LWN: Comments on "Seigo: building brand together"

Not so long really

nix — Thu, 02 Jul 2009 22:20:40 +0000

So KDE is just like the stuff proprietarys software companies turn out,
proprietary in every way except that all the developers' identities are
public and the developers are easily reachable and you can build it
yourself and the source is, uh, open and it's full-blown free software.

Just to get that cleared up.

(FWIW I find KDE quite transparent: the codebase is a pleasure to dig
through, clean and neat and everything in its place. I suppose you have to
be reasonably happy with C++ to be able to hack at it or grep it usefully.
But just compare to Mozilla or, God forbid, OpenOffice for a nasty C++
codebase.)

Not so long really

Digit — Thu, 02 Jul 2009 21:20:59 +0000

KDE4 has had the same effect on me as winXP had. both from kde3.5 to kde4, and from win 98 to xp, drove me further away from relying on others to build my os for me and steer it into somewhere that leaves me boggled as to who actually think it's welcome and worthy progress. when it was win98 to xp, it drove me to open source, away from corporations, now with kde, it's driven me out to the likes of openbox on crunchbang linux. getting me closer to playing with the code... or so it feels. i actually am getting, for the first time, allowed close to the code, encouraged even (and i'm still only talking about config files). where unlike previously with kde, and windows before that, the feeling was more distant, and disconnected, the view of it's inner workings more obfuscated by something opaque.
i now appreciate, and feel like i understand the value and importance of the idea of elegant tight code, and unbloated operating systems through and through, so to better afford the end user to get to know their software right down to the kernel and hardware.
but to do that without trying.... that is my dream ideal. software that even noobs picking it up can immediately come to learn it all. transparent software. not opaque. so the userbase is inherently brought up to speed of software developer elite, as expediently as possible.
to the user, kde seems to be heading more towards opaque, dare i say, like proprietary. though have a look at the kde forums, their brainstorming section... beautifully innovative? mmm. they do encourage the userbase in other ways though too,)
so although KDEQt is not the direction i would have prefered it be taken, i am still none the less quite gratefull to kde4, because. it has encouraged me to get more with the spirit of open source. :) of freedom software. of empowered people.

If you are under "great deal of stress" then you are doing something wrong

khim — Wed, 10 Jun 2009 08:48:54 +0000

Except that you are completely ignoring the fact that they do have other options. Users who've signed contracts committing themselves to indentured servitude to the KDE devs would clearly be the category you describe. As a ballpark figure... how many KDE users do you think are in that category?

Rough estimate? 90% or may be more. KDE will keep existing users for the same reason Windows keeps existing users: it's the devil they know. Sure, KDE4 is quite different from KDE3. The difference is on XP->Vista scale. And like there the new beast is different and unfamiliar, but all other choices are totally alien! The people who are complaining that KDE4 broke their favorite toy from KDE3 will not be happy with GNOME, XFCE or whatever - they will need lengthy adaptation period. Sure, some of them will endure bigger pain just to prove that KDE4 is DOA, but majority will choose lesser pain - and lesser pain here is switch to KDE4, not to WindowMaker or ION...

If your users are not happy, by and large they will leave if there is something else that works better and you compete in an environment where users are allowed that choice.

But the problem is: "everything else" is worse then KDE4! Not because other desktop environments are bed, but because KDE4 only broke 10% of things users took for granted in KDE3 and switch to GNOME or XFCE will break 20% of things users are using today (and "truly alternative" DEs will be even worse). 20% is two times worse then 10% so user will grumble, write hate posts on forums but will keep using KDE.

My job is to keep the users happy by facilitating them in doing their jobs with a minimum of stress and difficulty. I put myself through a great deal of stress in accomplishing that. And KDE4 is about as alien to that goal as I can imagine.

Beed there done that, moved along. I'm sorry for you (especially in today's environment where employers are ready to punish admins at the drop of hat), but you took the wrong position. You can not fix all issues users create. Ever. Users are too creative. Fix what you have time to fix and drop the less important issues. Don't worry too much about users - people are creative creatures, they'll manage somehow. Otherwise you'll eventually burn out and this will not add happiness to users. KDE4 is not yet ready to replace KDE3 in big company - like Vista is not ready to do the same with XP, but as distributions will start dropping support for KDE3 it's time will come. Sooner or later. Don't worry about KDE4 till then. And if KDE4 will be requested by majority of users before that time - all the better: you can direct holdouts to KDE4-lovers and remove stress from yourself.

Both, of course...

halla — Wed, 10 Jun 2009 07:49:12 +0000

"Debatable. There are other desktops out there which might have fewer of those glitches."

Irrelevant. The qualities or lack of qualities of other desktops have nothing to do with KDE being
usable or not. KDE stands on its own and is and has been since 4.1 a perfectly usable desktop for
many users, which is proven by the fact that there are many users using it and being perfectly happy
with it.

Both, of course...

sbergman27 — Wed, 10 Jun 2009 03:02:13 +0000

"""
New users:
Bascially KDE 4.2 is usable. Some glitches here and there, but...
"""

Debatable. There are other desktops out there which might have fewer of those glitches.

