Submitted by “KDE 4.1 will be what everyone expected 4.0 to be – a fully functional revolutionary Linux desktop. I took a look at the revision 777000 of this desktop environment and what you get is a visual changelog describing the current progress in terms of look and feel and the features.” It has Bejeweled KDiamond? I’m sold.
I still wish they would remove (or at least as an option) the plasma config thingy in the upper right corner. That is SOOO annoying. I have heard alot of people complain about this so I find it weird that it is ignored. KDE is suppose to be the “tweak as you wish” DE, but in its current state even gnome got more options it seems unfortunately.
I agree whole heartedly. There is time yet before the 4.1 release yet though.
I just have to say that although I disagree with this design choice it is not a personal attack on the Plasma developers. I have huge respect for them, and especially Aaron Seigo who is one of the people I respect the most within opensource.
Edited 2008-02-24 00:40 UTC
I would not call it an attack. But what purpose is the cashew once a user has set up the widgets he wants on his desktop? Especially does that purpose warrant tying up one of the four corners without any option to remove it. It served a purpose while presenting KDE4 as a technology preview but I am less convinced of its purpose on a production desktop.
Do you disagree with its technology or visually? From the visual aspect nothing is set in stone. KDE 4 will evolve during its lifespan and some technologies like Plasma may even make it into KDE 5 some day.
I still wish they would remove (or at least as an option) the plasma config thingy in the upper right corner. That is SOOO annoying. I have heard alot of people complain about this so I find it weird that it is ignored. KDE is suppose to be the “tweak as you wish” DE, but in its current state even gnome got more options it seems unfortunately.
Aaron is dead set against that for some unknown reason ( http://bugs.kde.org/show_bug.cgi?id=154535 ). You can still vote for the bug though, even if it’s closed (see comment #25). Then perhaps at least somebody writes the patch and distributions include it, regardless of what Aaron may think.
Thank you for the link, though I must confess it saddens me. If the developer can not explain or show why a request would not be considered then he is starting to lose touch with the users. I can understand that Aaron has a grand scheme that he is not communicating well and we “will see” eventually what that scheme is. However to not even entertain a patch to turn it off shows an inflexibility in choice that is distasteful to me.
I may lose some functionality that the developers consider valuable, but historically KDE let me make that choice. As an example in KDE3 I turn off all icons on the desktop as well as all tooltips and hover overs. This reduces the “functionality” of the desktop but it is the way I like it. KDE3 gave me the ability to make that choice. Indeed I was/am able to tailor most every aspect of my working environment (KDE3) to match my workflow. I dislike this direction of “we know what we are doing and where we are going. You are just along for the ride.”
Aaron, assume the end user is not an idiot and let him make the choice. Sure, given your uncommunicated scheme for plasma it will probably be best to have the cashew on, but let the user try and find out for himself. I try to minimize criticism of KDE4 because it is a work in progress and deserves a chance to shine. That said the lack of configurability, and worse the lack of reasonable responses from the developers why requests are being discarded. chafes and is beginning to fill me with apprehension.
Edited 2008-02-24 13:55 UTC
.. Answer attached to the wrong post. Please ignore.
– Gilboa
Edited 2008-02-24 15:29 UTC
Here’s something to remain optimistic: If it annoys enough people, someone will eventually release a patch.
Yeah, but the release of a patch doesn’t mean it will be accepted.
That thread made me a little bit sad. By the sound of it it seems like the KDE devs are starting down a we know best and giving the users config options are bad path.
When I first installed KDE4 on my laptop that ‘thing’ in the top right corner was the first thing that jumped out at me. And after finding out what it did, the first thing I did was try to find a way to remove it. On a 12″ screen it really does take up quite a bit of space and keeps drawing my attention, esp when I bring my pointer too close to it and it starts moving.
I hope this “we know best, so accept it and be grateful” attitude won’t be too pervasive in KDE4, since on the whole I think KDE4 show a lot of promise. I’m currently running KDE3.5 but with a bunch of KDE4 apps like dolphin, konq, ksysguard etc and I think they’re all a marked improvement. Also the KDE4+qt4 libs seem like an amazing step forward in ease of app development and I’m very much looking forwards to a chance to use them. I just hope they don’t trip on the goal line when it comes to the actual desktop environment.
