KMenu and Konqueror menus, too many disorganized control center modules) and some constructive suggestions on how to get over the bloat without losing the functionality.

Comments

If there are computers, most likely there are semaphores too.


By Quique at Tue, 2003/12/16 - 6:00am

Of course, you have to balance that against the cost of not reducing menu clutter, which is reduced efficiency for the user. I'm not surprised that you got more people complaining about the menu reduction than supporting it --- people who want the feature have a much stronger incentive to contact you than people who don't. The overall response on the 'net, however, has been that most users (with the exception of a vocal minority) seem to support the overal simplifications that have been made in the Konqueror context menu.

And yes, you can configure it using an XML file, but that's not a great solution. The XML file doesn't really look much like that the menu looks like when it is rendered. Unless you're familiar with the mechanism, its hard to get the look you want. At least thats what I noticed when editing the Konqueror XML menu.


By Rayiner Hashem at Mon, 2003/12/15 - 6:00am

I agree with you Rayiner, I feel this *very* vocal 'uber-geek' minority is driving people away from KDE to GNOME. They may be happy to ghetto-ize themselves, but I think it would be a shame for the project. Personally, I have already moved over to GNOME, precisely because it is so much simpler to understand for the people in our organisation who are largely moving over from Windows. GNOME is catching up look-wise recently as well.

So, yeah I'm behind Eugenia 100%.


By Tom Holte at Mon, 2003/12/15 - 6:00am

"I feel this *very* vocal 'uber-geek' minority is driving people away from KDE to GNOME."

Yeah, right. That's why KDE is leading over GNOME in every poll I saw, right?

Everytime somebody talks about "vocal minorities" you can just substitute everything he said with "I just made it all up and I don't have proof for anything".

FACT is that KDE is a lot more successful than Gnome at actual users. GNOME is more successful in finding companies (like Sun) pushing it.

A cynic might say that KDE-bashers want KDE to "simplify" only because they want it to fail and fall behind GNOME....


By Roland at Mon, 2003/12/15 - 6:00am

More importantly, are we supposed to write code for people who don't provide any input and only exist in theory, or the people who get involved and vocalize what they want? It seems to me that the vocal minority are the people who are claiming that we are listening only to the vocal minority.


By George Staikos at Mon, 2003/12/15 - 6:00am

Let's use the Konqueror context changes as an example. A few people here on kdelook complained strongly about them. Yet, for every review and preview of KDE 3.2 I've seen, people seem to love the changes. There are definatey *real* users out there who like the simplifications. Its a fact of life, however, is that people are more likely to vocalize disapproval than approval. I know when Kwin III got all wonky in beta1, I submitted a bug report. Yet, when it was fixed in beta2, I didn't send an e-mail praising the fix. I was happy and appreciative, but I saw no reason to vocalize it.

There is, potentially, a problem here. Its hard to judge what the users want. It comes down to: do the KDE developers agree that simplification is needed, or don't they? After all, these are their applications, no? There is no reason to be heavy-handed like the GNOME developers were, but if KDE-Usability put up a page saying "we want simplifications; send patches here," I think you'd see a lot more involvement from the segment of the KDE user community that wants to streamline the interface.


By Rayiner Hashem at Mon, 2003/12/15 - 6:00am

>More importantly, are we supposed to write code for people who don't provide any
>input and only exist in theory, or the people who get involved and vocalize what
>they want ?

Let me vocalize a bit then:

I'am a real user.

I like the KDE framework (although I never contributed a line to KDE) far more than GNOME. I like the speed of KDE. Iwould like to use KDE full time.

But I am using GNOME to write this just because of better usability. That pretty much says it all.


By Bilou at Mon, 2003/12/15 - 6:00am

Pay someone to set up KDE the way Gnome works, it's perfectly possible to turn KDE into a Gnome clone just by changing setting and without touching any code, making it look the same, and with all the menu entries removed etc. But don't expect it to become the default KDE.


By Datschge at Mon, 2003/12/15 - 6:00am

> it's perfectly possible to turn KDE into a Gnome clone just by changing setting and without touching any code

No, it's not. Seriously, I'll PayPal you $1000 if you can set up a KDE desktop, with KDE apps, that after a minute or two of normal usage I can't easily distinguish from GNOME. Hell, I'd be enormously surprised if you could even take a single KDE screenshot that looks convincingly like GNOME (and I mean real GNOME, not some fugly Geramik version).


