Saturday, November 24, 2007

Why KDE4 (might) suck!

kde4 rc-1
I've been a kde user for very long time, since from before version 3.0. I vaguely remember being excited with version 3.0 to the point I was using the beta releases on a daily basis.

Similarly I have been anticipating kde 4.0, the next major release since 3.0 I suppose. But it seems that is just its problem! Its too much of a major release to be pushing out in such a hurry. I have been following kde 4 from time to time by reading some of the discussions on the panel-devel list, trying out the alpha releases via the SUSE live cd and of course discussing with Siraj, our local kde developer, on the internals (technology and community wise).

While things have improved a bit, with each release of kde 4, its far from ready. We are now at RC1 and the announcement says its ready for prime time testing.

Building on this, the majority of applications included in KDE 4.0 are now usable for day to day use. The KDE Release Team has recently underlined this by calling on the community to participate in reporting bugs during the time remaining before the release of KDE 4.0 in December.

Meanwhile, preparations for the KDE 4.0 release event is taking place, with the main event taking place in Mountain View, California in the USA in January 2008. Make sure you don't miss it!


Well from what I saw on the Debian based LiveCD, its far from being usable on a daily basis. I sincerely hope this is because the Debian guys didn't compile or package it correctly, even though thats very highly unlikely.

And while I don't want to ruin the January release party at Google, which by the way, goes up to Jan 19, which happens to be my B' day (Whooo!), my feeling is kde 4 should have been pushed back to June/July to make give that polish.

Anyway rather than just rant on and on, let me put forward my main issues with it, coming from a kde 3.x background. My testing procedure involved booting up kde 4-rc1 debian live cd and taking notes on the iTouch. To get a bit more background to the issues I was facing, I mailed those comments off to Siraj and I have included his comments and my replies to his comments :) Even he seems a bit frustrated with some aspects of the current state of kde. I'm sorry I don't have screenshots. All this was tested in 30-40 minutes, which only goes to show that its too easy to be annoyed with kde 4.

Legend: My initial comment, Siraj's comment, My response to his comment :)

Bugs like



* I see a floating klipper with black background

> Not a bug, plasma is yet to find a way to using XEmbed windows
inside a QGV , so till that is ready clipper will float.


>>Well its a bug! I get this from time to time even on kde 3.5.x after compiling with library versions used with beryl/compiz. But that happens randomly on 3.5.x and is predictable on 4.0. This should be fixed before final release

* I can't adjust time using clock applet. How do I change time format from 12/24

> you can't ;) it's a known issue :P

* I can't remove some applets like Lancelot. Why not have a remove applet functionality inside add applets dialog

> I don't think the IPC (dbus call) to remove an applet is not ready yet

* I can't resize plasma applets!

> hehe, the clock you can! others u can't ;)

>>Thanks for adding that feature. But its still crude (you need to type in a value under configure menu). Resizing should part of (inherited in )all applet


* Its not clear how I add stuff to kde menu's quick access tab.Would've been easy if I could drag and drop from app tab


* Plugging in my iPod, USB disk didint fireup the what should I do pop up. Neither did it show anything on my desktop.

> I don't thing solid supports all that, but do u have the solid applet loaded ?

>>No didn't have this loaded. I think such applets, if there are any needs to be loaded by default.

* I can drag a file from file browser to desktop but it only shows icon not file name and I can't move its position

> I think the applet lock is on, unclicking unlock should let you move it
around.


>>I'd really like to believe that, except there is nothing that shows its the case. There was no indication in the icon (via an emblem of a padlock for example) and I don't recall an unlock option in the context menu or even a context menu :) Its also bad you don't see the file name. Imagine having 10 files on the desktop all having just the same icon!

* How do I get multiple desktops. Used to be more intuitive

> there is an applet for virtual desktops.

>>If there is, this should be on by default.

* Launching the pager applet crashed the desktop and upon reload all the applets I had added were gone.

> Kwin and plasma is having lot of problems, this is just one of those. Yeap, no history support implemented yet ;)

>>A critical must fix bug, since kde4 seems to depend so much on these applets to provide core functionality previously supported natively (via kicker, konqueror)


Usability issues (and sometimes bug like)



* On the top right seetings widget has zoom in/out. Zoom out shrinks the desktop to a point of no return. Why not have a zoom default. Also this should be implemeted better esthetically.

> this is a pre plan for ZUI, which mean the default should be keep to zoom out. ZUI is not ready so it's a feature not a bug ;)

>>:) its a bug, because its a missing feature ;)

* Lots of missing icons so defaults to ugly icon.

> Oxygen is doing some major rework, when done should be ok!


* Default clock in systray looks ugly

> Agreed, argued no use, users are stuck with that! Aaron likes it

* In add applet dialog its not clear what the star does. If to rate then why 1 star.