"""
Old users:
Even simpler. They have no options. KDE3 is not an option (it'll bitrot and die) and switch to other deskstop environment is bigger issue then switch to KDE4.
"""

Except that you are completely ignoring the fact that they do have other options. Users who've signed contracts committing themselves to indentured servitude to the KDE devs would clearly be the category you describe. As a ballpark figure... how many KDE users do you think are in that category?

As to the rest of your post... If you make users happy they will stay with your desktop. If your users are not happy, by and large they will leave if there is something else that works better and you compete in an environment where users are allowed that choice.

As a system administrator for corporate XDMCP servers myself, I can only reel in shock at your talk of keeping machines running to the exclusion of keeping users happy. My job is to keep the users happy by facilitating them in doing their jobs with a minimum of stress and difficulty. I put myself through a great deal of stress in accomplishing that. And KDE4 is about as alien to that goal as I can imagine.

Troll?

dkite — Wed, 10 Jun 2009 01:54:14 +0000

Please. What he described is real.

The dolphin scrolling thing was a response to overscrolling that made
dolphin unusable with a scroll mouse.

You are falling into a trap that is very common, and that I see in myself,
especially with stuff I write. After using something a few times, you
avoid paths that don't work or cause problems. Shortly thereafter, you
forget that the problem existed.

The fact that you don't see them anymore doesn't mean they are not there.

Derek

Troll?

nye — Tue, 09 Jun 2009 10:13:55 +0000

That particular issue is one which is still present and refuses to go away. It happens regardless of whether desktop effects are enabled. Anyway, I can live with it.

I'm going to presume that your previous reply was posted rather like mine - it was late at night and I regret not giving it more thought.

The thing is, there are *so many* complaints that people have about KDE4 which are so casually dismissed as if they don't matter, because it's 'just one little thing'. This is what annoys me so much because I'd like to see KDE4 being successful, but I don't believe that will ever be the way.

FWIW I do use KDE4.2 on a daily basis since it's good enough for me, but it's at least a couple of years away from being something I'd recommend for a non-technical user.

Both, of course...

khim — Tue, 09 Jun 2009 08:54:14 +0000

So... are we talking about KDE *gaining* users? Or KDE *not losing* existing users?

Both, but separately.

New users:
Bascially KDE 4.2 is usable. Some glitches here and there, but it usable. Bugs are sad, but will be fixed over time. The problem with "lost features" is not a problem: user never seen "lost features" so it's not a big deal!

Old users:
Even simpler. They have no options. KDE3 is not an option (it'll bitrot and die) and switch to other deskstop environment is bigger issue then switch to KDE4.

Does it mean users can happily be ignored? Sure as hell not. But the long agage stays: "If users are made to understand that the system administrator's job is to make computers run, and not to make them happy, they can, in fact, be made happy most of the time. If users are allowed to believe that the system administrator's job is to make them happy, they can, in fact, never be made happy."

Applied to KDE the principle becomes: "If users are made to understand that the KDE developer's job is to make tasks doable, and not to make them happy, they can, in fact, be made happy most of the time. If users are allowed to believe that KDE deveoper's job is to make them happy, they can, in fact, never be made happy."

That's why requests of users in form "I did this and that in KDE3 this and that way - how can I do the same in KDE4?" are discussed and answered (the broader "this and that" the better - question "how to rip audio-CD in KDE4?" will be answered faster then "how to disable Kickoff recently used tab?"), but requests in form "KDE3 had this and that, KDE4 does not, please fix" are just closed with WONTFIX resolution. This is correct and proper behaviour.

KDE4 is not KDE3. If some real-work tasks are impossible to do in KDE4 - it's big deal, if somethig exist in KDE3 but not in KDE4 - it's not a big deal. Please take a look on KDE4 from this viewpoint and suddenly it becomes much more finished and polished. Not perfect, but not a disaster you are saying it is.

Troll?

pyellman — Tue, 09 Jun 2009 03:51:50 +0000

Um, I saw the graphical glitches you describe (slivers of whatever is underneath the window) when I first ran my new installation. I disabled desktop effects, and they became rare. Then I installed the NVIDIA driver for my chip, and they've gone away completely (even with effects enabled). I'm not really a spinning cubes type of guy -- if I had to, I'd live without the desktop effects, but if your the type that submits a bug entry when the morph from spinning cube into shiny cylinder doesn't go just right (sounds like you are), you might want to check your video driver and perhaps boost your video horsepower.

Peter Yellman

Troll?

pyellman — Tue, 09 Jun 2009 03:40:40 +0000

No. YOU are wrong. The fact that audiocd predates KDE 4 doesn't shoot down my observations, it is in fact the basis for them. Investigate the blog at the entry I pasted; search for "audiocd". Also, google "kaudiocreator unmaintained". It will take a lot less time than what you have invested in your silly rant.

Your latest posting is unmistakable evidence that your feelings about KDE 4 left the rational plane quite a while ago; posting what looks like the entire KDE 4 bugzilla database is asshattery in the extreme. That being the case JUST STOP USING IT. Don't parade your asshattery around.

You are a datapoint. I finally went to KDE 4 -- 4.2.2 -- about 3 weeks ago. So far, it is working great. GREAT. Performance is definitely better than 3.10 (even with effects, which I don't really need, enabled on a low end video chip), but to be fair, I'm using it on a new, faster disk. My biggest problems have come from the failure of the NVIDIA installer to generate a working xorg.conf, and that the pcspeaker sound source (enabled in the default debian kernel I had installed) took precedence over my onboard sound.