By damiam at Tue, 2003/12/16 - 6:00am

The look and feel wasn't the point. Bilou statet that he can't use KDE due to usability, and I state it's easy to bring down and limit KDE to the level of GNOME. This has nothing to do with whether you can distinguish it from GNOME after using it or after looking at a plain screenshot.


By Datschge at Tue, 2003/12/16 - 6:00am

Look and feel is a large part of usability. Look at Eugenia's screenshot and comments on Gnumeric and KSpread. There are a lot of little details that GNOME gets right and KDE doesn't. (Full Screen under the Settings menu in Konsole? What were they smoking?) A lot of this sort of stuff can't be fixed just by editing configuration options.


By damiam at Tue, 2003/12/16 - 6:00am

You should choose better examples than the ones you pick up since it's indeed possibly to fully customize the menu, it's a simple xml file for each app which contains the menu and toolbar structure. Apps' menu structure just have no GUI to be user editable since the KDE HIG states that a menu's purpose is to offer an alternate way to access all functions an app offers.

And regarding Full Screen under the Settings menu I agree that it's suboptimal. But I personally wouldn't put it in any other menu but instead make it part of the WM (reasoning being that I don't want the app to decide whether I can run it fullscreen or not).


By Datschge at Tue, 2003/12/16 - 6:00am

That was only an example of the poor design choices that have been made, not a concrete "Fullescreen in the settings menu is THE problem" statement. I don't have a Linux box in front of me at the moment. When I get home, if it matters so much to you, I'll find a better example.

Making fullscreen part of the WM is in interesting idea, but I don't see how it'd work. Generally apps designed to run fullscreen (media players, terminals, office products, web browsers, etc.) have a completely different UI layout in their fullscreen incarnations. Since that would already have to be hardcoded into the app, there's not a lot of point in making the actual fullscreen switch a WM option.


By damiam at Tue, 2003/12/16 - 6:00am

Yes, please get better examples. Input of any kind is always nice to have.

And I don't see any real problem with fullscreen being a part of the WM. Windows with fixed geometry can't have fullscreen, but the WM, having to handle that after all, knows about that. The UI layout, with respect to usability, shouldn't be radically different between a normal window and fullscreen, surprising the user never had been good usability.


By Datschge at Wed, 2003/12/17 - 6:00am

Even though many of us are completely happy with things as they are, KDE does have (minor) usability issues. If that were not the case we would not be hearing the things we are hearing. But what is hapenning, as it often does, is that some of these views are not being acknowledged, and a polarization is occuring.

Those of who adore KDE because it lets us set it up they way we want it, have the most to lose by letting KDE be tampered with; when the evangelizers point to MacOSX and Gnome as their champions, we start to get really scared, as though they were recommending castration!

But the KDE developers are a very intelligent bunch. I am convinced that if, despite any reservations, the KDE maintainers can take the message to heart, then the result will be a good one. Even though we might not yet be ready to admit it, a future KDE that is simpler to use, and as powerful, if not more powerful, will be something we will embrace once we can see that no permanent damage was done.

George mentions the issues with slimming down menus/context menus. That is a really tricky area! I followed some of the conversations on the Bugzilla lists, and almost went mad that people were advocating the removal of the options I use all the time. Of course, by the same token, I would happyily agree to the removal of everything you need most, provided I don't need it! There is a pattern emerging here: what I wan't to see, is not necesarilly what you want to see.

Win2K introduced the concept of hiding rarely used options. Provided options are not moved around, and the complete menu can be viewed again with ease, there are no negative usabilty issues with that approach. Although care needs to be taken with the algorithm used in the decision making process, this little bit of statistics is AI at it's easiest. If I use an option then I am going to want to see it, and if I don't, then I won't. No XML editing required.

Could we not move slowly through these issues by taking simple non-invasive approaches like these?


By Dominic Chambers at Tue, 2003/12/16 - 6:00am

As a KDE user since 2.x (I wouldn't dream of defecting to GNOME, I can't stand what they've done with 2.x) I want KDE simplified to enhance usability. The clutter on the toolbar and context menus makes them useless to me, because it forces me to scan through the toolbar to find what I want. If I'm going to be scanning anyway, I might as well just use the main menu, where at least things are spelled with with words and things are hierarchically arranged.

