As the leading Open Source desktop, many from the private and public sectors have expressed interest in KDE and are currently using it on a daily basis. A message consistently received from those users has been that KDE's usability could be better. While the need for greater usability is not unique to KDE, what is unique is our ability to directly and positively affect the usability processes in KDE via the same Open Source methods that have been at the core of its success.
KDE is delighted to have received invitations from corporate users to leverage the expertise of their own usability teams to help improve our software. In order to take advantage of these generous offers, the current usability processes in KDE were reviewed so as to discover how best to marry these new opportunities with the KDE project in general.
To facilitate these efforts a new moderated email list has been set up: [email protected]. Subscription to the list is open and the list will be publicly archived. The list will be moderated to keep the signal-to-noise ratio as high as possible.
The primary goal of this new email list will be to find, promote and implement methods to integrate ongoing usability efforts with existing KDE design methodologies. KDE developers will be encouraged and given the tools to include usability as a primary concern at the earliest phases of application development.
To provide additional support for these efforts, a new website is being developed that will become the incubator for a set of revamped and extended KDE UI Guidelines and house industry research papers, studies and findings in order to allow greater communication and cooperation between academic, commercial and Open Source usability efforts.
Comments
Nope, not a 'good joke' at all.
GNOME's technology may be nearing abysmal right now, but they ARE working hard (well, partly because there's no other way to get any kind of result with the GNOME API, granted), and now that they've got D-BUS and other KDE-inspired technologies to build on, they may very well begin to catch up. It's not a as much a matter of time, as one of man-hours, and given that Miguel de Icaza certainly makes up a certain lack of engineering competence with lots of excellent connections in the industry, accumulating man-hours is much less of a problem for GNOME than KDE.
See Evolution for an excellent example of what I mean. Totally HUGE piece of software, does loads and loads of stuff manually, that would be automatic with the KDE API, etc. Yet they (literally) poured millions of investor dollars into it, hired a bunch of developpers to work on it full time, and there you have Evolution: poorly built application that works pretty darn well.
In short, if we (KDE developpers and other contributors) remain sitted on our asses, other alternatives WILL eventually catch up.
So I wouldn't call it a 'good joke' at all, no. Because they're not playing in the same league as KDE right down doesn't mean KDE should believe itself the winner forever.
And how exactly is Evolution poorly built. Can you clarify that please? It seems you are stating an opinion on something you have no knowledge of, and immediately assuming that opinion to be true, and then coming up with a conclusion. That, my 'friend', is called "begging the question".
> And how exactly is Evolution poorly built.
Just guessing, it's not built from components? AFAIK its developers plan to break it up so you can start mail-only or organizer-only applications.
And GNOME/GTK is doing rather well for an environment that is really poor. I can now count at least 3 IDEs that use GTK or GNOME. Anjuta, Monodevelop, Scaffold, plus Eclipse which can be built using GTK.
GNOME technologies work, and that is why people use them, duh. And KDE ones work well too. This whole thing gets really ridiculous when people start making unsubstantiated statements like the one I am replying to. Do not fan flame wars by saying stuff that is unnecessary or outright false.
It's an irony of life that "integration" is the harder to do right the more benefits from integration there actually are.
thanks for giving me one more reason to work hard on KDE. who said ignorance had no value? ;-)
seriously, i appreciate your skepticism, but your broad-stroke painting of the KDE project is dead wrong. there are those who don't grok usability or even care much about it. that's true in every project, including GNOME.
there are many KDE developers interested in and working towards a more usable KDE. how do you think this project started, exactly? this emerged from discussions between several people at the Kastle conference in August; this isn't my brain child, but the desire of others currently involved in KDE to join in an make KDE better when it comes to usability.
the goal is to improve the processes, and we have more support to do just that than ever before.
It is great news. The Usability list sometimes look like a big mess with never-ending discussion rehearsing the same points every couple of months. But I've found that there was a sort of auto-organization process with people starting to do more work than talk, eg the work of Benoît Walter on KCM modules, the propositions of Frans Englisch and the toolbar effort of Aaron.