I'm learning to like it, which IS what I expected to have to do. One reason I might be having an easier time adapting is that I didn't expect KDE 4 to be an EXACT copy of KDE 3; I don't expect Dolphin's scrolling function to be EXACTLY like Konqueror's, etc., etc.

Seriously, dude -- just stop using it. I would, if I found myself in the place you are. Take a breather. Exercise your freedom of choice, and do it with some aplomb.

Peter Yellman

Troll?

nye — Tue, 09 Jun 2009 01:47:44 +0000

Okay, the audiocd ioslave existed in KDE3 with the same functionality, so you're just wrong. This is one of many examples of things which have simply been removed from KDE4, and which certain people like to claim is just one isolated example, as if to say that each point, one at a time, doesn't matter.

I'll play.

In the first 3 hours of using KDE4.2 I kept a log of the more annoying errors. Of course, there were far more problems than this in that time but these are the ones that I could describe reasonably well and reproduce sufficiently that I couldn't imagine anyone with KDE4.2 had not seen them. I got bored after that and haven't logged any more, and I didn't log the less annoying things (there are some which have bitten me since and keep hitting me constantly, but I've given up on reporting anything which isn't a clear, major bug unless it causes a reproducible crash - any regression will always be labelled as a wishlist item and ignored because xxx4 is a new application and apparently it's unreasonable to expect it to have equivalent features to xxx3.

But I digress; this is my log of the first few hours (It appears that I removed several of the entries as they were fixed, but at some point I got bored even of that, so some of these may not be problems any more - treat it as indicative and if you are deeply resentful about anything that might have been fixed in the mean time, remember all those other issues which have not). Intended for personal reference and unedited:

------

Does not honour ~/.Xresources when set to apply colours to non-KDE applications. Maybe a separate setting to apply this to the console? Or maybe the user Xresources should be reloaded after merging KDE settings.
Actually, krdb does attempt to source .Xresources, but fails because its wd is not ~, and my .Xresources file has relative paths, because it's not possible to include $HOME or ~ and have it expanded, and I don't want to hardcode the absolute path since this file is shared between machines where that path varies.

(Snow not working for me)

Graphical glitches:
locked desktop continues to show previous contents
horizontal/vertical lines and corners appear in odd places - possibly showing the shadow of a window in the wrong location? No, sometimes it appears to be showing slivers of whatever is underneath the window. But what about when it happens on the desktop?

Mousewheel in dolphinpart only scrolls a tiny amount and doesn't appear to be configurable.

Since first login: kdesu <application> no longer works, synaptic no longer asks for root privileges, and login fade to desktop no longer works (works when doing 'Test Theme' in Splash Screen configuration).

When logging out, desktop is slowly desaturated, then after clicking the logout button it becomes resaturated, sits there for a short period of time, and suddenly changes to black before the greeter reappears.

Would be nice to have more control of the appearance of the panel beyond just switching the plasma theme - eg. KDE3 had the option to set transparency.
While we're at it, a one-line panel taking up the same amount of vertical space as in KDE3 would also be good. Obviously it's resizable, but lots of icons seem to go fuzzy at that size - could this be that the padding is greater so the icons are being scaled slightly smaller than in KDE3? This seems better in Debian so possibly there's something buggy in Kubuntu's packages?

krunner sometimes doesn't seem to launch what you typed. Particularly seems to happen when typing in part of a command name and tabbing, but not consistently.

krunner sometimes appears to be showing two options superimposed

krunner sometimes wont launch an application in $PATH if there is an .desktop file describing an application of that name. This causes problems such as "Service '/var/lib/menu-xdg/applications/menu-xdg/X-Debian-Applications-Network-Communication-krdc.desktop' is malformatted." The application can be launched from an xterm as usual. When trying to duplicate this for testing, the option just to run the specified command appeared as it should - possibly this is related to the previous bug, and that option was hidden beneath another when I first tried.

Session saving doesn't seem to work - eg. running Konqueror is killed, and when manually started next session the crash recovery prompt comes up.

Dragon Player only rarely makes sound (even playing the same file may or may not make any sound).

Konqueror crashes frequently when customising toolbars (after clicking OK)

Some other would-be-nice desktop effects options: when setting menus to be translucent, apply the translucency only to the background, not the content - I bevieve this is likely to be what's desired most of the time. Would like some options for the blur plugin - don't remember what I was trying to do...one thing was make menus highly translucent, and very blurred, in contrast to the usual (very good!) behaviour of blurring proportionately to the opacity.

Would be nice to have dolphinpart separated out so that Konqueror can use it without needing to install Dolphin.

Kickoff menu doesn't fade in like other windows/menus. It just appears immediately with a bit of flicker.

Disable Kickoff recently used tab?

Changing between Kickoff and Classic menu types unregisters keyboard shortcut

Removing phonon-backend-xine and replacing with -null for testing:
After launching systemsettings without restarting KDE, Multimedia section causes
'Error: Unable to use the Xine Multimedia Backend:
The shared library was not found.'
Clicking OK just causes the box to reappear, as does the close button.
Eventually had to xkill systemsettings

================================================================================

** Known bugs:

Doesn't appear to be possible to change Login Manager settings as there's no button to become root.