On my desktop, I have probably tweeked every single feature in KControl. Yet, I wouldn't mind some of the more esoteric stuff broken off into a seperate "advanced tweekers" app, if that would get more users and support for KDE.

Simplification doesn't have to mean a loss of power. Done properly, it just means that power is easier and faster to get to.


By Rayiner Hashem at Mon, 2003/12/15 - 6:00am

"The clutter on the toolbar and context menus makes them useless to me, because it forces me to scan through the toolbar to find what I want."

First, it's pretty extremist to call something "useless" just because of some unused icons, secondly, the whole point of icons is to recognize them easily and once you found the wanted icon you shouldn't have any troubles finding it again if it is in the same place, no matter how many icons are around it, thirdly you are arguing against your own point: EXACTLY BECAUSE THE DEFAULTS ARE NOT PERFECT FOR EVERYBODY, we need customizability. Don't like the toolbar? Configure it! If some self-proclaimed "usability expert" would say that toolbar-editing is only for the "vocal minority" and remove the ability to edit toolbars, you would be screwed, so be careful what you wish for.

I never said that the defaults can't be improved. I just say that under no circumstances remove configuration settings.

"On my desktop, I have probably tweeked every single feature in KControl. Yet, I wouldn't mind some of the more esoteric stuff broken off into a seperate "advanced tweekers" app, if that would get more users and support for KDE."

Some points:

-) How would that make anything easier? If you don't find your configuration setting in app A you have to look at app B or vice versa, it's getting more complex and NOT easier.

-) Different users have different fields of expertise. I am a programmer, but I have no doubt that an average secretary would need many more settings in KWord than I. So who's the "advanced user"? Please, get it in your head: There is no such thing as "perfect setting for everybody" and there are also no "experience levels". Every user is different. There is no "average user" and finding settings that "only average users will need" is impossible.

-) Getting help is getting harder. Novice users will of course ask advanced users, who will of course not know what options are in the novice settings and what aren't. Yet another layer of complexity and confusion.

-) The whole issue is overblown anyway. I just need to set settings *ONCE*, but I use applications *EVERY DAY FOR YEARS*. Streamlining applications AND ADDING MORE CONFIGURATION SETTINGS TO DO SO is much more important than streamlining configuration.

I have helped a lot of users with Linux and that's what the real problems are:

-) They just fear the unknown. They are willing to spend days, even weeks in solving some Windows problem, but will give up on the slightest Linux problem. KDE/Linux just needs more visibility and success stories. Marketing to put it in other words.

-) Hardware problems like "my USB printer doesn't work" or "my PocketPC doesn't synchronize".

-) And of course the biggest problem is software compatibility. Every Windows users has a lot of applications he doesn't want to lose. If you really want to get masses of Windows users to KDE, integrate Wine into it. Make sure Wine-apps are automatically added to the K-menu and make sure they work.


By Roland at Mon, 2003/12/15 - 6:00am

First, it's pretty extremist to call something "useless" just because of some unused icons
>>>>>>>>>>>
Not really. What is the point of the toolbar? The toolbar is there to allow you to quickly access actions without having to go to the menu. If it is just as quick for me to go to the menu as to wade through a huge toolbar full of extraneous icons, then it is completely useless. Its even more important for the context menus, where the icons are too small to really be useful as a guide. In that case, a menu with 18 items (like the original Konqueror one) is nearly useless (at least its still context-sensitive), because you have to linearly scan through all those items to use it.

secondly, the whole point of icons is to recognize them easily and once you found the wanted icon you shouldn't have any troubles finding it again if it is in the same place, no matter how many icons are around it
>>>>>>>>>>>>>
Then you must have a better memory than I do. Its easy to keep it all in your head when there are less than a dozen entries, but when you get Microsoft Office style monster toolbars, there is little you can do but search through them linearly.

Don't like the toolbar? Configure it!
>>>>>>>>>>>>>
I do. I spent hours getting every single KDE app I use regularly to have sane toolbars. I don't think other people should have to go through that though. I assert that the way I have my toolbars configured is more appropriate for the majority of users than KDE's default ones. I could be wrong, but the fact that toolbar minimalism is a part of GNOME's, Microsoft's, and Apple's HIGs lend support for my argument.