I must agree.
The “we are right, the user just doesn’t understand our way of thinking” way of thinking was the source of GNOME’s unfortunate spatial-browsing-is-the-default-with-no-UI-way-to-change-it decision (GNOME 2.6?) that led me to KDE in the first place. (Thankfully the GNOME devs reversed the decision later-on and added the required UI support)
While I do understand that there are other urgent issues at hand, and having the UI to remove/disable/hide this control is less urgent then say, having a customized panel, this bug should have been postponed, instead of WONTFIX’ing it.
Other then that, KDE 4.0.x is coming along nicely. (At least on my rawhide machine)
– Gilboa
Yeah, this sort of thing is pretty much why I switched. It’s such a weird role reversal.
If I remember correctly, this also contains the zoom control, and that would actually be very useful if you for instance log in to X terminals with different screen sizes and resulutions, or use a laptop with a dockingstation, that have a larger screen atached to it than the one on the laptop.
So, I don’t think it should be removed totally, make it optional would be the way to go.
Annother, perhaps even better, way to solve the varying screen size problem would be to automagically zoom to a desktop size where all icons was visible regardless of screen size/resolution.
Or they could just make it so that xrandr works with KDE4… The zoom tool is of questionable value, especially when that could be replaced by an “effect” or key combo.
If I remember correctly, this also contains the zoom control, and that would actually be very useful if you for instance log in to X terminals with different screen sizes and resulutions, or use a laptop with a dockingstation, that have a larger screen atached to it than the one on the laptop.
Not certainly what 99% of people needs.
Me, I don’t understand why I need this on my KDE 4 desktop. It’s not just that I don’t use it. Is that if I use it (ie: I use the “zoom” feature), the desktop zooms. I mean, why I’m supposed to be able to use a zoomed out desktop?? They could at least disable it in 4.0.x and reenable it when it does something useful.
There has been quite a discussion about the zoomfeature on aaron’s blog, and to put it bluntly: it’s useless for users, and is only in there to find bugs in it. Which did work, a few bugs were actually found. The useful part might come with KDE 4.1, or even later on…
I’m an ardent KDE user– and I’m happy with the KDE 4 mantra of ‘sane defaults’ with ‘clean config screens’, and that thing needs an: ‘on’, ‘off’, ‘auto-hide’, and ‘auto-hide on widget lock’ option someplace. I don’t hate it, it just has to go away on my single head CRT setup because there is a good chance it will never do anything I’ll ever use. Why this is shaping up to be the GNOME trashcan debate of the KDE world I’m not totally sure…
Seriously.
Blackbox+Themed+Wallpaper > KDE || Gnome.
Seriously. Please preface opinions, because that can’t be a fact, because in my opinion, you are wrong
A character map? Ok..
Keyboard LEDs? Completely useless.
Luna Applet?
Seriously, couldn’t they come up with more useful stuff to include?
Keyboard LED applet? Just ask anyone with a Logitech wireless keyboard..
Very useful IMHO, but you won’t know untill you actually need it.
That said it should have an option to display text next to the virtual leds, so you know which one is which. Haven’t tested it myself, so I can’t speak to whether this is already an option.
I do have one, and I have never needed such tool. If anything, just a small “popup”, like the one in the XP login window would be more than enought.
Do the developers think that I’m going to care to take a look at the plasma desktop or the menu bar to look at this applet each time I need to type something?? Seriously.
I’m surprised to see this kind of things when there’s so much work to do in very very basic things like the menu. Look at that menu in 4.1 in the link, “OMG, transparencies!”. I really hope it’s not the final version, because I’m humbly waiting for a Plasma version that doesn’t makes me login out from the KDE 4 session.
OK, that’s your opinion. As for me – I absolutely hate that stupid popup thing on Windows and actually had to install a panel applet for it. Horses for courses…
Which is exactly why so many people want to make this an option, and not have it forced on them one way or the other. No one is saying that the KDE4 design is so horrible that it cannot exist. They just want the choice to change or remove it if they personally don’t like it. On the other hand Aaron is saying that the current way of doing things is so perfect that people should be forbidden from changing it. That is why people are getting upset.