> hehehe, maybe for the looks!

* Kde menu based on kicker looks unpolished and ugly

> :( life!

* Would be nice if I can clear recently used histry on that tab itself

> you will just have to wait :)

>>I can wait. Just wish kde can too instead of being released in Jan without all these


* Left and right click has the same effect on k menu button. How do I
add a custom item to menu?


> you can't for now :)

* I can't right click on taskbar. What happened those functionality.

> not implemented yet !

* Right click on desktop and I can't create new file, folder, launcher icon. WTF. Bettr to offer this along with option to hide icons from desktop. Dont cripple it

> not ready :)

* Where is trash icon? Should be in desktop or dock. Its not even an applet. It should be there by default.

> trash widget should be there but not ready yet :)

These are just some of the issues I found after playing with it for 45 min or so using the debian live cd. Definitely not RC1 quality. More like beta 1 or alpha 3!

kde4 rc-1 kde4 rc-1

So in conclusion, kde 4, IMHO doesn't offer much that isn't already available via other add-ons but as a result as compromised on usability and most of all flexibility which differentiates it from say GNOME. Worse yet, it seems to be missing a lot of the things we currently have with kde 3.5.x and earlier.

As a result, I think kde 4 has diverted from what its current users love and expect of kde. Its as if it has tried to be something its not. So my main complaint is that it needs a lot of work to get the same polish which we currently enjoy with kde 3.5.x series and should take the time needed before releasing prematurely. There is a saying in FOSS - It will be released when its ready!

In the mine time lets continue making kde 3.y.x incrementally better :)

35 comments:

Dean said...

Sounds like a bad case of developerits -- ie -- Gosh, look at all this cool stuff we can do. And my, doesn't the code look so much nicer.

I'm sure the users will like it.
Right?

In fairness, it does sound like the KDE folks have tried to take a big step for the future, making a more versatile, maintainable, and reusable platform.

They have a great track record of getting stuff out on time in decent shape, but it has always been my experience that late betas and early rcs were very usable.

If that's not the case now, it does seem like cause for concern.

Not panic, maybe, but concern.

Henaway said...

They should really pull the plug on the release party ... because while there ARE a few things to celebrate with kde4, this baby ain't ready for primetime.

When THAT MUCH core functionality is broken, it's not RC quality. Maybe for Microsoft, but not for FLOSS. I tried it out on livecd and couldn't do much with it at all. Almost useless.

And you're not alone on the clock issue. Ugly as hell, and "Aaron likes it" is NOT a good reason for a COMMUNITY project to keep something so many people seem to dislike. If it was a one man project, fine, it's your baby. But there are more than one developer on one of THE most important projects in the linux world. Take the ego blow and do the right thing with it. It's just a clock. Make the dang thing configurable like the 3.x series.

No, I don't see myself flipping the switch to KDE4 for quite some time yet. Maybe it will be usable everyday by the 4.3 or 4.9 release. But I sure won't trust my desktop to it before a couple of point releases that add CORE functionality like I've had since kde 2.x.

Tío Rubo said...

Most of your complaints are about Plasma, which is the least ready part of KDE4. Most of the underlying infrastructure is ready, as well as the programs.

So, it's already been said that KDE 4.0 won't have feature parity to KDE 3.5.8, but that doesn't mean it will suck.

Ben McDonald said...

Almost all of those problems were with plasma and once plasma is working KDE4 should be great.

Plasma is being fixed up at an amazing pace and it is now possible to resize all applets. Check out http://www.youtube.com/watch?v=HGYGzTDHhPg&eurl=http://commit-digest.org/issues/latest/

As for now liking the clock, I think there are already 3 different clock applets and I'm sure there will be many more. Plus the clock can be re-themed by changing the svg, so I'm sure there will be another one you like.

blackbelt_jones said...

I don't like the look of this "Dolphin" thing. After my last upgrade, it suddenly replaced Konqueror as the default file manager in Kubuntu. The first time I tried it, I tried to open a terminal window. Instead of opening from the present working directory, it opened konsole from my home directory. That's all it took to get me to reach for the harpoon.

blackbelt_jones said...
This comment has been removed by the author.
John said...

"There is a saying in FOSS - It will be released when its ready!"

Actually, no, the saying is "Release early, release often".

The platform is ready, the apps are ready, Plasma will fall over the line sometime mid-December, and KDE 4.0 will then be ready for a January release.

All your complaints seem to be about the fit and finish of the Plasma desktop, and you infer from that that the entire underlying framework is not yet ready. This is not the case, most of the remaining Plasma work is just a case of connecting the framework up to the desktop, fit and finishing work.