If some self-proclaimed "usability expert" would say that toolbar-editing is only for the "vocal minority" and remove the ability to edit toolbars, you would be screwed, so be careful what you wish for.
>>>>>>>>>>>>>>
Nobody is saying that we should listen to self-proclaimed usability experts. However, it would be wise to take into account some basic, well-understood, non-controversial GUI design research. This is where GNOME went wrong. Instead of smartly applying basic principles, they went hog-wild and tried to apply a single simplistic ideology (less is better) all over the system, even where it wasn't needed. We don't want to go the way of GNOME, but that shouldn't stop us from taking advantage of GUI design knowledge when its appropriate.

I never said that the defaults can't be improved. I just say that under no circumstances remove configuration settings.
>>>>>>>>>
I'm not saying we should remove configuration settings. Maybe move some KControl stuff to a seperate utility, but not remove them. I use pretty much every KControl feature. I'd hate it if they were removed. But I can put up with starting a seperate program if it'll make the mainstream happy.

How would that make anything easier? If you don't find your configuration setting in app A you have to look at app B or vice versa, it's getting more complex and NOT easier.
>>>>>>>>>>>
App A would have well-chosen defaults. It would be for users who didn't even know there was an App B. If they couldn't find the feature they were looking for in there, they'd just think it didn't exist. App B would be for advanced users. It really wouldn't be a problem of hunting between the two apps, because App A would only have a few options, and App B would have everything else. This mechanism works surprisingly well for MacOS X and Windows.

Different users have different fields of expertise. I am a programmer, but I have no doubt that an average secretary would need many more settings in KWord than I.
>>>>>>>>>>>>>>
I'm just talking about KControl. The basic desktop needs to be accessible to everybody, which is why it is worth it to go to a little extra effort to make something like KControl more accessible. Its not worth it for something like KWord.

and finding settings that "only average users will need" is impossible.
>>>>>>>>>>>>>>
Not really. First, there are average users. Most people using Windows machines don't do anything with them but launch games, run Office, check email, and surf the internet. Microsoft (and Apple) design the defaults for those users. They don't seem to be having any problems finding those settings, because its really not hard. And you're really in no position to claim otherwise, because they (as in Microsoft, Apple, and even GNOME) have done research on what 'average users' need and you haven't.


By Rayiner Hashem at Mon, 2003/12/15 - 6:00am

Yeah, right. That's why KDE is leading over GNOME in every poll I saw, right?

And do you believe that situation is going to continue?

Everytime somebody talks about "vocal minorities" you can just substitute everything he said with "I just made it all up and I don't have proof for anything".

Or you could substitute, "I have a problem here but everyone says it's not a problem."

FACT is that KDE is a lot more successful than Gnome at actual users. GNOME is more successful in finding companies (like Sun) pushing it.

Well, KDE is a lot older, and has had a lot more time to attract users, I will give you that much.

A cynic might say that KDE-bashers want KDE to "simplify" only because they want it to fail and fall behind GNOME....

Please don't think I'm a KDE-basher. My first linux desktop was actually KDE, and I had been using it since the 2.x days. In those days it was so much better than GNOME, but I see GNOME catching up quickly, especially in the area of user friendliness.

Now I realise this is a hollow talk, "get up and do something about it, I hear you say". I know I should but I have neither the skills, nor the time, nor (and this is the most important thing) the patience to argue with people who seem unwilling to see (what I see as...) a broader desktop world out there.

Maybe you're right and I'm wrong, maybe I'm right and you're wrong, only time will tell, but *please* listen to us, if only for your own sake.


By Tom Holte at Tue, 2003/12/16 - 6:00am

> Well, KDE is a lot older, and has had a lot more time to attract users,
> I will give you that much.

According to the history books, GNOME is 76 months old, whereas KDE is 86 months.
Makes a difference of 13 %, hardly 'a lot'.


By Bernd Gehrmann at Tue, 2003/12/16 - 6:00am

Well, at least I have put up some evidence supporting my point:

- MS Office is extremely bloated, yet popular among low-skill-level users
- OpenOffice is also extremely bloated, yet more popular than KOffice or Abiword on Linux.
- Computing hardware is becoming "bloated". Nowadays a lot computers come with serial, parallel, PS2, Firewire, USB1+2, LAN, several types of card-readers, a "game port", audio ports and loads of other ports, often even a redundant VGA-port. In comparison, the so-called "legacy-free" PCs and motherboards almost all were unpopular. Bloated won again.
- All distributions are "bloated" and include millions of features which won't get used in some particular installation, yet almost nobody runs a streamlined "Linux from scratch" installation.