Edited 2008-02-24 21:20 UTC
OK, then we are probably talking about different things. I have personally voted to reopen that bug ( http://bugs.kde.org/show_bug.cgi?id=154535 ) report and absolutely agree that the zoom/add applet should be hidden when your desktop is locked. As for the keyboard LED applet – nobody forces anyone to add it either to desktop or panel, so I am quite satisfied here.
(Edit – add bug URL)
Edited 2008-02-24 21:52 UTC
How about this: current trunk supports Apple Widgets. Pretty much all of them. So if there are in your opinion useful Mac OS X widgets, Plasma now has useful widgets.
Now this looks *much* better then 4.0. That launch menu looks so much better than the SuSE one. However it is very important to understand that “first impression” matters and lasts. Still what I would like to see is the desktop menu, right click on empty area on the desktop. Is there an option to “Line Up Icons”, “Sort Icons by Type”, “Auto Arrange Icons” — some of these options were available in KDE 3.x and how about that “Run” option in the desktop menu, this is extremely handy! I have to agree with one of the KDE developers, he said something like — “KDE 3.5 shows that we can produce some high quality ‘fine’ work” and yes I agree with that, for me, Mandriva is irreplaceable piece of work and I support it all the way. I have no doubts in the capability of the KDE deveopers, I know their work. This might make me consider to start experimenting programming for KDE. I have no doubt it is so much easier than messing with the Windows C APIs. I have done QT before, way back in KDE 1.x days. Despite my negative comments last time, I really, really love KDE.
And a bit off topic. Do you realise everyone got flamed? Microsoft released Vista and they got flamed, Apple released a new version of OS X and they got flamed too and now KDE developers released 4.0 same thing happened to them.
Edited 2008-02-24 02:34 UTC
And something else, on the link above, txf has posted the following comment:
I didn’t mind 4.0.1 but I found it awfully slow if not as unstable as 4.0.0 . What I’m MOST interested in whether I will still get horrible window drawing performance, or that dragging windows across my desktop don’t suck up cpu time, If 4.1 is stable and relatively fast then I’ll be happy. I am not expecting it to be as lighting quick as 3.5, but at least a usable speed
This is a very good comment! I would like to have this answered too. The window drawing performance was horrible. It was as if you didn’t have any graphic drivers installed on an LCD panel. Moving toolbars around was painfully slow. Has this been fixed?
Edited 2008-02-24 02:45 UTC
Most of those issues are due to X.org and bad drivers. The NVIDIA drivers are almost totally unable to accelerate anything (embedded Intel hardware with Free drivers performs better than the latest NVIDIA hardware – what a waste of money…) same goes for AMD/ATI. You simply have to bug your hardware provider for free drivers.
This is why we brought out KDE 4.0, btw – to expose the horrible issues in other parts of linux like X, GTK+, drivers etcetera.
I wonder why I love the ubuntu desktop, but I find KDE apps / infrastructure so much better.
Everyapp from KCompare to Amarok (KDE) to QT is better than anything in the GNOME world
Even the initiatives like phonon sounds amazing.
Yet before settling on Ubuntu, I actually tried Kubuntu, and I just didn’t like the desktop. It’s just too much for me. I love the simplicity of the gnome ubuntu desktop.
Makes me wish Gnome would let the KDE handle all the infra stuff, and the GNOME guys just design a good desktop on top of that
“I wonder why I love the ubuntu desktop, but I find KDE apps / infrastructure so much better.”
It’s a not uncommon sentiment – on the Ubuntu forums, where polls show that people generally prefer GNOME to KDE by a *factor of two*, we see rather striking results like this:
http://ubuntuforums.org/showthread.php?t=456040
“Plasma” seems to me to be the most concentrated effort poured into the “desktop” aspect of KDE in years and is built on a top-notch infrastructure – it will be interesting to see if it is enough to win some people over from the GNOME desktop environment over the next few years.
If they knew how to design a good desktop, they would have done it by now. *ducks*
The problem is that it is probably not possible to design a good desktop anymore, or at least it is extremely hard.
The reason for this is that users already have made up their minds on what a desktop should look like, and that is something that resembles the MacOS or Windows desktops. If it looks different from that, it will be considered bad and not used even if it, given a chance, would be much better.