The new devel list will push forward this process of auto-organization. We need more people that take care of finding all inconsistencies in toolbar, menus. Critical eyes on help messages can also be very valuable to push clarity and consistency. If we have volunteers (don't need to be a programmer) who specialize in even tiny janitorial works on usability and believe in step-by-step improvements, we will really be more productive than endless talks about one-click vs double-click, MDI vs SDI. We will then succeed in open-sourcing usability.
> Changes do happen, but reluctantly, and only with a lot of kicking and
> screaming. Gnome do indeed have the right vision and seems to be genuinly
> committed do needed changes. So my guess is that KDE will never be able to
> catch up, and as Gnome gets technically better, it will be the winner of the
> desktop wars.
So Gnome has the right vision? Why? I think Gnome got blinded by Suns usability study and now think it's the Holy Grail, thereby alienating itself from both their developers and users. Of course, usability is important and KDE developers will need to realize you can't just throw every possible feature in a GUI. But still, though some changes may be painful for developers they are definitely welcomed by our users (seen the positive reactions to usability improvements in KDE 3.2?). As long as we've got good developers focusing on good features and others who will fit those features nicely into the GUI we will have a very balanced desktop. So no, KDE won't be able to catch up, we'll just keep our lead ;)
> the developers > a) Are in general opposed to do the needed changes
And the GNOME developers are different how? AFAIK most of the accessibility and usability patches are contributed by Sun and not written by the applications' maintainers.
> The style should communicate to the mainstream user and her values.
> style-elements that speaks to the office workers
For you the office workers are the mainstream!?
> For you the office workers are the mainstream!?
Indeed.. office workers are != mainstream.
Home users are far more important.
why, exactly, are they far more important?
Well, there´s more of them, I suppose.
> It is good that KDE tries to take usability seriously, but I am slightly skeptical. Out of experience from what I have seen here and on the mailinglists the developers
Did you participate in GNOME development pre-GNOME 2.0 (and post GNOME 1.4)?
There was a MUCH larger attitude against usability changes. Major flamewars were the order of the day.
This was at a time where KDE was against any changes that would hinder usability (for example, the 'hicolor' iconset was always default instead of prettier icon sets because it was easier to use). Sadly, I saw some of these decisions rolled back during KDE 3.0 and 3.1. 3.2 is a good step forward, and I think 3.3 will be even better.
Rock on KDE (and GNOME!)
> Out of experience from what I have seen here and on the mailinglists the
> developers:
> a) Are in general opposed to do the needed changes
> b) Lack the needed understanding and vision pursue such goals
Unfortunately, I have to agree that 'a' is true in *some* cases. I have often -- when all I expected was to start a discussion -- found unrelenting and irrational opposition.
However, I believe that 'b' is totally false. It is not a question of understanding and vision, it is one of methodology. The methodology needed is, in most cases, a discussion of the question. So, the only problem is 'a'.
An example of this is the discussion (or lack there of) of bug 73851 that has occurred on KDE-BugZilla and various KDE lists (and in private mail). When I started discussing this, I was not clear on the exact problem or the solution. It was only through discussion with others that I was able to clarify my thinking on the matter (my final proposal has now been posted on KDE-BugZilla). Such discussions need to be critical and pointed for good engineering to occur, but this would have gone much better without the 'you don't know what your are talking about' remarks. What is needed is rather 'just exactly what do you mean' -- by having to explain your ideas to others it forces you to refine them.
--
JRT
Doesn't anyone get the irony inherent in the phrase
"Suns usability study"?
From the folks that gave us CDE (jointly),
browser applets, the horrible Metal L&F, etc...
Gee! Good thing Gnome has those folks to point
to the right direction!
>CDE (jointly),
wow there. Sun was not behind CDE. That was HP, IBM, and Dec. Sun only came on board when it was apparent that they had lost with their own desktop (openview).
BTW, I was part of those wars back then. In those days, Sun was doing PR that constantly said that they won the desktop war. I would argue that nothing has changed.
"A message consistently received from those users has been that KDE's usability could be better."
I don't want to sound elitist here, but what the heck do users know about usability? Think about it. They are not usability experts. They may have a sense that the overall interface is somehow less than ideal, but they do not have the expertise to identify the real problems or submit correct solutions.