Splash Screen icons display perfectly in preview, but flicker in actual use, and appear to be cut off such that the blurry-fade effect is cropped.

Constant visual glitches in notification area.
Same thing?
System tray: icon background is corrupted.

** Reported Wishlist:

Window decoration: option to close window by double-clicking menu icon. This exists for some windecos, eg. Plastik, but doesn't work

Middle-click on desktop to get task list for all desktops.

Would be very nice to have a keybinding to minimise all windows. Showing Plasma dashboard is not the same because I'd like to be able to selectively restore windows, for example.

Fuzzy clock: allow changing font size, and vertically algn to the centre, not the top.

Would be nice if Login Manager config included some reminder that not only do you need to log out, but kdm must be restarted in order to read the changes. Selecting 'Restart X Server' from the login menu is insufficient. Even nicer would be for it to re-read its configuration when the user logs out.

Also would like a GUI option to choose between su and sudo for kdesu, though this isn't something anyone should ever really have to use - is there any way to guess this?

Kate doesn't remember recent files

** Bugs resolved as 'WONTFIX':

Desktop grid/cube/etc. show task manager from starting desktop in all desktops.

** Configuration problem:

In KDE3, I liked getting a larger preview of a file as the tooltip.

Having panel cashew in bottom-right corner is a usability issue: it's an uncommonly used function and prevents other widgets from making use of the infinite space. In particular, desktop pager which I use for switching desktops with the mouse wheel.
->Cashew is hidden when widgets are locked.

** Reported bugs:

systemsettings: multimedia section crashes immediately (no crash dialogue) - fixed by installing phonon-backend-xine (previously using -null).

** Need more thought:

When desktop switching is set to use the cube, it's extremely unintuitive that dragging a window from desktop 1 to desktop 3 can't be done by going left or right twice, as it's inconsistent with the animation used.

Dragging windows between desktops is inconsistent with desktop cube switching animation

When switching desktops using the cube, the cube rotates around the vertical axis to the destination desktop, so for example moving from desktop 1 to 3 shows two rotations. Dragging windows from desktop 1 to 3 can't be done like this - if the window is dragged right twice, then the second flip goes in the opposite direction to what's expected and returns to desktop 1. This confused me for a while until I realised that it's using the desktop arrangements from the pager widget, and I can drag windows to the top or bottom of the screen to move them - though it still animates two horizontal switches.

At first I was going to suggest that desktop layout be considered as a line when the cube (cylinder, etc.) is used as the desktop switch animation, and the pager layout is taken when the grid is used as the switching animation.

This would give the same behaviour as in KDE3 - moving a window from desktop 1 to desktop 3 meant going right (or left) twice, even with the standard 2x2 layout in the pager widget, which would be consistent with the layout of the cube.

However, a comment in bug 177639 indicates that this behaviour is incorrect, and the pager should be treated as the definitive authority on how desktops are laid out[0], so perhaps the cube could be altered such that the animations shown are consistent with the layout as kwin understands it?

[0] I don't really like that, because I want a 1x4 layout as in the cube and as in KDE3 but I want the pager to be as small as possible, and set it to 2x2. I'd rather these be decoupled, as I see the pager as being a way to cram a little overview into the smallest space reasonably possible. Note that using the scroll wheel on the pager or the desktop acts more like the cube does - it increments the desktop number without regard for its nominal layout, which is the behaviour I find most intuitive.
But if that's what the spec says then fair enough.

Seigo: building brand together

brianomahoney — Mon, 08 Jun 2009 22:27:43 +0000

I absolutely agree, we need code bug-fixed and extended not an endless, useless focus on branding and UI appearance rather than fuctionality.

It isn't that Windows functionality is all there ie Exchange-Calendar, Active Directory ...

Do something useful, not masturbate the UI Graphic Design, leave that to Vista/Windows 7.

What other choice is there?

yokem_55 — Mon, 08 Jun 2009 22:22:18 +0000

The main metric that matters over the longer run though isn't number of users a project attracts. It's number of developers, and for KDE, that number has been going up since KDE4 began shipping. As ridiculous as Mr. Balmer's fated dance on that stage several years ago was, he was absolutely right. The broader the base of developers a project has means that it is proportionally more likely to better meet users needs.

Seigo: building brand together

coulamac — Mon, 08 Jun 2009 21:50:02 +0000

"GNOME
releases at similar speed, but it would be fair to say that their releases are just incremental
updates, and not major overhauls."

Your statement is not necessarily true. GNOME releases twice per year. Whether a new release is considered "incremental" or "major" depends upon what changes are ready for that particular release. For example, the addition of GIO/GVFS over Gnome-VFS in GNOME 2.22 was a fairly major subsystem overhaul. You might consider that GNOME release "major," and not "incremental" as a result. Other releases have been more about polishing the UI, reducing memory, squashing bugs, and speeding up the applications without major overhauls. You might consider those releases "incremental."

In the recent year or so, a lot of work has gone into the next generation libraries and UI like Clutter, gnome-introspection, gjs, seed, GSettings, gnome-shell, Zeitgeist, and the new accessibility framework. These things are only now starting to filter into GNOME proper. So, your implication that GNOME is stalled is not accurate.