Sure, we won't have all the features from KDE3, and Plasma will not be a complete desktop, but we can't wait for 6 more months, we need to ship now to get the platform being used and the problems shaken out, and to allow all the downstreams time to work with it before they ship 4.1 in their releases in Q3 2008.

This is the way KDE works, always has: build a solid foundation, get it out the door, then spend several iterations applying the polish on the top.

You like 3.5 the way it is? Thats great, stick with it until 4.1 comes out which will be the complete, finished 'shrink-wrap' equivalent you're apparently looking for.

Bud said...

Few people had said that most of my complaints were around Plasma and everything else (e.g. apps) was ready.

While I'm not sure that everything else is ready as advertised, I think Plasma should be usable since it contributes to a significant amount of user interaction.

What people have to realize is kde is a desktop environment which the user mostly interacts with. So while it might be ok for an underlying framework to be broken, when the UI is broken or very limiting then thats what most people will think of. Not how wonderful things are under the hood.

The other thing that factors in is ofcourse kde's popularity and maturity. Previous versions of Kde has already set the bar (not to mention other desktop environments).

Some have suggested I should wait till 4.1 but that contradicts with the kde release note as well as target user. Is kde 4.0 a developer only release and not meant to be used by general users and distributions? If so, why not name the release as such?

There was a comment about "Release early, release often" as opposed to "Release when ready", which I some what reject.Yes FOSS is generally released incrementally and often but that is for point releases. Usually major releases are not done until the community feels they are ready for wider use.

e.g: 1) KDE :) KDE4 has been in development for a long long time. Why wasn't KDE4 released earlier without Plasma so it just feels like kde 3.5.x but was actually kde4 under the hood? Because they needed to finish the job. Did they finish it? Not IMHO.

2) The linux kernels 2.6 was in development (as 2.5) for about 2 years before they started going beta and RC.

When things change drastically FOSS projects tend to take that extra time and get things usable. Thats why most projects take a long time to reach 1.0

People say apps are ready, but last time I checked, I didn't see kopete or amarok or raptor. It also doesn't seem to have the compiz like effects that some had integrated to kwin. But I'm willing to over look all that if just the desktop was usable.

To me it feels like KDE4 has run a marathon non-stop and is taking a break while it only has 200m left to go. I'm saying, run that 200m and have the release party.

If you think this is the only review thats gonna say KDE4 isn't ready, well thats really wishful thinking. When kde4 is finally release a lot more people are going to have a look at it. And if things don't drastically improve between now and Jan then there are sure going to be a lot of bad press! And for a kde user like myself, thats not going to go down that well.

Henaway said...

I find myself in agreement with Bud here. When the AVERAGE USER fires up KDE4, they won't give a flying fig if it's a problem with plasma or oxygen or solid or any of the sub-projects. All they will see is that KDE4 is broken. And that will make a lasting first impression.

I write ads for a living. I know the full extent of what it takes to overcome bad first impressions and less than stellar press reports. It takes a lot more than another point release. Look at the damage the bad press around the betas of Vista were for Microsoft. Now the product may actually suck as bad as reported, and in my experience it does, but how many people are refusing to try it based SOLELY on what they've heard others say? Lots. Lots and lots and lots. It's keeping the Redmond beast awake at night it's so bad for them. And they have billions in ad budgets to fight public perception with. Does KDE have that luxury? I don't think so.

I'm excited about KDE4, I really am. I think that ONCE IT'S READY it will blow us all away. But I don't think it's ready for a 4.0 final, and won't be for a few more months. And if the devs want to prove me wrong and shut my mouth, I welcome them to it. But I'm not going to upgrade my desktop environment to downgrade the functionality I need.

I just hope it doesn't release too soon, get a black eye, and have to fight until 5.0 to get it's standing as THE best desktop back.

Aaron J. Seigo said...

let's see ...

the systray, resizing applets, moving applets, device notifier, clock in "systray" (it's the panel, not the systray, but whatever) aesthetics, menu aesthetics and right click on taskbar have all been addressed along with many more things.

the final layout of the panel hasn't been decided yet, though there will certainly be the pager on it.

some of Siraj's comments are also off the mark. he's hardly involved with KDE development these days (though still an awesome guy =)

it's sad your rant of a blog entry made it onto sites such linuxtoday, as it's so woefully out of date with the state of 4.0 that it's interesting as a historical note and little else. but it's sensationalist and controversial so of course it will. congrats on that.

it's also pretty sad that you focus on one app and some of the icons that are missing and gloss over absolutely everything else.

there are a few more issues left to go in 4.0, some of which you also note and some you don't (but hey, you only gave it 40 minutes ;). i think your blog title "geek with an attitude" pretty much sums up this entry. =/

i'm sorry you didn't like what you saw there, but as others have noted it's "release early, release often" and until we start threatening with "rc" status too few people out there bother testing things or start working on porting their apps.

but hey, it's easier to stand on the side lines and throw rocks, right? i see one bug report from you on bugs.kde.org (against kde3 amarok), don't see any posts to panel-devel, etc .. *shrug*

chris said...

