Slashdot Mirror
KDE Developers and Usability Folks on Cooperation
sultanoslack writes "Over at NewsForge a story just popped up on the usability experts from OpenUsability and some of the issues on working with KDE development teams, specifically the KDE PIM team. There's some interesting content on the different working styles of the two groups as well as a little bit on some of the improvements that were part of the recent KDE 3.4 release."
Just wanted to share my excitement about this.
I think it's great that the KDE Devs have no problem acknowledging that KDE could even be better if it focuses more on usability.
Don't get me wrong, KDE is far from the usability nightmare some folks want to make it, however it certainly has issues and it certainly can use some polish. (As can probably any other environment out there for that matter)
Now getting usability expert on board to solve these issues sure is the right way and if KDE 3.4 is anything to judge from, there are great things to come for KDE.
Rock on!
Maybe it's just me, but I wouldn't have the only screenshot on the whole kpilot page making it look like the thing barely works:
...
/dev/pilot does not exist /dev/pilot /dev/pilot
KPilot has been reported to cause data loss
Starting KPilot daemon
Daemon status is 'not running'
Pilot device
Trying to open
Could not open device
The thing might work great, but that screeny certainly isn't confidence instilling.
http://pim.kde.org/components/kpilot.php
Sure, I have a lot of choices under GNU/Linux. Too bad that for every choice I make I become more and more limited.
Bored? Browse Slashdot with a +6 modifier for Troll comme
I think the answer is the same -- good work is evident to others in both cases, even if the observer isn't an expert.
Speaking of enlightenment, have you seen Enlightenment 0.17?
Now that's new, different, powerful, and I can't wait until its done.
Raydude
Seriously, I totally agree with you. I think kprinter actually is a very good example of the problems kde faces. Kprinter is technically an awsome tool if you take the time to really get into it. However the problem is, for simply setting up a local printer it is way to confusing.
No, but your post is a very good example of the problems KDE, Gnome and the rest face, when it comes to adoption by Windows users. If something has a configuration option or two less than Windows does for the same thing, then it's not configurable enough. If it's got a configuration option or two more than Windows does, it's too confusing. I just pulled up the default Windows and vendor provided printer configuration panels on Windows 2k, XP and 98 SE and printer properties in Kprinter for KDE 3.4. Kprinter appears to have just a couple more options than the default Windows equivalent and a couple less than the configuration tools provided by Epson and Canon. The most significant difference I see is a matter of depth. In Kprinter and the Epson tool all the options appear to be presented at the same depth, a handful of tabbed pages in one window, while the Windows and Canon tools have buttons on some pages that open additional windows. The latter have most the same options as the former, they just hide more of them.
Making the world a better place, one psychotic episode at a time.
News flash:
both GNOME and KDE suck big fat balls when it comes to the user experience.
What a worthless test of usability...
You should take note of several things about that screenshot.
First, EasyTAG isn't a Gnome app, it's a GTK app. It doesn't use the GStreamer framework, or GnomeVFS, or GnomeUI, or GConf, so including it is incorrect.
Similarly for Abiword, that project has consistancy across platforms as a goal, and is also not a gnome project. It does go reasonably out of it's way to play nice with a Gnome environment, however. It should also be pointed out that the toolbar requirements of a word processor are very different than those of an email client, and I'd say having the buttons the same size across both would be less usable. This goes doubly if you're trying to win converts from Word.
The three real gnome applications in your screenshot each have 'File,' 'Edit,' 'View,' and 'Help,' along with two or three application specific menus. That seems about right. Also, the most commonly used/important functions from each set of menus is on the (similarly styled) toolbar, each with text beneath the icons(that's configurable system wide).
This random attack of yours might be useful if you actually made some suggestions about specific changes that would help, so let's hear some.
You'll find that most people who are doing programming work for free don't give a flying fuck about 'consumer behavior.'
Are you kidding?