E.g. the amount of information we have on our desktops gets bigger and bigger, but the main way of structuring it is still hierachies. If we compare with the development on the internet, we see how people go from navigating hierachies on yahoo to searches on google. It would be reasonable that the same thing should happen on the desktop.
Designers of future desktop will also need to think a lot more like ordinary office workers and less like sysadmins and developers. People in general are not interested in what they find in /etc, /dev, /lib, /bin,… They are just interested in that it works.
Users need to find, create and manage their own data.
Users need to find and manage what they share to/from other users.
Users need to find other users in a way that relate to everyday life, rather than to some cryptic user id.
Users need to find meta data about other users and data.
E.g. They need to express things like, “I want to share this info with Jane in the HR department” rather than to share it with some user doejane03 in group hr.
In real life we determine if we have the right person by looking at them, or by looking at meta data available about them, until we have enough info to determine that we have the right person, only if there happen to be two Jane Does in the HR department that look the same and have the same phone number,… we should have to resort to a computer related way of referring to her.
We also need to bind what we know about users and their data to actions and events. E.g. If Jane at HR calls you on the phone and you answer, the latest e-mails, and documents she sent or shared with you should be presented to you.
There also need to be better developer tools that can be used by people with more domain knowledge than programming knowledge. VB, c#, python and similar requires real programming knowledge. That way sysadmins should be able to create things ike workflows, or tie certain activities to various projects depending on what templates that was used to create various information.
E.g. If I create a new travel expence sheet, I should just be able to throw it in an outbox, and it shoulld be automagically turn up in my bosses inbox, and when he signs it it would continue to the right person in the financial department that grants the payment, and then the money should turn up on my account, no e-mail addresses involved in the process, and all of it should be programmable with more or less drag and drop by somebody who knows how our organisation works.
Functionality like this should not belong in a separate office suite, they should be handled at desktop and OS level so that all types of information can be handled consistenly.
To make all of this work, the desktop need to be multiplatform, and not just run on X11 but also on Windows and Mac. KDE is moving slightly in this direction by making applications easily portable and to have some semantic desktop features, but a lot more need to be done. Above all, desktops need to start talking the talk of non computer experts.
So, yes it will be interesting to see what the response on KDE4.x will be from Gnome, XFCE and other desktops.
To be succesful they need to think in terms of:
other users,
my information,
other users information,
information about me,
information about other users,
information about information,
actions, events, and projects or activities.
If developer can express these terms and avoid tying them to computer related concepts the free desktop have a great chance of being the best desktop ever built. This will not only have the effect of making it easier to understand to the user, it will also raise the level of abstraction so that new technology and ways of working can be incorporated more easily.
E.g if you say “send information to Jane Doe”, you don’t specify if it should be by e-mail, file sharing or a message sent as an sms to her cell phone, You leave it to the system to find out the best way of sending it. If some new technical means of communication should become available to us that could be used as well, without changeing the user experience.
In most regards, I agree with your posting, but I’d like to point out special problems that occur:
1. Computers are not easy. Not today’s ones. Applications, desktops and OSes are just abstraction layers to pure electronic circuits.
2. Users are not always intelligent (educated, experienced, …) enough to handle a computer. This is not to be seen as an insulting phrase, sadly, it’s the reality. No matter what you do, you need certain skills (e. g. driving a car – know how to drive and know the traffic rules; paint a wall – know how to handle a brush and wet paint; build a model – know how to use a screwdriver).
Today’s desktops assume certain skills to be present. Some of these skills are:
– realize categories,
– apply abstracions,
– manage associations,
– know the concept of hierarchies,
– know the concept of relations,
– know the concept of linear progress.
This does not apply to files, to any data, arbitrary objects, but to activities, projects or “stuff” in general.
I think it’s not neccessary to mention that these abilities, right next to the abilities to use spoken and written language (reading, writing), are essential parts of most societies in the world – they are their basal cultural means of communication and interaction.
The question for desktops is: How much of this “knowledge” can be assumed? Where does it make sense to give assistance, where does it annoy the user? (On other topics, the question looks like “How much can be dumbed down?”)