K3b

jpetso — Mon, 08 Jun 2009 21:24:48 +0000

While being a relatively hidden feature, K3b is really *the* way to grab
audio CDs. Just as easy as it's flexible. KAudioCreator was a joke
compared to K3b (and audiocd:/, naturally). Granted, K3b's 4.x port still
hasn't been released, but unless your distro has completely dropped even
the KDE3 libraries, it should have a copy of K3b lying around somewhere in
the package repositories.

Troll?

pyellman — Mon, 08 Jun 2009 19:32:12 +0000

Umm, nah.

The "this way takes one fewer clicks, therefore must be upheld" arguments are instantly boring. I haven't counted the exact number of clicks, but I'm guessing the difference between the two ways comes to 3 or 4 clicks. Idea? Create a link to audiocd:/, reduce your clicks by one. The loss of functionality in regard to directory creation is definitely a bummer.

The subtext to my response, though, was not really about the comparison of functionality between the two methods, but about how you unfairly and inaccurately used this example to slam KDE 4 in general. If anything is "legion" it is comments like this in regard to KDE 4; look no further than other parts of this page. Kaudiocreator is not and was never a basic component of KDE; it was "add-on" software created and maintained by an individual and the simple fact is, it is no longer being maintained. I imagine that at some point a conversation was engaged in which the developer of kaudiocreator was asked and agreed to port as much of the functionality of the kaudiocreator as he could to KDE proper before withdrawing his time and effort. If anything, you should be lauding the KDE team for preserving some usable portion of the functionality of that software which will give you time to find another software option, if you should find that necessary.

In short, you are a victim of "a favorite piece of utility software of mine has been orphaned", not "KDE 4 screwed me". This kind of thing happens all the time -- it's happened to me innumerable times on Windows in the past. If this were a thread entitled "My Favorite Little Utility Software Has Been Orphaned", I would nod my head, and say "yeah, that really sucks, dude". Instead, I will suggest that you be more sensible in your criticisms, and, if it's that important to you, invest a little time figuring out if you can still get kaudiocreator working on KDE 4.

Peter Yellman

Troll?

nye — Mon, 08 Jun 2009 18:03:52 +0000

Using the audiocd:/ IOslave might seem comparable if you've never used the app, but in reality it's a joke. Rather than 'lauch app; click go', it's now 'open file browser; browse to audiocd:/; browse to subdirectory in appropriate format; select all; copy; browse to destination directory; create 'Artist' directory if nonexistant; create 'Album' directory; paste'.

(The directory creation steps are because the ioslave configuration doesn't support creating files in subdirectories - if you try to put a '/' in the output file name, it escapes it.)

KDE4 is like this all over.

Troll?

pyellman — Mon, 08 Jun 2009 17:46:49 +0000

It took me a few minutes to figure out how to rip cds in KDE 4 (4.2) also. I'd never used kaudiocreator, so I Googled for a couple of reviews, and found this this one http://benjamin-meyer.blogspot.com/2002/02/kaudiocreator.... (kinda old, yeah).

You will see from the screenshots in this posting that most of what kaudiocreator does can be achieved in KDE 4 by configuring "Audio CDs" under System Settings --> Advanced, and then using the kioslave audicd:/ in Dolphin to select your format (I've now got mp3/lame, wav, ogg, and flac -- just like in the screenshots) and doing a drag and drop. In fact, the similarities are such that I wouldn't be surprised to learn that over time, kaudiocreator's functions were simply absorbed in this manner.

It's true that the functions aren't as neatly arranged in a single app as they were previously, but I don't think this is really a show stopper.

PS -- I went ahead and searched the developer of kaudiocreator's blog (Benjamin Myer, above), and indeed, he's the one who did the work to incorporate kaudiocreator's functions into kde audiocd.

Definitely a case of 6 of one, half dozen of the other.

Peter Yellman

Troll?

nye — Mon, 08 Jun 2009 12:19:43 +0000

>I have just been burned by Debian's transition to KDE 4.x and what I perceive as an ongoing immaturity of the KDE 4 framework. Korganizer, one of my most important productivity apps, stopped working with a WebDAV-base

Yeah that transition wasn't too fun. My biggest annoyance (with that transition in particular; my annoyances with KDE4 in general are legion) is that kaudiocreator doesn't exist any more and KDE4 doesn't have anything comparable, so I've had to 'upgrade' to abcde.

Seigo: building brand together

Janne — Mon, 08 Jun 2009 12:19:13 +0000

When I talk of "usability", I'm not referring to a HIG that the desktop
follows, but to general feeling that how well the desktop works with some
particular user. OS X (for example) has an extensive HIG and usability-
guidelines, yet many people feel that it's not usable, since they can't
make it work in a way that they want it to work.

I would say that if you do not like the UI you are using, it will
negatively affect usability, no matter how extensive and well-thought that
UI's HIG is.

But even if we talk of strictly "usability" and "likability", then we are
back to square one. What makes KDE4 "unusable"? I know for a fact that
there are lots and lots of people using it at this very moment, and they
are liking it and finding it usable. So obviously the complaints mentioned
here that KDE4 is not "usable" are more about "likability" than they are
about "usability" as such, since "likability" is a subjective feeling,
whereas "usability" is more objective (I'm simplifying the issue, but I
hope you understand what I mean). People like different things, and that
includes stuff like music, movies.... and yes, desktop-environments and
GUIs.