>> as it's so woefully out of date with the state of 4.0

True or not, there's an obvious lesson if you care to learn it.

A Release Candidate should be just that - a candidate for final release. This was apparently nowhere near being that as there were obvious pieces of functionality that hadn't even been written.

If you abuse the RC term you're going to get much harsher feedback than if you label it as an Alpha or Beta. I suggest Public Beta as a more appropriate term.

Bud said...

I will reply in reverse:

>it's easier to stand on the side lines and throw rocks, >right?

Aaron, sorry that your taking this a bit personal but the intention was not to throw rocks but convey that I don't think kde rc-1 is of rc quality. I've tested the previous betas and alphas but didn't speak out because they were alphas and betas and I could see good progress was being made.

>i see one bug report from you on bugs.kde.org (against >kde3 amarok), don't see any posts to panel-devel, etc ..

Guilty as changed on not putting out more bug reports but I don't think I need go on panel-devel and be an ass* saying kde4-rc1 is broken.

I am seeing the tremendous momentum on that list and see a lot of things are happening.

I'm not a kde developer but am an interested user. I've even came to one of the irc meetings (I think the first) where you were missing for a couple of hours.

>but as others have noted it's "release early, release >often" and until we start threatening with "rc" status >too few people out there bother testing things or start >working on porting their apps.

I don't agree this is a good approach but if it works for you then I guess its ok.

>it's also pretty sad that you focus on one app and some >of the icons that are missing and gloss over absolutely >everything else.

By one app, I think you mean plasma? Yes so everyone seems to keep reminding me. Whether its one app or many doesn't matter because plasma seemed to have crippled the user experience. Whats a desktop environment without the user experience. I'm sure there is a reason for not called fluxbox or icewm as desktop environments. I'm sorry it makes you sad. It made me sad too after seeing what I saw.

>it's sad your rant of a blog entry made it onto sites >such linuxtoday, as it's so woefully out of date with the >state of 4.0 that it's interesting as a historical note >and little else.

I only submitted it to lxer. Yes its sad it got picked up by more news sites. But I'm happy to hear that the points I raised are now history. I look forward to rc-2

>let's see ...
>the systray, resizing applets,

Ok this is good news. I'm sorry for not running the svn version of kde, but that doesn't justify the state of rc-1. I look forward to trying out rc-2

Aaron J. Seigo said...

"but that doesn't justify the state of rc-1. I look forward to trying out rc-2"

ok, let's talk about this whole "what it was called" thing. we had very good reasons for calling it rc1. here are most of them:

- we need to make a release; we've been working on this code base for a long time now and it needs to get "out there"; that's the "early, often" thing

- it's really hard to get developers in release critical mode after 2 years of free form development, sometimes it takes a shot in the mouth

- it's really hard to get people testing releases before they are released; this is hardly an issue for our usual minor releases that we traditionally do every 9 months or so: there's relatively little to test, many of the goalposts are fixed rather well. not so with a major release like this one.

so instead of doing Yet Another Beta which would not have caused the movement needed at this time, we took a deep breath and called it an rc. we all knew that taking this difficult step was going to be painful. we knew people such as yourself would write things such as you did.

still, we accomplished what we had to. that may be difficult for many (most?) people to grasp (it requires a certain knowledge of the project's status as well as some big picture thought)

when you make as brazen a set of statements as in your blog entry here, i will certainly be around to offer some reality correction.

that's the beauty of open discourse.

Aaron J. Seigo said...

"Ugly as hell, and "Aaron likes it" is NOT a good reason for a COMMUNITY project to keep something so many people seem to dislike. I"

oh, btw, i *don't* like that clock. in fact ... i made the people who were doing it change it.

and probably more than any other developer out there in the free software desktop world right now, i spend a huge amount of time listening to, processing and responding to the community.

you have no idea how much time i spend not dealing with my own issues and desires, but working on those of others.

so, you really missed the mark on that one. on the upside, you were successfull at showing what a complete asshat you can be. *applause*

blackbelt_jones said...

As far as I'm concerened, do what you like, just don't mess with Konqueror. Dolphin is nice, I can see why you'd want something that, but I hope you can accept a little messy redundancy, and not try to push everybody into a neat, preprogrammed desktop "experience" It's reading about how "the two applications won't compete; they'll complement each other", that gave me physical symptoms. For God's sake, let them compete.

http://dot.kde.org/1172721427/

This is the only "official" thing that has been published on the subject.

quote:

Konqueror is not going to disappear for KDE 4, although its user interface may yet see some adjustments as its primary utility will not as the default file manager.