Users love features (= bloat).

I made several points illustrating that. If you could come up with just one example of an "unbloated" application/product triumphing over an already established app because of it's lack of bloat, you would have some credibility, but like that it's just claims and hot air.


By Roland at Tue, 2003/12/16 - 6:00am

Okay firstly, I assume:

bloated == features / features you don't use

Which makes 'bloat' at least 50% subjective.

To a few of your points:

1) I'm sure you would accept that MS Office's market share is not purely based upon its feature set.

2) I agree at most 50%. OpenOffice has more features (useful to you or not) than KOffice or Abiword.

But I feel you are missing the point. I far as I can tell (and speaking for myself,) the people here are not arguing for removal of features, but for simplification of the *default interface*.

In terms of proof, just look at the default toolbars for an Office XP application compared to a KOffice application. The initial setup for Office XP is far less intimidating to new users, simply because they are not immediately innundated with buttons. As they become more familiar with the application, they can add toolbars.


By Tom Holte at Thu, 2003/12/18 - 6:00am

as a non-kde and non-gnome user (E0.16+Rox filer are fine for me, my computer is "only" 600 MHz), I can say that KDE made more good decisions to convince me to use its software and environment that Gnome in the past 2 years.
Many very useful features were removed from GTK and Gnome in the name of "HIG" (Holy Integrist Guideline?), and drove me away from Gnome (think about Metacity which has absolutely no functionality for managing windows as the most obvious example of bad decision). If only KDE had a lighter look as gnome... (yes "look": colour scheme, icons, toolkit appearance). And if only I could DnD from Konqueror to Gimp 1.3.x. (ok, I have tp check with the newest Mandrake if this works now)


By oliv at Tue, 2003/12/16 - 6:00am

>And if only I could DnD from Konqueror to Gimp 1.3.x. (ok, I have tp check with the newest Mandrake if this works now)

Jesus, it finally works!!!! I can close the old bug report about DnD not working between Gnome and KDE :)


By oliv at Wed, 2003/12/17 - 6:00am

It seems like you have just designed yourself a well defined task: write an XML GUI file editor for KDE. Then you would have a GUI menu editor that would make it easy to customize your menu. You have to realize that everyone draws their line in a different place and it's impossible to make everyone happy. We do provide the tools for everyone to make himself happy though.


By George Staikos at Mon, 2003/12/15 - 6:00am

> Icons on buttons are not on by default because not everyone has a computer as fast as yours. It costs to add those icons.

What, should icons be removed from toolbars as well? It only costs as much in toolbars as in buttons.


By anon at Mon, 2003/12/15 - 6:00am

No cookie for you. Pushbuttons, ex. on messageboxes, are created and destroyed regularly throughout application execution. Toolbars are created only once at application startup. Furthermore toolbars only show icons by default, not text and icons.


By George Staikos at Mon, 2003/12/15 - 6:00am

I like cascading menus, as long as there is only one submenu.


By Quique at Tue, 2003/12/16 - 6:00am

There's one thing I don't quite get: are the M$ "rarely used menu items" (the "v" at the bottom of a menu) really such a bad idea? Why on earth not have a powerful context menu and just expose whatever is frequently used? That way if you want to do something less-than-usual, you just expand it. Doesn't take too much time, does it?

Kuba


By Kuba at Tue, 2003/12/16 - 6:00am

Everything is subjective, and most of her suggestions would drive me nuts.

I think the KDE folks are surely capable of figuring out what needs to be done on their own. I am immensely happy with kde, and while there are a few small issue for me, I find that KDE is head and shoulders above any other environment, including Mac OS X. I just don't want it ruined by following well meaning suggestions such as hiding configuration options.

About the only thing I liked in her review is the spacing suggestions and Plastik as the default. _maybe_ the lines around the toolbar.


By TomL at Mon, 2003/12/15 - 6:00am

Most KDE users might now be able, but will KDE users in the future, who might have a lower skill-level?