We can't objectively claim that "KDE4 is unusable", since there are lots if
users using it right now. If KDE4 is objectively speaking "unusable", why
are those people using and enjoying it? Because they are stupid or insane?

No, what we have here is a subjective feeling that "I do not like KDE4".
And there's nothing wrong with that feeling. But to claim that KDE4 is
objectively speaking unusable, is wrong, since there are lots of people who
are finding it perfectly usable. It just happens that no GUI can cater to
everyone.

Seigo: building brand together

hppnq — Mon, 08 Jun 2009 11:56:26 +0000

Ah, but you missed that "branding!!!" should be part of any, uh, branding effort nowadays.

Seigo: building brand together

hppnq — Mon, 08 Jun 2009 11:38:45 +0000

Yes yes, you still find it unusable. And I do find it usable. What does that tell us? That we have different wants and needs. Some people like GNUstep....

Actually, all the desktop environments you mention employ a definition of "usability". It might not be a bad idea to take a look, and to chew over the difference between "likability" and "usability".

Moemntum != quality

nix — Mon, 08 Jun 2009 11:38:20 +0000

Hey, even MS has something to learn about late delivery when compared with the Hurd. (Of course that is hurt by the fact that everyone who develops for it seems to want to rewrite it from the ground up, so it's had how many total rewrites? Three?)

Not so long really

nix — Mon, 08 Jun 2009 10:45:53 +0000

If you want a usability nightmare, look at what the late unlamented XForms widget set did with its menus. Ew, unusable must-hold-down-mouse-buttondom.

Seigo: building brand together

wstephenson — Mon, 08 Jun 2009 09:54:13 +0000

In openSUSE we cater to this need by providing alternative branding packages so you don't need to recompile just to get the default look.

The distro install gives you kde4-kdm-branding-openSUSE, kdebase4-runtime-branding-openSUSE, and kdebase4-workspace-branding-openSUSE, but these can be replaced with the 'upstream' defaults eg kdebase4-workspace-branding-upstream.

By using our packages you keep all the integration work that makes KDE run better on the underlying openSUSE infrastructure.

Moemntum != quality

khim — Mon, 08 Jun 2009 08:46:12 +0000

GNOME releases at similar speed, but it would be fair to say that their releases are just incremental updates, and not major overhauls. I would say that KDE certainly has the momentum.

GNOME only releases incremental updates because it's usable desktop. KDE4.0 was so messy it was almost unusable. So of course KDE has "momentum". I presume when GNOME 3.0 will be released it'll be "GNOME time": the GNOME desktop will be unusable like KDE4.0 and GNOME will have this "momentum". Only then we'll have fair compatison.

MS has promises us great things for 15 years now (even longer, actually), and they have ALWAYS failed to deliver.

Yup. Object File System for NT was promised in 1991 and "almost ready" in 1994. It's still not delivered. Compred to that KDE developers are saints.

Seigo: building brand together

Janne — Mon, 08 Jun 2009 05:19:03 +0000

"Firstly, I must confess that the Language Nazi in me compels me to point out, in a constructive
and good-natured way, that the term "ingenuousness" is probably the requisite term. :-)"

Is that about something I said? I must point out that English is not my native language.

"Meanwhile, KDE4 had a long gestation, with many setbacks and delays. And when it was finally
(and prematurely) released, its state was surprisingly poor. And it has been about a year and a
half. And *maybe* 4.3 is the charm. (I'm not making any commitments on that.)"

It's been 1.5 years, and in that time we have had two major releases. And both releases were big
improvements. That's more or less the same time-period where Microsoft releases a service-
pack, which are nowhere near as complete makeovers as major KDE-releases are. GNOME
releases at similar speed, but it would be fair to say that their releases are just incremental
updates, and not major overhauls. I would say that KDE certainly has the momentum.

"So while Microsoft's foibles were the furthest thing from my mind when I originally posted... you
have demonstrated to me a very interesting parallel."

The thing that is striking between the two is the level of difference between the two. MS has
promises us great things for 15 years now (even longer, actually), and they have ALWAYS failed to
deliver. And every time when they release their latest and greatest, they immediately start to
explain to everyone "the NEXT version is going to be really great! Honest!".

KDE has delivered us great software for years. Yes, KDE4.0 was a miscalculation on their part,
but they have been VERY fast in fixing it. In one year they went from "unusable mess" to "hey, this
is pretty damn cool!". And they give us that for free.

Yes yes, you still find it unusable. And I do find it usable. What does that tell us? That we have
different wants and needs. Some people like GNUstep....

Troll?

Janne — Mon, 08 Jun 2009 05:10:31 +0000

"At any rate, I will stand behind my statement, supported by responses by others in this thread,
that many still do not consider KDE4, in its current 4.2.4 guise, to be usable."

You made two claims:

a) KDE 4.2 is not usable

b) There are no real differences between 4.2 and 4.0

Point A if perfectly valid. There are plenty of software out there that have lots of users and fans,
but which I find to be not usable. Does that mean that the software in question is crap? Not at
all. We just have different wants and needs. It would be dumb for me to start telling you that you
DO like 4.2, since you and you alone can make that judgement.

Then there's the claim B, which is utter crap. Everyone (yes, EVERYONE) can see that 4.2 is a lot
better than 4.0 was. No, that does not necessarily mean that it's suitable for you for example,
but to claim that they are more or less identical is ludicrous.