Can you imagine how ominous this seems to a fanatical Konqueror user? Sure, I'm overreacting, but I'm not getting a clear idea what's appropriate. Am I supposed to read this and NOT freak out?

Henry S. said...

I know it is subtle, but the phrase "the majority of applications included in KDE 4.0 are now usable" means something quite different than "Plasma is now usable."

Try testing the *applications*, and ignoring Plasma and the control center.

Also, if it helps people understand things: The big, nasty, in your face problems are usually the easiest to fix. The hard, time consuming part of development is in the subtleties and details.

Aaron J. Seigo said...

@blackbelt_jones: 'hope you can accept a little messy redundancy, and not try to push everybody into a neat, preprogrammed desktop "experience"'

rejoice, blackbelt jones! for we are not doing that at all.

both konqueror and dolphin are in kdebase in all their glory.

they have different use cases, different user bases ... consensus is we need and want them both.

that has always been our stand, and i don't see that changing. konq is too good =)

in fact, part of the reason for introducing dolphin is so that people who need a dolphin like thing don't try and screw up konqi ;)

as for adjusting it's UI to be more web browsery, i expect we'll eventually see a dolphinesque web-browser-only around webkit sometime in 4.1 or 4.2, relieving any of that UI pressure from konqi.

so konq will continue on as it always has with these other special purpose apps servicing a different group of people.

blackbelt_jones said...

Aaron, thank you for helping me to relax. As long as Konqueror is permitted to remain its big bad self, I can completely get behind Dolphin, a great-looking, user-friendly file manager that new users will love, and that's important.

And of course, Dolphin should be the default, since Konqueror power users will be be able to make the adjustment. (Even I expect to figure it out eventually!)

Kongratulations on all your fine work!

blackbelt_jones

Bud said...

kde4 release has been delayed by a month to fix some of the unpolished issues (aka usability experience) and is very much welcome, as you will see by the majority of user comments.

http://dot.kde.org/1196525703/

kde4 rc-1 seems very much outdated (5,000 bugs or so behind according to 1 comment) and yet at the time, it is still the latest release. I hope they do an rc-2 soon, so that fixed issues wont keep getting re-reported.

Robert said...

With some experience as a Software Engineer in big projects nearing Release deadlines, I have tried to test KDE4, as a user.

What I find, is many of the KDE4 Applictions included by Open SuSE 10.3, work fine under KDE3.

OTHO, the KDE4-rc1 release is IMO Alpha Quality. Basic things like Konqueror are very broken. Realistically a Beta should have the Plasma stuff ready in featyre complete (for KDE 4.0) form.

But bugs are getting fixed, and there's been a lot of discussion. My main concern is in the change of character of KDE4. KDE has suited Win Refugees, but also Power Users, who appreciate it's take on the UI, and things like kio_slaves in Konqi.

What I fear is "Gnomisation". GNOME drives me barmy, if Dolphin breaks the Konqi Browser/FileManager concept, which I love, then I'll probably try XFCE and go lightweight. Always been a text in terminal type, KDE I've used since Version 1.0.x, and has been my fave heavyweight DT by far.

There's probably a conflict, lots of KDE4 ppl are "done" and eager to move to the improved toolkit and ship.

Those who aren't done, and have integration issues, most likely tend to underestimate the time to ship.

Still given time, it'll probably come together very well, and then we'll see about UI changes.

Aaron J. Seigo said...

> What I fear is "Gnomisation".

not happening.

> if Dolphin breaks the Konqi
> Browser/FileManager concept

not happening either.

blackbelt_jones said...

Well, I think it's official. It's officially been released,and last night I installed it on opensuse. Either I can't understand the interface, or KDE4 sucks. Or both.

What I can't get past is the vastly curtailed keyboard shortcuts. Or perhaps I am mistaken? I click and I click, and I just can't find any additional keyboard options. I just don't see how it's possible that someone made a conscious decision that I wasn't going to want to directly open a terminal window with a keybinding.

This is pretty fundamental stuff. If it's going to make me use the mouse more, or type more, I just don't care about anything else. I still don't know what plasma is... and I'm not the least bit curious.

blackbelt_jones said...

Okay, you know what? To hell with this. I can't prove anything, but I have a strong suspicion that no matter what anyone tells you, as a file manager, Konqueror has been marked for death. I've been searching up and down the toolbar-- and the button to "show hidden files" seems to be missing. Okay, maybe it's SUSE that is causing the trouble, but a file manager that can go almost anywhere might as well not exist. This has all the ear marks of sabatoge. This sounds like paranoia when talking about KDE, but if I accused Microsoft of doing something like this, it would seem perfectly reasonable.