Users are going to request interface changes not on the basis of usability, but on the basis of their personal familiarity. If they're familiar with Windows, then their requests are going to be based on the Windows interface, which is a poor example of usability.
This shows the necessity of looking at the requests and complaints to figure out what the problem really is.
Someone may complain about something not working as expected. So the questions are what is expected? What actually happened? What were you doing?
It's amazing what you can learn if you listen to people. They may not know what is going on, may not be able to identify the problem, but they usually communicate that something isn't right. In my real life work, I do this all the time with comfort issues. The clients don't know what is wrong, but they know something is wrong. I ask them questions to ferret out the details that help me troubleshoot the problem. Then I fix it, and send them a big invoice for my expertise :)
It is too easy for 'experts' to essentially couch their own preferences and prejudices in complicated language.
I like what I see happening here. Aaron is trying to get experienced people to contribute. He is collecting data from users. And throwing in his own experience and knowledge. Then he takes all this, mixes it, and comes out with incremental improvements.
Derek
they know when it's hard to use or when things are inconsistent or when it confuses them. they may not know WHY, but they usually know WHEN it happens. =)
nobody said that the masses were suddenly going to shape KDE by democratically voting on UI issues or anything, simply that the need has been identified and we're doing something about it.
huzzah!
"they know when it's hard to use or when things are inconsistent or when it confuses them"
While usability is concerned with consistancy, among several dozen other factors, it is wholly unconcerned with difficulty or confusion by untrained users. Yes, seriously! They only reason its a factor with KDE is because somehow the notion has cropped up that it needs to be dumbed down to the level of the greenhorn newbie, and screw the intermediate and advanced users.
This may initially seem unintuitive to most people, so let me explain a bit more. Confusion among newbies occurs not because the UI is unusable, but merely because they are unfamiliar with it. And they think the UI is hard simply because they haven't learned how to use it. Just because a newbie doesn't know what something is for in KDE, does not mean that that something must be eliminated. These new users come primarily from Windows. If KDE does not exactly duplicate the Windows UI, then they will think that KDE is unusable. This is not true.
There are genuine usability concerns in KDE, but they should not be confused with newbie confusion. Heck, they were confused the first time someone stuck a mouse in their hand and pointed them at Windows, but they got over it. It may mean slower adoption of KDE by the unwashed masses, but it's worth the price to get it right.
p.s. I am not disparaging this usability effort. I am supremely heartened to see that it is involving real industry usability experts.
"Users are going to request interface changes not on the basis of usability, but on the basis of their personal familiarity. If they're familiar with Windows, then their requests are going to be based on the Windows interface, which is a poor example of usability."
not true, none of the games I know have windows interfaces, yet i never heard of someone having trouble with their user interface
The weakest spot by far is usability, though I have to say I'm pleased that it was significantly improved in KDE 3.2 and that it's on the road to more rapid improvement.
I just hope I won't have many usability nightmares, like those with Khotkeys 2.
Well done Aaron, this sounds really promising.
KDE usability actually isn't *that* bad, and in my opinion Gnome, Windows and Mac all have their problems too.
MS Word is a good example of bad usability - despite widespread use and familiarity, few users can get it to behave correctly. MacOS 7 was a good example of usability - very simple and yet techie users had some really neat keyboard shortcuts.
Oh, and I like the usability of Firefox/Thunderbird vs the original Mozilla.
This is all relevant; look around at what's out there, it's easy to get used to KDE, especially I suspect for developers. Windows XP is good for many users, and yet sometimes they achieve this by sacrificing security. Don't knock it, just try to better it - aim for the same usability without the security compromise, and see how close you get.
KDE needs simplification. I wouldn't mind if 3.3 had, if not a feature freeze, then a very high usability focus.
In Konqueror, if you select some text in a webpage and right-click, the menu has 14 items in it. That's a lot to look through if you just want to Copy. The Copy item is actually 'Copy Text', which is non-standard and less recognised than 'Copy', but the longer name is 'needed' because further down the menu is an option to copy the whole web page. Surely this latter one is out of context if you've highlighted some text?
If you select some text in a Kmail message and right-click, there are 19 items. None of them allow you to copy the text to the clipboard.