KDE definately is cluttered in places where it doesn't really buy you anything, just loses you efficiency and ease-of-use. And I say that as a full-time KDE user. Would anyone, for example, mind the advanced options in KControl being moved out to a seperate app like in Microsoft Powertoys? GNOME clearly goes too far in the direction of minimalism (hell, even MacOS isn't that sparse!) but KDE needs to move more towards the center as well.


By Rayiner Hashem at Mon, 2003/12/15 - 6:00am

"Most KDE users might now be able, but will KDE users in the future, who might have a lower skill-level?"

So let me get that straight what you want:

You want to piss off existing users and developers to please non-existing "future" users who *might* like a dumbed down interface?

Well I think that sums the stupidity up pretty well.

"KDE definately is cluttered in places where it doesn't really buy you anything, just loses you efficiency and ease-of-use."

That's nonsense, simply because you usually change settings only *once*, but you use an application for *years*. So if you (oh my god!) need an extra 20 seconds to find some setting it's 20 seconds lost. If you lose 10 seconds every day because some smartass thought that the "average user" doesn't need that setting and (re)moved it, you lose an hour in the first year - and you get frustrated.

"Would anyone, for example, mind the advanced options in KControl being moved out to a seperate app like in Microsoft Powertoys?"

I would.


By Roland at Mon, 2003/12/15 - 6:00am

Hear! Hear!

I would be pissed off too if "advanced" settings (advanced for what? for whom?) would be moved to a different application. If people complain they can't find the options now, how are they supposed to find them if you split them over several applications?!


By André Somers at Mon, 2003/12/15 - 6:00am

Hold on. Nobody is talking about splitting things up over several applications. There would be a seperate "advanced" tweeking application for power-user features, like Microsoft's PowerToys. Its not several applications, but one.

Look: I take KDE's configurability to the extreme. I use nearly every option in KControl, all the apps I use regularly have had their toolbars reconfigured, etc. Hell, I regularly patch my themes to change things I don't like! Yet, I still can't understand why anybody would have trouble installing and using a seperate application for advanced tweeking.


By Rayiner Hashem at Mon, 2003/12/15 - 6:00am

Last time I looked, powertoys was a separate application from the normal Microsoft configuration (however, it has been awhile ^_^). I really don't see the point of splitting options that belong together into separate applications. Sure, split them over different tabs, even use an Advanced button if you must, but please keep the options together. I don't want to have to look in two applications to find the option I need, because I don't know if the author decided it's an advanced option or not.


By André Somers at Mon, 2003/12/15 - 6:00am

Its probably not as bad as you think. Under the two-application scheme, KControl would have only the most simple and obvious options: changing fonts, changing themes, changing resolutions, etc. The advanced app would have everything else.


By Rayiner Hashem at Mon, 2003/12/15 - 6:00am

You, distributions and everyone else are free to create as many option tweaking tools or levels as they like, you don't even need to be able to program to be able to offer customized solutions, just use kcmshell and add the components you want. If you don't know what components are available search for kcm_* in KDE's library folder.

Example (me using KDE 3.1.4 here right now, so components my have changed their name): You only want access to the desktop background picture, icon and keyboard setting, execute following command (or better create it as icon):
kcmshell background icons keyboard


By Datschge at Wed, 2003/12/17 - 6:00am

Very good, Roland! You worded this perfectly, and the point is right on!


By George Staikos at Mon, 2003/12/15 - 6:00am

You want to piss off existing users and developers to please non-existing "future" users who *might* like a dumbed down interface?
>>>>>>>>
Nobody is talking about dumbing down the interface. I am not, repeat, am NOT, suggesting we go down the GNOME path. I am, however, saying that we should simplify the interface in places that would make things easier and more efficient for everybody. Its not too different from what KDE is doing already. Does anybody really remember, or miss, the three or four options that were removed in 3.2's KHTML menu? If those changes are acceptable to the KDE community, I think that there are a lot more changes like that which could be made that would be equally acceptable.

That's nonsense, simply because you usually change settings only *once*, but you use an application for *years*.
>>>>>>>>>>>>>
In this respect, I'm not talking so much about settings, but extraneous toolbar and context menu items. These slow you down every time you use the menu or toolbar. Since these are configurable, if it really helps your workflow to have them, you can always add them, right? And by your own logic, the extra 20 seconds it takes you to add those options back is nothing compared to the efficiency gain over using that application for years.