And then there's the claim that "others support" your claim that 4.2 is not usable. I would be
careful with that kind of support. People comment when something doesn't work, they don't
comment when something does work. If you use a piece of software, and you have problems, you
will probably complain in a forum and/or ask for help. If you use some software and you don't
have any issies, what are the odds that you would go to a forum and say "I just wanted to let you
know that I have no problems with this software"?

What other choice is there?

sbergman27 — Mon, 08 Jun 2009 00:37:12 +0000

So... are we talking about KDE *gaining* users? Or KDE *not losing* existing users?

There is a difference.

Of course, then there is also the matter of how any influx of new FOSS desktop users might be divided.

From an "overall users" standpoint, I'm not sure that the KDE project is positioning themselves so well, despite how much their existing clique of users might or might not be enamored by KDE4.

What other choice is there?

khim — Mon, 08 Jun 2009 00:06:05 +0000

OSS users have never had that much real choice in browser offerings.

OSS users - probably not. But then these "OSS users" are somehow happy to use proprietary Flash plugin and proprietary nVidia drivers so I doubt Opera was unused just because it was not OSS. I think it's just inertia: Firefox was quite different from Mozilla Suite, but still much closer then Koqueror or Opera. There were a lot of complains but few switchers. I suppose the same will happen with KDE4.

So I'm not at all sure that the strategy is going to work in favor of KDE4 in this case. In fact, I suspect that it will not work well for them at all.

Why not? It's like Linux => BSD switch. Sure, KDE4.2 does not offer all options presnt in KDE3.5, but GNOME and XFCE offer even less! And if KDE3.x will bitrot (and it will, no doubt about this) KDE users will be faced with tow choices: go with KDE4 (somwhat lacking, but familiar), or switch to totally alient desktop. I doubt many will switch.

Sorry, but this is bull-shit

sbergman27 — Sun, 07 Jun 2009 23:27:40 +0000

What you say is reasonable.

Particularly "Mozilla used it". OSS users have never had that much real choice in browser offerings. "Kernel used it" is a more interesting case. OSS users have a choice of kernel; The *BSDs are out there. And yet... moving away from Linux is a more major step, for a current Linux user, than would have been changing browsers. That brings us to "Gnome used it". Which is the most interesting question.

Certainly, in the transition from Gnome 1.x to 2.x, so long ago, certain users were alienated. Some of them just wanted some twiddly configuration options back, and didn't hear about gconf-editor, which is unfortunate. They may well have ended up with KDE. Others really did need to be able to change the anti-aliasing style of the text embedded in the pixmap of the middle button of the window decoration button group which they had moved from the upper right to the lower right of their windows, for critical productivity reasons, and are probably now spending most of their time editing and re-editing Flux-box configuration files in Emacs and engaging in passionate battles with people who do the same using vi instead.

But for most users who are not so weighed down by productivity-critical desktop interaction minutia... there is a choice between desktop environments.

So I'm not at all sure that the strategy is going to work in favor of KDE4 in this case. In fact, I suspect that it will not work well for them at all.

Seigo: building brand together

sbergman27 — Sun, 07 Jun 2009 21:35:32 +0000

Firstly, I must confess that the Language Nazi in me compels me to point out, in a constructive and good-natured way, that the term "ingenuousness" is probably the requisite term. :-)

That said, the comparison with Microsoft's releases is interesting. Not that I was initially comparing KDE4 with anything from the State of Washington. And ignoring the fact that we are (mostly) comparing full OSes to desktop environments... in broad terms, XP was a big improvement upon Win98, which was a significant improvement upon Win95, which was a huge improvement upon Win 3.x, which was a big improvement upon DOS. You imply that there have been many releases of Windows which have not been notable steps forward. The two notable big embarrassing gaffes were, of course, ME and Vista. What can I say about ME? It was ME. More recently, Vista had a long gestation with many setbacks and delays. And when it was finally (and prematurely) released, its state was surprisingly poor. And it took about a year and a half to arrive at a reasonable level of quality.

Meanwhile, KDE4 had a long gestation, with many setbacks and delays. And when it was finally (and prematurely) released, its state was surprisingly poor. And it has been about a year and a half. And *maybe* 4.3 is the charm. (I'm not making any commitments on that.)

So while Microsoft's foibles were the furthest thing from my mind when I originally posted... you have demonstrated to me a very interesting parallel.

Thanks. :-)

-Steve

Seigo: building brand together

sbergman27 — Sun, 07 Jun 2009 21:00:34 +0000

"""
Without wanting to heat up a flame war
"""

I would like to comment upon that, because I think it is an important point. I do not consider the strong disagreements in this thread to be a "flame war", although I certainly understand the sentiment and your choice of terms. LWN.net has few flamewars. And when they do occur, the flame-retardant nature of its general readership tends to keep the resulting insurance claims down. ;-)

This is not even funny - KDE developers accuse distributors because they are doing what the KDE developers iteself did in the past!

khim — Sun, 07 Jun 2009 17:05:12 +0000

Unfortunately, in the F/OSS world we like to build little fences around our plots of land and then design the gardens in them like the unique little acres of wonder we feel they are.