If most people have used Windows, and most other people have uses Mac apps, Mozilla, Open Office, etc. etc, then meeting their expections as much as possible *is* improving usability.
Also, the KDE Control Centre is horribly scary, and even as a seasoned Linux user I look at it and wonder if I can be bothered to go through so many pages. The Windows approach of having an 'Advanced' button is actually a good one here - but if you don't like it, maybe it would be best to come up with some other way to make it easy for the user to differentiate uncommon options from advanced.
Personally, I'd want to play with about a third of the options, and only dig deeper if I want something specific. For less advanced users, this would be more of a *need*.
There are other things too, but a lot of it would come down to cleaning up the look and feel further, lots of small refinements and efforts to make things more bug-free, fault-tolerant etc.
Please, let's have a seriously cleaner and more usable KDE 3.3... and any KDE developers who haven't used Windows XP extensively, I urge you to try and do so!
I suppose you will donate all the computers running Windows XP so we can, uhm, waste our time with it extensively?
> In Konqueror, if you select some text in a webpage and right-click, the menu has 14 items in it. That's a lot to look through if you just want to Copy. The Copy item is actually 'Copy Text', which is non-standard and less recognised than 'Copy', but the longer name is 'needed' because further down the menu is an option to copy the whole web page.
I'm planning to disable the kuick plugin that provides the "Copy To" for khtml though. The only reason that this wasn't done in KDE 3.2 was that I forgot to add support for popup plugins to check certain new context-related flags before it was too late (I didn't have kdeaddons installed)
> Surely this latter one is out of context if you've highlighted some text?
I was planning to implement that originally, but there are usability problems with that. Some people forget they've forgetten to highlight text. This is the same reason why you'll get the items for a page when you click on a image.
> If you select some text in a Kmail message and right-click, there are 19 items. None of them allow you to copy the text to the clipboard.
yes, kmail's context menus are a mess
If you are afraid of pain, you can not learn, and not progress. The guys behind GNOME were not afraid of pain when they started enforcing the HIG rules. They rubbed quite a few people the wrong way, but they are successful in that they have managed to make people change the way they do things. Now everyone who is developing for GNOME wants to follow their guidelines, because it makes things easier for them too.
Also, the rule enforcing one app per use is especially important. Take the Kedit, Kate situation. These are both very good apps, and there is a case to be made for including both of these in the DE, for some people at least. But if you are not a developer, you do not need these. So a choice needs to be made, and there will be pain.
> Also, the rule enforcing one app per use is especially important. Take the Kedit, Kate situation. These are both very good apps, and there is a case to be made for including both of these in the DE, for some people at least. But if you are not a developer, you do not need these. So a choice needs to be made, and there will be pain.
The /only/ reason that kedit still exists is that kate doesn't support BiDi. It will in the future (KATE_GOES_BIDI branch), and when that's done, kedit will be removed.
It has -nothing- to do with developers.
KDE is working, learning, progressing and changing just fine the way things are handled so far, no need for unnecessary additional forced "changes" and "pain". And no way there is someone being able to force something like that anyway. Your example of the for uninformed outsiders problematic looking Kate/KEdit situation shows that you really aren't in a position of being able to judge about how things are handled within KDE.
Just my two cents worth:
First, on this forum setup: In the 'Important Stuff' below, one of the items is:
>>Try to reply to other people comments instead of starting new threads
Yet when I click on Reply, all the other comments are gone unless I'm folowing a specific thread. Makes it kinda difficult to quote other comments. Couldn't the forum be set up using frames or something? As it is, I have to have reply screen open in a new tab....good thing konqueror & mozilla have tabs.
>>On KDE & useability. IMHO KDE doesn't really have a 'useablity problem', it is the currently most useable Linux GUI available. The GIMP's unusual ui has a useability problem, KDE is pretty easy to get on to no matter what OS you're coming from. This is not to say it's perfect, there is room for improvement but generally it's pretty good. Holler's post ('1st impression counts, 19/Feb/2004) lists some good examples of where improvements are wanted.
The 'problem' is getting improvements happening.