One important thing you mentioned: Users aren’t usually interested in processes, they are interested in the result. Applications, desktops and OSes should be part of the toolset that helps them getting their stuff done, but of course, it cannot do the stuff for them without any interaction. This is where desktop environments are important. Currently, they are not able to detect what the result of the user’s activity will be, so it can only provide possible means.
Let me add the following:
The first impression is the most important one, and it’s the most tightened one. So if a change of mind should take place, there needs to be a really “heavy” argument. 🙂
Please don’t confuse “people in general” with the average users you’re obviously talking about. People like system designers, administrators and operators have to know what they find where in the file system that is the fundament of any desktop. So even the desktop designers have to know certain things, and they have the difficult task to find a way to map a subset of them onto GUI elements.
As I said, abstraction.
But a computer is no real life, it’s not a representation of real life. I agree, it’s part of real life. And people don’t treat computers like real persons: Some even tell them their credit card number and very confidental information, assuming it would be safe. Please believe me: I really wish users would be more responsible with their data and think: “Would I do this in real life, too?” before doing the usual stupid things… The explaination is simple: While in real life a stupid action is followed by some kind of punishment immediately and feelabe, it doesn’t happen at the computer. (RL: “I burned down my house. How do I get it rebuilt?”; !RL: “I deleted all my files. How do I get them back now?”)
The concept isn’t new in fact, but it is extremely complicated to predict what the user expects. Just assume you’re sitting on a document you’re writing, the phone rings, the screen flashes and displays some holiday photos – would you like this? 🙂
Yes, this would help. I think you cannot leave the one without the other because software development is still tied very much to programming concepts and source code. You need to know much about the people you’re writing software for in order to create good software. In a closed market (e. g. testing software for psychologists) it is easy, but for Joe Q. Sixpack…?
This would require standardized and unified interfaces for nearly everything. Good approach, but only possible with team players.
I think that’s happening in KDE.
It will, I agree, and I’m interestedly following the development.
In my opinion, most desktop developers try to think this way, but they get confronted with the barriers of established concepts very soon. In some cases, users expect things to be complicated so they would not be able to adopt to a more easy way of doing things. One example: Microformatting in documents: Instead of thinking “this is a headline, this is a paragraph, this is a table”, users think “this needs to be +5 pt and bold so it’s recognized as a heading, here I need a line break between each line, and here I need to put in many many spaces to adjust the items so they occur in columns”.
Whenever you delegate choice to the system, the content of the delegation should be chosen wisely. The system would have to “know” much more than the user does…
Without any intention, feel free to read http://www.rinkworks.com/stupid/ in order to know more about certain users. 🙂
Actually, from my experience as a consultant in many large organisations, I feel that this, quite often, is the case. If I want to send a message to a collegue, he could be in a meeting, work from home, or travelling across the world to our branch in India.
He is likely to be the one who knows the best way of conacross the world tacting him at any given time, and he could tell that to the system. In fact to some extent he probably does that allready by, programing the switch board with a voice messeage or redirecting his calls to his cellphone.
Combine that information with who he is meeting, who is trying to contact him, both sender
and reciever would be more effective at their work.
It is also clear that the desktop of tomorrow needs to be able to react to much more than just keyboard and mous interactions. Phone systems are obvious things to integrate, other things could be GPS info from cell phones or info from building security systems (that could physically locate the one you want to contact, and hopefully make a good decision on how).
I also think that the future desktops will be much more about communication than they are now. Today they are basically a electronic combination of the file cabinet, ring binders, and type writers of the 1930:s
in most offices. With a better utilization of the information there allready is, and by adding some more
they could be so much more.
Then, there is of course the home desktop to consider, but given all the patent and DRM stuff that the media industry currently insists on, I think it would be a better strategy for FOSS to start making the free desktop the best office system there is. Once that has happened people will want to be able to work from home, and when that happens the media industry will find ways to make money out of these home users.
I had my hopes in KDE 4.1 but now I see I’ll have to waith for KDE 4.2.
You don’t think you’re being a little foolhardy considering 4.1 isn’t going to be released until July?
This is just a “so far” preview.
Can the panel now be resized? Also are they going to include more options for configurability of the desktop and other pieces of functionality in this release? How many feature regressions are they planning to still have from KDE 3.5.x for this release?
I am pretty sure they added that in 4.01.