And if you'll recall the history KDE4 developers started this trend. They converted all icons names to the freedesktop.org guidelines - true, but they refused to follow Tango style guidelines! May be they wanted to give KDE4 unique look-and-feel, or may be Tango developers were uncooperative... it does not matter: they did to Linux desktop what distributions are doing to KDE.

If they want some unity - they should discuss this with GNOME developers first: there a lot of distribution out there but there are just two major "Linux desktops" out there - and if developers of these two can not agree on look and feel then how the hell they hope to convince distributors?

I'm not saying KDE should drop Oxygen, I'm not saying GNOME developers should drop Tango - it's up for developers for both camps to decide. But as long as themes of GNOME and KDE are different I see no point in trying to work with distributors.

Not so long really

dkite — Sun, 07 Jun 2009 16:16:11 +0000

I'm curious. What exactly about Qt creates a 'usability nightmare' for
users?

Derek

Sorry, but this is bull-shit

khim — Sun, 07 Jun 2009 14:44:41 +0000

The strategy you reference only works when one holds a monopoly. In our OSS world, where chaos^Wchoice reigns, it's a no go.

Sorry, but it's tried and true strategy. GNOME used it, Mozilla used it (with switch to Firefox), kernel used it many times. It's perfectly valid and clear strategy: you need to satisfy 80-90% of developers base (not user base!) and after split lovers of "old things" will eventually have unusable piece of code (GNOME 1.x) or some second-class citizen (Mozilla Seamonkey) and eventually most users will stop using "old junk".

Sure, it's much easier to do with proprietary offerings - there you can push new version down the people throats (see Windows Vista), but even OSS world can not save the platform if it lost enough developers. KDE3 lost most of them - I certainly know very few guys who are interesting in supporting it, let alone enhancing it.

Most people who loudly complain that KDE4 is a mess a KDE3.5 is a way to go are end-user - and that's not enough to keep old system alive. KDE 4.2 is "good enough" to sway enough user for this strategy to work and if KDE 4.3 will be better - it'll just make the switch easier...

Troll?

sbergman27 — Sun, 07 Jun 2009 13:47:41 +0000

"""
Ok, by claiming that you see no improvements from 4.0 to 4.2.4 you clearly show that you are just trolling.
"""

Kragil, I'm disappointed in you. Playing the 'T' card like that. I expected that you would be above employing that rather questionable strategy. Here is my statement which you were replying to:

"And no, I did not see differences between 4.0 and 4.2 which were significant in comparison to the difference between 4.0 and where it would need to be for many to consider it usable."

I phrased that rather carefully. And if you are going to reply, please reply to that, and not to something different which you imply that I said.

At any rate, I will stand behind my statement, supported by responses by others in this thread, that many still do not consider KDE4, in its current 4.2.4 guise, to be usable.

And I should also point out that your fish is tainted. (I'm genuinely sorry to have to say that about what you describe as your only fish.) Slack ships with KDE 3.5.10 and Debian with 3.5.9. Of course, your claim is that they are "going to" ship KDE4. (And I'll assume that you do mean to imply "in their next release".) Well, with Debian, that could be a very long ways down the line, indeed. (Judging by "recent" history... likely about 18 months from now.) Which leaves Slack. And somehow I find the implication of Slackware users representing "everyone" to be bizarre enough to have made reading your post worthwhile.

Seigo: building brand together

stumbles — Sun, 07 Jun 2009 12:28:47 +0000

Which is why I for go a distros attempts at "branding their" KDE and just compile it from source to get all the wholesome goodness the KDE devs intended... without the monkeying round by a distro

Seigo: building brand together

cantsin — Sun, 07 Jun 2009 11:09:22 +0000

Without wanting to heat up a flame war - as a user of a couple of KDE-based applications (rather than the whole desktop environment), I have just been burned by Debian's transition to KDE 4.x and what I perceive as an ongoing immaturity of the KDE 4 framework. Korganizer, one of my most important productivity apps, stopped working with a WebDAV-based iCal calendar used for scheduling a study programme I supervise. Apparently, the KIO file access abstraction layer of KDE 4 is to blame. My solution was to ultimately abandon korganizer and switch to Mozilla Sunbird/Iceowl. I am hearing similar complaints about lost or broken functionality due to the KDE4 framework from users of, among others, Amarok.

This is highly annoying because KDE, while arguably a bit clunky in its user interface design, always shined in the quality of its framework and subsystems, from network-transparent file access to printing, through textbook-like object orientation and componentization of the KDE middleware. I tend to attribute a lot of this software design maturity to the craftsmanship and experience of the developers who initiated KDE, such as Matthias Ettrich and Mathias Kalle Dalheimer (who engineered the first Linux port of StarOffice before co-initiating KDE). It seems to me as if KDE, nowadays, is led by a different generation of developers who give more priority to look-and-feel - and what I perceive as an anxiety to visually compete with Windows Vista and Mac OS X - than mature back-ends. There is a real danger that, with these priorities, KDE is repeating the mistakes of Enlightenment. From my empirical observation of Linux desktop usage, it seems as if KDE has shrunk to being the choice of a mere minority of desktop users; the troubles with KDE 4.x and the rise of Ubuntu have unbalanced competition on the Linux desktop.

Troll?

cantsin — Sun, 07 Jun 2009 10:39:17 +0000