In order to view hidden files, I have to open Dolphin. And since Dolphin looks so much like Konqueror, I usually have no idea which I'm using at any given moment. I know a train wreck when I see one, and this is a goddamn train wreck. I'll try the next KDE 4when it comes out, but I'm fully prepared to hunker down with my stable debian and pray for a good long time before the next release.

When I'm this pissed off, it's usually a prelude to me finding out that I'm completely wrong. If that's what happens this time, I won't be afraid to admit it.

blackbelt_jones said...

Of course, sabatoge is a moot point. Based entirely on what I was running earlier today in SUSE 10.3, Konqueror is being killed. Whether it's murder or negligent homicide is just academic.

Robert said...

Konqi's not being killed, the web browser part is working better than ever. It was very buggy in rc1 & rc2 but seems better than ever now in 4.0.1 off the OpenSuSE build service (read the OS weekly news and update the repositary to get latest KDE 3.5.8 & KDE4 stable 4.0.1 or latest which ever is your preference.

According to reply to a bug report, the file manager kpart is in Konqi and the same "views" are usable. At time the KDE4 I had was crashing lots, so it wasn't checkable by me.

To me the release was just a couple of months premature, 4.0.1 seems perfectly acceptable as a X.0.0; with not all 3.5.8 features available.

I share your confusion when running KDE4, for some reason as a past KDE 1,2 & 3 user, suddenly I'm finding things non-intuitive and I get lost and frustrated. Due to low quality of the rc's it was not possible to see if that was due to bugs, or feature changes.

With all the developer emphasis on components with kool names like phonon, oxygen, plasma; I'm kind of wondering if they've not seen the wood for the trees. Some things are improved, but KDE was always a complete desktop environment, lots of features but very usable, and little need to read up documentation. Though a little study did pay off, as intro to neat features.

I really hope, we're not going to be stuck with the default Dolphin, there's something very annoying about it. Konqi may have confused many, and I would have liked to be able to have seperate settings for web/file mode on opening a window, but it was one of main reasons I liked KDE 1,2 & 3. If KDE4 is going to insist on a dumbed down filemanager, it'll be a bitter pill. Aaron Siego has said that's not happening, but there seems to be a disconnect between user KDE 4 reviewers, and the developers, with developers not able apparently to point to bug reporting users, how to get the expected behaviour when they're setting bugs to INVALID.

blackbelt_jones said...

Yeah, this is a premature release, and I got way too paranoid.

I installed the unstable Debian, which is still running KDE 3.5.8, and I upgrade daily. When the Sid maintainers, who know much more than I do, decide it's time to to go to KDE 4, that's when it'll be ready for me to take a look at.

Until I see the whole thing laid out the way it's supposed to be, I'm not going to be able to form a decent opinion, but my God, if they mess with Konqueror...!

When I ran Kubuntu, I discovered that if I clicked on a link in Konqueror, it would open in Dolphin. Hope they fixed that!

I don't understand how anyone could find Konqueror "confusing". You click on the icon, what's the big deal?

Praveen VK said...

I am disappointed by KDE4. lots of features of KDE 3 are absent. No freedom to control or change many settings. It's looks are good, But we need more than just look. More windows users are using the linux nowadays, I assume that caused the release of lame KDE4.
I liked the freedom and options that were available in KDE3. Now I feel like someone tied my hands in KDE4, I am no longer able to change or control many things.
Now using the simple looking Gnome, because no choice. I agree KDE4 looks good, But restricted.
Please bring back the freedom that we were enjoying.

Aaron J. Seigo said...

@Praveen: "Please bring back the freedom that we were enjoying."

Ignoring for a moment that there are already many more features in KDE4's applications than in KDE3 (so I'm guessing you're talking about a rather small number of things in the workspace apps), you may want to try out KDE 4.1.

blackbelt_jones said...

It's not finshed. I installed KDE 4.1, and I found it pretty frustrating. If you want a finished, workable desktop, don't bother.

But don't condemn it by it's present form. It's not finished. I think it's probably going to be great, but I honestly don't know, and I can't possibly know at this time.

I've said this before, and poor Aaron must think I'm following him around, but releasing KDE4 way before it's ready, and then telling users to be patient, was an epic PR blunder. KDE needs to be responsible for that, but don't mistake bad PR for bad software. Just keep using KDE 3.5,(there's even a new release coming up!) and let KDE4 get made before you issue opinions on it, favorable or not. In the current atmosphere, saying something nice about KDE4 will probably just prod someone else into saying something nasty.

Whatever you do, don't install KDE 4 without previewing it first with a live CD.