It wasn’t added in 4.0.1, but openSUSE and Kubuntu backported the patches from plasma in trunk to make it happen. I think the KDE devs backported it for 4.0.2, though.
When KDE has a list remotely as broad as this:
http://www.apple.com/downloads/dashboard/
Then I will be impressed.
Apple Dashboard widget support is being worked on by Aaron and Zack, literally as we speak. Well, maybe not literally – but when they wake up, in a few hours time
So it shouldn’t be long now, especially now that one of the dependencies (Qt 4.4, with integrated Webkit) is now in KDE SVN trunk.
Qt4.5 is when Qt gets Cocoa.
I’ve got KDE 4.1 Trunk. The Qt integration of WebKit is minimal, at best.
http://developer.apple.com/macosx/dashboard.html
WebKit Widgets are not Dashboard Widgets that leverage Quartz and OpenGL directly. They rely on CSS/Javascript/XHTML and the Canvas and other elements implemented in Safari.
WebKit for Cocoa/OS X is much broader than WebKit for Qt and to think otherwise shows you aren’t looking at the same WebKit source trunk from Apple as I am.
What does any of that have to do with the price of tomatoes?
Support of OSX widgets was already coded into KDE4, it was waiting for the official integration of webkit into Qt, and now that KDE trunk moved to 4.4 last week, Aaron (and Zack, I believe) are working on making it happen.
Some of the OSX widgets that rely on embedded OSX binaries or OSX hooks for part of the functionality obviously won’t work, and aren’t targeted to, but the ones that are basically webkit-oriented HTML, CSS and javascript should be easily implementable.
At any rate, let’s just wait and see what they come up with. It’s speculation until they do.
I moved from KDE a while ago, I had some dispute with the developers and I must say that I am still not feeling any urges to go back to KDE. What did they do with it?
Everything takes up real-estate, the new menu is one of the most awful things I’ve seen and I still can’t see any good reason to switch to it.
It is just my opinion but it looks weird and I will not use it in this state, ever.
Sorry.
Exactly! This is the one thing I don’t like about KDE. Everyone else, windows included, seems to have figured out the value of real-estate on the screen. KDE has not. Other than that I just love it.
“Everyone else, windows included, seems to have figured out the value of real-estate on the screen.”
Can you explain why you need screen estate on a desktop with no applications running (which is what you seem to imply)?
I can only see the “screen estate” jibe as being an issue within a browser when you’ve got a plethera of toolbars removing space required for the webpage.
Personally once i’ve selected the menu, that is what i want to see because i am looking for something on it, the desktop behind it is irrelevant.
That’s my reason for griping about those plasmoids that seem useless. They take up so much room for something with little value.
I wonder if they tested those things on a 12″ 1024×768 laptop screen? I can see it possibly working on a 24″ widescreen, but on my laptop so many things in KDE4 took up far too much space.
As someone with a 24″ widescreen, I tell you that it *still* takes up too much space.
I stand by my point, the KDE team either lacks graphics designers, or those they have suck a lot.
What’s wrong with the new style? The new lancelot menu is really ugly and seems to be a bad vista imitation but otherwise I think KDE4 looks much better than KDE3.
Edited 2008-02-24 12:55 UTC
I don’t think that is the problem. The problem is rather that there is too little communication between the graphics designers and the usability people. Graphics is often judged just from the artistic point of view, and that is not enough to make good desktop design.
One example: the icon to chance fonts is a a double T.
Does that stand for true type? or is it an antenna symbol?
Just changeing the T to an A would making the icon much more distinct, and at least we would avoid confusing it with an antenna symbol. From an artistic point of view the A and the T could be equally beautiful, but the A gives better usability.
See: cognitive psychology: identification of prototypes (rather than arbitrary examples) of object categories.
If “TT” stands for “true type”, it would be okay, but the icon represents the font setting in general, so “font” or even “text” would be the category to be represented. Here, “A” would not only be a better chosen example, but it would be a valid prototype of “font” or “text”. “ABC” would be possible, too, but to much elements within the symbol.
All I’ve got to say is that, for the most part, I hated the way the new desktop looked from the screenshots. However, after using it I’ve come to like it, including the widgets. I think they’ve taken it in a new and right directions.