It would appear 'will' (post 19/Feb/2004) has a point:
>>... developers
>>a) Are in general opposed to do the needed changes
>>b) Lack the needed understanding and vision pursue such goals
Lars (Friday 20/Feb/2004) gives the example of revising the drop-down Konqueror menus. A quick search of the bug list reveals that this has already been reported as Bug 53772 (and several others, with over 300 votes). Bug list reports it as a wishlist item and it is marked 'resolved'. How is it resolved?, I have just installed the latest KDE 3.2 & still it has the same problem?
For KDE (and Linux generally) to progress, issues like this are going to have to be taken seriously and dealt with. This is one of the few disadvantages of open source and particularly volunteer labor open source. In a closed source environment, 'Sales' or somebody with authority can identify a problem (big or small) and enforce a solution but with open source there is no authority structure. This forum may (probably will) help identify problem issues but who's going to see that they get fixed?
>>by Gil: "As the leading Open Source desktop" Can we really be making statements like this?
Yes. To be the best you must believe you are the best.
and finally
>>by David Johnson on Thursday 19/Feb/2004
>>I don't want to sound elitist here, but what the heck do users know about >>usability?
You can't be serious. A 'user' is the ultimate test. If the user can't use the product, what good is it?
> Lars (Friday 20/Feb/2004) gives the example of revising the drop-down Konqueror menus. A quick search of the bug list reveals that this has already been reported as Bug 53772 (and several others, with over 300 votes). How is it resolved?, I have just installed the latest KDE 3.2 & still it has the same problem?
Please compare KDE 3.1's context menus with KDE 3.2's.
The Bug report starts out:"The context menu in konqueror-the-web-browser should *only* contain entries relevant to the currently selected element. "
Select text, right click in 3.1.3 gives:
Back
Reload
Open in New Window
Open in Backbround Tab
Open in New Tab
Open with
Add to Bookmarks
Actions
Select All
Stop Animations
View Document Source
View Document Information
Security
Set Encoding
In 3.2:
back (but not fwd)
reload
Copy Text
Select All
Bookmark this Page
Open With
Preview in
Actions
Copy To
Stop Animations
View Document Source
Security
Set Encoding
(It is of course different again in this editor box)
Admittedly the ability to Copy/Paste on right click is a welcome and very much overdue improvement but what are all those other things? How is 'Stop Animations','View Document Source' or 'Preview In' relevant to selected text?
Does this really solve the Bug?
As an aside, I would have expected Select text->rt click->preview in embedded text editor would have presented me with the selected text displayed in the embedded text editor but instead it gives the html code for the whole page. This is probably going to confuse a nOOb who will then maybe click the up arrow trying to back out of it and will soon become thoughly lost.
Yeah, I guess that hasn't been optimized yet. I'll take a look at. Compare KDE 3.1 and 3.2's context menus for links however.
I should also mention that reopening that bug report probably won't do any good; it has far too generic stuff in comments.
// Yet when I click on Reply, all the other comments are gone [...]Makes it
// kinda difficult to quote other comments. Couldn't the forum be set up using
// frames or something? As it is, I have to have reply screen open in a new
// tab....good thing konqueror & mozilla have tabs.
Konqueror also let you split the screen up into two halves, which would get you the forum "set up using frames".
// In a closed source environment, 'Sales' or somebody with authority can
// identify a problem (big or small) and enforce a solution but with open source
// there is no authority structure.
I doubt it's a problem of (authority structure free) open source versus (dictatorial) closed source environment. It's more a problem of many people agreeing that something should be done while there is no single clear vision but multiple competing solutions by those people how the end result should look like, leading to confusion what is actually wanted. As fault already noted, the report 53772 you are referring to is actually an excellent example of that. Solution to that issue would be encouraging commenters to stop watering down reports and instead commonly agreeing on a single solution, but seeing how controverse these specific cleaning up issues are this will be hard to achieve.
>> split the screen up into two halves...
So it does! I had completely forgotten about this feature. Leads me to wonder how it could be made more visible - perhaps an icons on the main toolbar??
This is one of the things I like about KDE...almost every app is chock full of really good features. Way ahead of Gnome.