Aaron J. Seigo said...

@blackbelt_jones: "I installed KDE 4.1, and I found it pretty frustrating"

i hope you made sure all the frustrating things are reported on bugs.kde.org.

i do wonder what you found pretty frustrating.

"releasing KDE4 way before it's ready,"

define "ready". it wasn't ready for production desktop usage in 4.0.0 and nobody pretended it was.

that was not the *point* of that release.

"and then telling users to be patient, was an epic PR blunder."

it turned out badly PR wise, yes. and you know why? because most of the journalists and bloggers who cover free software right now have very, very little idea of how free software actually works.

they are busy trying to fit our processes into the processes they are familiar with: the processes of companies whose products we're eclipsing in half the time it took them to and with a fraction of the budget.

"KDE needs to be responsible for that"

we need to be responsible for making the right technical decisions and then communicating what it was. we actually did both of those things.

people chose to neither listen nor comprehend, and then turned around and spread all sorts of very lame opinions, where "lame" is defined as "talking about that which you know not".

you might also want to keep in mind that "take responsibility for" implies changing how things are done in the future. while there is certainly always room for improvement, radically shifting from what we did would have resulted in a far, far less useful stack of software in both the near-term (e.g. this year, evidenced by current progress) and in the long term (e.g. 10 years from now).

as you go around advocating your "taking responsibility" meme, you may want to take that into consideration.

you may also want to consider educating the people who *actually* dropped the ball, namely the people who led (and continue to lead) the most amazing storm of half-truths and misinformation i've ever seen come out of the free software community. only the antics of Microsoft and SCO have been more audaciously wrong headed in the last decade or so.

kde communicated clearly, the community beyond our immediate borders decided to play an antagonistic game of us-and-them.

responsibility, indeed.

i wonder if we'll see a follow up blog on as kde4 matures? here we have a panning of 4.0rc1 and then ... silence.

blackbelt_jones said...

I just joined the KDE-promo mailing list, so I'm going to take the liberty of speaking in the second person.

We need to be responsible for our public relations, period. It seems to me that's what public relations is... not leaving it to others what the message is. If we had put the message out correctly about what the release was supposed to be, the public would have gotten the message, period. What can we do better? What can we do to make sure the public gets the message? These are the only questions worth answering, because that's the only way things will get better.

Anyway, I take back all my criticism. After some research and some thought, I may say it all over again, but if I do it will be documented, it will be specific, and it will registered internally, and not publicly.

blackbelt_jones said...

About bug reports:

I'm pretty sure that when I installed that desktop, it was a beta, and there was something on the website that said "there are a lot of bugs, don't submit bug reports". Words to that effect.

But I didn't necessarily encounter any bugs. When you don't understand how the desktop works, that's not really a "bug".

It's not finished, that's all. It's no problem. I'm totally cool with it, because YOU explained to me that KDE 3.5 was still being maintained, and we weren't going to be pushed or rushed. I hadn't gotten that message before that. Now we need to ask ourselves what we can do to get that message out. If the bloggers and pundits don't understand what free software is, those are the conditions we have to work with.

Responsibility doesn't mean blame. It means "the ability to respond". See how that works?

i wonder if we'll see a follow up blog on as kde4 matures? here we have a panning of 4.0rc1 and then ... silence..

You say that like it's a bad thing. Didn't we agree that it's not finished? Do you really want a review for every unfinished development version? Until it's finished,silence is good.

Anonymous said...

@aaron j seigo: "i hope you made sure all the frustrating things are reported on bugs.kde.org."

Well, I tried 4.1.1 today, and it was less than a salutary experience. I had tried 4.0, and didn't like it; I've now tried 4.1.1 and didn't like it.

Problem is, I can't really report "I didn't like it" as a bug. 'Tain't.

Mind you, I think Plasma has a great deal of promise - but right now, I found I couldn't move icons around on it once it started. Knotify4 shot to 97% and stayed there as long as I was moving the mouse (okay, THAT'S a bug, but "knotify4" isn't on the listed programs on the bug reporter).

KDE4 messed with KDE 3.5.9 settings (seriously). Shouldn't have. Maybe that's a kubuntu thing. But you know, the "you can run them side by side" didn't work out. I uninstalled kde4 pretty quick.

Right now, I'm glad y'all are still bugfixing 3.5 (just upgraded to 3.5.10, to get rid of the mis-set 3.5.9 settings).

Because... 4.1.1 is still not ready for prime time. No doubt it will be - but I'm looking for behavior along the lines of 3.5. I want Konqueror to behave in much the same way it does now. I couldn't even FIND konqueror in 4.1.1. Lord alone knows why, but it wasn't anywhere.