What are you comparing it to? I have worked on a lot of large object oriented code bases (my day job is maintaining a LARGE finite-element code written in C++) and working in KDE is atleast an order of magnitude better than any large piece of code I can think of.
There are very well laid out docs that describe the core pieces of the system and tie it all together with the Qt heritage. The inheritance trees is KDE are very good... with just the right amount of inherit and extend mixed in with a good amount of functionality (some codes go too far down the OO paradigm path while others don't go far enough).
If you're not comparing it to other object oriented code bases then you are probably comparing it to Gnome... which, quite frankly, would be hilarious. Gnome is a HUGE mess (architecturally) compared to KDE. To do anything you have to use a huge kludge of disparate libraries that never follow the same design patterns (because they are external to Gnome). Most of the time you end up implementing your own widgets because the standard base isn't expressive enough (one of the reasons why toolbars look different in a lot of Gnome apps)... and in general it is just really tough to get anything done.
As for no-one ever cleaning up anything... that is just rediculous. KDE has gone through SEVERAL large cleanups... including a fairly recent one which switched everything over to stricter namespaces (to prevent collisions).
I am not saying that KDE is in any way simple... but there's no way it could be with all it does. It is a LOT of code, but I feel that they manage it better than most.
Friedmud
IFF they want consumers to use and appreciate their work, they need to.
Not saying they should or they shouldn't: But that attitude is why the usability on Linux is so bad.
Why yes, I AM a rocket scientist!
to paraphrase:
"...sure xfce4 "seems" more consistent & integrated than kde but personally it seems like nothing but a Mac OS X ui hack, looks just like it dont it? I'm all for choice but after hearing so much about how crappy the Mac interface is what do we get in xfce4, same old thing. I use kde exclusively and will never change..."
- just pointing out what has to be said: there's only so many ways to go in UI without a radical redesign of the computing world in general, and either you're going to look to someone like you're copying Windows, or copying Mac OS, or copying BeOS, or copying an Amiga. I like KDE, and I like xfce4, but let's face it, they BOTH have a lot in common with what's come before, but so what? Sure, fine, you can argue KDE is a lot like Windows, but you know what? I'll take my KDE over Windows any day of the week, and the more they improve, the happier I'll be.
-Jay
I saw a hack once for MacOS X Cocoa that allowed you to hold down some modifier keys and drag buttons around in an app to really reorganize them. In fact you could drag them to another application and put them there, or to a floating toolbar or a menu. That's the beginning step of user interface customization.
That kind of functionality may be cool for a geek who knows what he's doing, but it's a total nightmare from the helpdesk point of view. I don't think that this enhances usability at all. On the contrary; in the hands of a configuration-happy noobie, it would probably render the desktop unusable faster than you can spell out "reorganize".
Usability isn't about "user interface customization". It's about a well thought out interface that the end user will be able to figure out how to interact with efficiently, thus actually minimizing the need for user customization.
I used to be a sceptic. These days, I'm not so certain.
KPDF has seen such major improvements in usability in 3.4 that I was amazed, it is one of the best if not the best PDF readers in existence currently. Adobe could learn a lesson from KPDF. I really hope they wont follow the same approach as Gnome, just dumbing everything down and leaving the users who really need features like SCP over VFS, Tabbing and Splitting in Konqueror etc.. standing in the rain. But so far it looks very good. They did not dumb anything down, but understood usability to make a better ui but leave the power functions in (which can be locked out via kiosk if needed) One of the biggest problems Gnome had, was that they went the usability for idiots way and left their main base, which mostly are power users standing in the rain, the way, we take it out you will never have it in again.
Sorry, but no.
How would people that have problems with vision set a High-Contrast color-scheme? Or do you want to have everyone use one so that some 5% of all people cann see better?
How would blind people interact with your perfect immersive environments, without telling the environment so? How would deaf people who cannot hear sound notification tell the environment that?
Sorry, while your comment is true in that config options are not a solution to most usability problems, the absolut you use is not quite true.