I would.
>>>>>>>>>>>>>
Why?


By Rayiner Hashem at Mon, 2003/12/15 - 6:00am

"Nobody is talking about dumbing down the interface. I am not, repeat, am NOT, suggesting we go down the GNOME path. I am, however, saying that we should simplify the interface in places that would make things easier and more efficient for everybody."

If you argue about tweaking the defaults, I agree.

If you talk about removing/splitting/seperating options, I strongly disagree.

MS Office is probably the most bloated application in existence, yet nobody can claim that it isn't widely used. Users want features, features, features. Every user wants different features and in the end the only way to satisfy all users is to have a bloated app.

"Why?" [a seperate kcontrol would be a bad idea]

Because:

-) I love to scan kcontrol for new configuration settings and new features. I often found interesting features that way which I would have never found if they were hidden.

-) Searching for configuration options in seperate apps is more complicated than searching in a single application, it's just that simple.

-) It would not make KDE more attractive, to the contrary, it would needlessly add confusion and frustration.

It's not fair to compare KDE with Windows. We are not preinstalled on almost every desktop computer. We don't have any serious gaming library. However we can compare KDE with Gnome, Blackbox, FVWM, etc. And KDE is the *MOST SUCCESSFUL OF ALL* so KDE must be doing something right.


By Roland at Mon, 2003/12/15 - 6:00am

If you argue about tweaking the defaults, I agree.
>>>>>>>
I'm thinking bigger than tweeking, but nothing fundemental. Basically, I'd like to see the stuff that was done from 3.1 to 3.2 (especially to Konqueror) be applied to the rest of the system.

MS Office is probably the most bloated application in existence, yet nobody can claim that it isn't widely used.
>>>>>>>>>
I don't think anybody *wants* to use MS Office. We can do better than that.

-) I love to scan kcontrol for new configuration settings and new features. I often found interesting features that way which I would have never found if they were hidden.
>>>>>>>>>>>
So? Nobody is saying to hide them. They'd just be moved to a seperate app, and you'd scan that. New features wouldn't generally get added to the basic KControl.

-) Searching for configuration options in seperate apps is more complicated than searching in a single application, it's just that simple.
>>>>>>>>>>>>>
You have to consider the two types of users at work here. New users would just use the basic app. They'd not even be aware that the other app existed. Advanced users would already know the functionality of the basic app, because it would have only basic options and wouldn't change often, and would know that everything else was in the advanced app.

-) It would not make KDE more attractive, to the contrary, it would needlessly add confusion and frustration.
>>>>>>>>>>>
You can't be sure of that. Indeed, experience indicates that it has worked fine for Windows and MacOS.

It's not fair to compare KDE with Windows.
>>>>>>>>>>>
You're right, its not. We're beyond Windows. But Windows isn't a good target. Its GUI was never very good. It still isn't. Microsoft Office's interface epitomizes the whole "toolbar/context menu clutter" problem. Our goal should be to match the ease (and efficiency) of use of GUIs that people actually liked, like MacOS (Classic and to a lesser extent X), BeOS, NeXT, etc.


By Rayiner Hashem at Mon, 2003/12/15 - 6:00am

Drop it. This idea of separating config options stinks. It _is_ more complicated. I think everyone can agree that KDE usability can be hugely improved without pissing anyone off. This _would_ piss people off, including me. Since there is so much low hanging fruit, why even bother persisting with these other half-baked ideas? Until the simpler changes are made we can not truly determine the best way to proceed forwards.

There are four areas where I can imagine things could be improved in an incremental stepwise manner:

1. Solve the context menu issue (I like the Win2K solution of hiding items
that are never used anyway -- an adaptive inteligent system).
2. Perform usability analysis and figure out what should go where (toolbars
and app menus). See where problems occur and optimize the out-of-the-box
settings for the mythical average user.
3. Improve the settings panels in various apps and control center so that
they are better worded/organized.
4. Provide task oriented How Tos to complement the application manuals.
Something more advanced then this is a window, based on areas were users
actually get lost.

If there was a real will to do it, all of that could be achieved for KDE 3.3, and then, and only then, should we ask is we need to go any futher.

Just my penny's worth.