My laptop is an old HP n5430; it works fine under 3.5.10 (now). When 4.1.1 sucks 97% of its cycles for internal programs, for minutes at a time - well, I can't use that.

So... sometime around 4.2 or maybe 4.4, I'm sure I'll upgrade. The features are fine - the speed and behavior are not.

blackbelt_jones said...

For Christ's sake, KDE is already forked! We have two KDEs, and they're both valuble and important. Why do we need to put ourselves through this aggravation? For us holdovers, the plan seems to be to continue to develop KDE4 until we can run it as an ersatz KDE3. Is this really the simplest way to proceed? It seems to me like more work for the developers, more complaints to listen to, and I feel like my desktop is under siege.

How feasible is it to keep the maintainance mode going indefinitely? People talk about converting KDE3 to qt4 and crazy stuff like that, but how long could we keep KDE3 going without any significant change and just security fixes? Would maintenance require less advanced skills than development? so that, with a little help, guys like me could do most of the work?

blackbelt_jones said...

I've said this elsewhere, the problem isn't KDE4. It's the sense that KDE3 is living on borrowed time. If KDE 3 wasn't on the chopping block, I'd love KDE 4, and I'd certainly use it from time to time. I love the idea of having a lot of attractive information (time, weather, newsticker) on the desktop for those times when you just want to play internet radio, or whenever I'm not directly working with the computer.

For working, I'm not crazy about Plasma. It feels like a wall between me and the system, and for working, those widgets seem like glorified applets, only applets don't require all that locking and unlocking, moving around and resizing.

But I could easily live with Plasma, since I usually run KDE applications from fluxbox. When I use The K Desktop environment proper, I'm often generating text files, and it's helpful to be able to copy files directly to the Desktop... but for that limited use, Gnome will work just fine... and, like I said, I intend to use KDE4 and enjoy it from time to time. What has compelled me to make such a shrill, shrieking public spectacle of myself in every Linux forum I can find is the fate of Konqueror.

The amazing thing about Konqueror is that it contains just about all the graphical capabilities of the entire K Desktop environment. There are GUI applications menus, a window for detecting and mounting hard drives-- and every directory in Konqueror can be made to behave exactly like the KDE Desktop, with Customized desktop launchers (in Konqueror they're called "Links to applications" and "links to locations").

So I can create my own super-desktop by running KDE applications from a Window Manager, where the menu and keybindings are text files that I can literally write myself from scratch. Thanks to Konqueror, I can bring ALL the GUI power of a full-featured desktop to my light, fast super-customized iconless window manager.

Konqueror is at the center of almost everything I do (except, ironically, web browsing.) I just did a search, and the word "konqueror" in my current fluxbox menu 128 times, either as part of a command or as part of a menu header.

Also, because Konqueror is both a file manager and a webbrowser, there are opportunities to do all kinds of interesting and creative things with html. Essentially, you can do file management with html. It's very cool.

Ah, but in KDE 4.1, The mighty Konqueror has simply been ruined. There's no other word for it. No filter bar for instant locating of files and directories. No support for the "storage media" function. There's also some pretty severe bugs, but the bugs don't bother me, because I know they will be addressed.

However, in the context of all the current shaking up, and the rather paternalistic effort to push konqueror as a web browser and dolphin as a file manager, I just can't be certain where they're going with the star of my Desktop. At times, my thinking takes a conspiratorial turn. I mentioned that Konqueror no longer has a filter bar... but Dolphin does. Is someone giving Konqueror the Tonya Harding treatment, deliberately crippling it so dolphin can prevail? That's crazy, isn't it? I can't believe I even brought it up.

So where's the Konqueror filter bar, motherfuckers?

Dolphin is okay, but it's not Konqueror. It's not all the power of the DE in one interface. And that wouldn't be a problem, except that Konqueror isn't Konqueror anymore. That's more than a problem; it's an emergency, but I can't be sure that anyone at KDE cares.

No, Mr. Nichols, it's not yet "time to fork KDE". After all, we just got a new version of KDE 3 maybe two weeks ago. Only when 3 has been abandoned will it be time to push the "fork" button, and maybe never. The devs tell us that future releases will address all our concerns. I wouldn't be surprised, but it's too important to take for granted.
We should all relax, take extra care to respect each other, trust that it's going to work out. and maybe we should find out where the "fork" button is located.

I'll say it again, the KDE developers are bona fide culture heroes in my book. Future generations may look back at them, and their collegues, with the same sense of gratitude that we feel today toward those monks who copied manuscripts all day long.

The free software model will work this out for us. Forking KDE would be nearly impossible, unless it turns out that a forked KDE is necessary. Then it becomes inevitable. I don't intend to be a regular user of KDE, but I feel pretty certain that we have KDE4 because KDE4 is necessary.flbvg