BTW it is Windows, not Gnome which is the rival in the 'desktop wars'. The primary choice faced by any potential system installer is Windows system or Linux system. The choice of which desktop then defaults to the choice of which distro which will be largely influenced by other factors such as geography. Red Hat probably dominates in North America, SuSE in Europe. Each of these has their own reasons for their choice desktop, I doubt if useablity comparisons are very big as both are generally considered pretty much equal in this area.
It does appear that KDE could implement some improvements in the methodology of implementing improvements, however. At present this is pretty much limited to the bug reporting system, possibly something more like an open forum might generate more ideas and may be able to focus the ideas to get a consensus of what is really needed in the way of a solution. One problem is visibilty of the forum, even this one is not easy to find from the KDE homepage. I happened on it quite by accident. I knew about the forums at kde forums.org but I never would have gotten here from there. Maybe 'wish list' forum at KDE-forums with possibly a separate link on the KDE home page. Whatever it is, I think it would have to be directly and obviously visible on the KDE home page, if I'm a nOOb, that's where I'm going to start.
Well, how could we see "improvements in the methodology of implementing improvements"? I agree that for "n00bs" it might be preferable to hang around at kde-forum.org instead at bugs.kde.org, but for the rest of the KDE using/contributing/developing community to notice suggestions on kde-forum.org results there better be reported to bugs.kde.org for the suggestions not to get lost. So I have to disagree that limiting ourselves to bug.kde.org regarding suggestions is bad, actually I see it as the only way ensuring that all kind of reports and suggestions are tracked and searchable anytime without having to keep track of multiple sites, message boards, irc channels and mailing lists and whatever other place could be used for just that. The best solution would probably be that each forum, mailing list etc. has at least one person who regularly checks whether specific suggestions and problems mentioned at his place are already reported on bugs.kde.org, and when that's not the case he reports them there himself.
>>So I have to disagree that limiting ourselves to bug.kde.org regarding suggestions is bad, actually I see it as the only way ensuring....
Yes, probably you are right. It would be too much to expect developers to keep track of numerous forums, etc. Still, the present setup (maybe) has some problems in that
1. it is not easy for a nOOb to 'effectively' use the bug system. I speak from experience, I'm still pretty much a nOOb having used Linux for only about a year and I'm only just now beginning to feel like I can 'use' it as opposed to feeling my way around to see what this or that will do. The main problem in using the bug system is trying to figure out if your bug really is a bug, is a repeat bug, is maybe just a wish list item or is just a feature that you haven't figured out yet. Is is rather off-putting to a nOOb to be told to RTFM, especially when it is difficult read the FM.
2. the bug system is not very 'open' in the sense that the only people who will see your problem are those who maybe have a similar problem ie there are no 'browsers'. An open forum might trigger some discussion which could develop a more appropriate resolution to a problem, for instance the items list on the rt click menu example mentioned above.
As a total nOOb, I am/was/would be hesitant about using a 'bug reporting system' whereas I would be quite comfortable using an open forum, especially for something minor. Also, you would never get any 'first impressions' comments in the bug system whereas you might in s forum.
Anyway, that's about the best case I can present for a 'Bugs & Wish List' forum as an additional way of getting user feedback.
Anticipating a 'Do we really need an additional forum?' response, the answer is 'I don't know'. If the present forums are producing feedback which is actually reaching the developers and initiating action, then there is no need for another one but if this is not the case, then maybe there is a need for an 'official' forum.
You know what, I fully agree with you. ;) The problem is that reporting bugs and wishes which should be resolved by programmers are technical issues, so the tracking system will never be able to lose the obstacle that at least some basic technical knowledge is there for reporting such stuff.
As I wrote before the best solution would probably be that every forum has at least one person who has the technical basic knowledge and regularly checks whether specific suggestions and problems mentioned at his place are already reported on bugs.kde.org, and when that's not the case he reports them there himself.
It looks like kde-forum.org has a special forum called "Feedback - Post here for Ideas & Feedback for KDE and the KDE Developers" for feedback purpose already, containing suggestions etc which should go to bugs.kde.org. I will look if anyone is taking care of the job I mentioned before, and if not I might do it.
Actually, if you go "Settings->Toolbars->Show Extra Toolbar" you will have a an icon on the toolbar for splitting the screen and joining it up again. Also, you can edit your toolbars so you could put the icon on the main one if you preferred.