By Dominic Chambers at Tue, 2003/12/16 - 6:00am

I dunno how much low-hanging fruit there is. The developers made some pretty basic and obvious changes to the KHTML context menus, and there was an outcry over it. I think almost any simplification you do to KDE is going to meet with a lot of resistence.


By Rayiner Hashem at Tue, 2003/12/16 - 6:00am

I think there's lots. Context menus should certainly only show items when they are contextually relevant, and that isn't always the case with KDE 3.1. I made a lot of noise too, because people were proposing the removal of items I use all the time, so I can understand that.

But, and I have made this point twice already in this forum, and am feeling like a naughty repeat poster, they could very easily hide items that are never/rarely used (statistically speaking) like Win2K does for the Start Menu. That is a good idea that solves the problem, harms nobody, and is definitely low hanging fruit.

The same technique could be used for the KDE menu to solve the too many text editors dilema. Other possibilities include cleaning up some of the setings menus. Some are almost incomprehensible (e.g. some of KMail's dialogs) and many could be better worded or organised.

Measures like this are simple, if not technically, then politically, since they do no harm to anyone. Since they can be made before more drastic measures are taken, and since we will not no how much further we need to go until then, that is what we should do first. That, in my opinion, is the low hanging fruit.


By Dominic Chambers at Tue, 2003/12/16 - 6:00am

> 1. Solve the context menu issue (I like the Win2K solution of hiding items
> that are never used anyway -- an adaptive inteligent system).
Well... if they ever decide to implement this, I really hope it can be turned off. I absolutely loathe this feature in windows (the few times I'm forced to work with it). Really. If I click on a menu, I'd actually like to see that menu. I find it confusing that the menu doesn't seems to be complete.


By André Somers at Wed, 2003/12/17 - 6:00am

why not just have the "basic" controls as the first few controls in kcontrol? or perhaps stick them in a seperate tree that defaults to being unfolded? the rest of the options can be in their normal places, with the tyree folded up so there is no visible clutter for basic users.

another idea would be to make the kcontrol folders actually a kcontrol module themselves that contains the basic, most used options for everything in that section. that way you have the basic controls a singlwe click away, and by clicking on the icon for that module again, it unfolds and shows the rest of the modules.

just a few ideas


By ben klopfenstein at Tue, 2003/12/16 - 6:00am

I agree with some of the criticisms. Having used KDE 3.1.2 for a while, I felt the Kontrol center was not presented properly. I haven't used KDE 3.2 since my comp is down.

There should be a distinction between Superuser options (those that require the root password) and User options. They should be kept separate. For example the Login manager option (kdm settings) clearly doesn't belong to a non-priviledged user's option. What does he get from the option apart from a disabled dialog box (and an admin mode button)? Sure some are a bit difficult to decide, like font installer, which belongs to both places?!?. That could be decided, but the point is that it removes a lot of clutter from that one Kontrol center app, and we have a clear reason to differentiate them.

Overall, it give place for expansion of the Kontrol center. If someone writes a Kontrol center app to, say, set up hardware, we know where it should belong. On the other hand if someone writes a working integrated theme app (not the 3.1.x legacy crap, hope that's done with), we know where that belongs too. The Information on hardware etc. and perhaps kuser should be in superuser Kontrol center, while user password, user information etc should be in the user Kontrol center.

>Konqueror's context menu is a mess, why would I want to zip a web page or use Cervicia with it, is beyond me

I agree with that. The developer's viewpoint shows. If we want new people and non-developers to use KDE, that option should be shunted somewhere else. Zipping (or 'war'ing a web page) makes sense as an option, but I doubt the context menu means 'war'ing the web page.

I would like to hear if some of these changes are already in the works (as it so often is:-) Thanks.


By Rithvik at Mon, 2003/12/15 - 6:00am

Personally, I make heavy use of this. But it belongs in Location/Save, not as tool option.


By a.c. at Mon, 2003/12/15 - 6:00am

"But it is so damned easy to develop for this platform that some people seem to... over-develop."

I dont know if its true, but I like the sound of that. Maybe KDE should use it to atract developers


By Mario at Mon, 2003/12/15 - 6:00am

No sure about over-develop, but compared to anything else, ( Windows, Gnome, WxWindows ) It is the easiest.


By Leonscape at Mon, 2003/12/15 - 6:00am

Pages