On Monday, thanks to the amazing help of Jill Ratkevic, Bruce Perens and Xandros, I was able to attend the Desktop Linux Conference to represent and speak about KDE. I followed a talk by Xandros, another KDE based product, and their demonstration was an impressive example of how KDE is being used to do great things. My talk focussed on what KDE is, what it provides, and where we're going with 3.2. It's standard for KDE developers, but important for those who are looking to move to Linux on the desktop. It was a long day!
There was an excellent turnout for the event, and the facilities were ideal. The talk schedule did change quite a bit, both in speakers and in timing. My talk was scheduled for 3:15 but didn't start until approximately 4:00. At that time everyone, including myself, was tired but we made the best of things.
I spent much of the day listening to other talks and talking to an uncountable number of people about Linux and KDE. I could write forever but it would bore you and take too much of my time, so here are some key points I learned:
- We still haven't educated people enough regarding FreeQt. People who should be well educated about these things were still trying to tell me that if Trolltech is acquired, KDE would end up being a proprietary platform.
- Our PR is worse in North America than I had originally imagined. Many people use KDE, few know what it is, and I suspect some think they're using other things even.
- OpenOffice.org wants to work with us. We need to work with them. Let's get going folks!
- Government regulated/approved accessibility is no longer optional. If we don't do it and do it now, we might as well give up hope of seeing KDE accepted into government and corporate institutions.
- We need to come up with an alternative plugin scheme to Mozilla's XP based plugins that can be used across all Linux browsers, and perhaps on other platforms too. Browser plugins are a mess.
- The Linux Desktop, according to business types, is Linux+Desktop+Mozilla+OpenOffice. We need to integrate better with OpenOffice, and we need to better educate people about Konqueror.
- The real target for most businesses (IBM and others) is thin clients and hybrid configurations. Ironically KDE is best suited to this, especially with Kiosk mode (something people were demanding at the conference, and apparently NovellXimianSuse will provide a proprietary Ximian desktop solution for).
- Businesses want a corporate partner to buy their free desktop from. Go figure.
- Novell is really stressing the "ISV support" point. It's as though they're taking a lesson from Microsoft - give away inferior product to destroy the competition and then provide expensive peripheral applications and vendor lock-in. Ironic that they took direct shots at Red Hat for apparently trying to do this.
- We need to be involved with the Desktop Linux Consortium. We need help from everyone to speak about KDE, and additionally about Linux on the desktop in general.
- We need to continue our work with freedesktop.org. The X server enhancements coming from there will be very good too.
- Let's get involved with Fedora. They want our support. There is mutual benefit to be had.
- Linux on the desktop is a 4-6 million machine pie right now. That's 1-1.5% of the market. There's only one direction to go from here. We need to make sure that KDE remains competitive and retains our majority portion of that pie.
- There are some great case studies out there, and we need more of them.
This is a summary of what I learned and saw at the conference. Of course it's not a complete picture but I think it illustrates what the direction is for Linux on the desktop, what we need to do, and what's possible in the future.
I think the report says IBM will be going with Novell/Ximian...
No, it doesn't say that.
yes business needs MS office compatibility. IMHO a key succes factor.
What's wrong with a small champion? kittyhooch.com supports Quanta Plus sponsoring Andras Mantia full time and makes a difference. Maybe if a few more KDE users would get their cats snockered with the best we could support KOffice too. ;-)
While we're at it... the Linux kernel got big with a lot of small supporters and now is crucial to a number of businesses and enjoys many large supporters. Who's to say that when KDE is as old as Linux that won't be the same case? Remember it's the cummulative strength of small supporters that convinces large supporters they can't beat it so join it.
Xandros uses KDE, but:
Xandros web browser: Mozilla
Xandros Mail client: Evolution
Xandros Office suite: OOo
If they're going for a simple, integrated desktop, I would think they'd want to use KDE apps as well...
Also, it sounds like they've added lots of neat technology like hardware detection, fast user switching, non-destructive NTFS partitioning, config gui's (for wifi network, dual-head display, etc)...are they sharing any of this with the community, or is it proprietary? (not necessarily complaining either way, just curious)
Xandros 2 was just released and it sips with KDE 3.1.4, but I do wish they would ditch XFM, it is better than Konqueror for many taks but keeping it better is a huge taks considering how many people work on Konqueror.
"Our PR is worse in North America than I had originally imagined. Many people use KDE, few know what it is, and I suspect some think they're using other things even."
It's not KDE's PR! The only one of the big player (really) supporting KDE was/is SuSE. Everyone else was using GTK/Gnome.
"OpenOffice.org wants to work with us. We need to work with them. Let's get going folks!"
Really? Q's roadmap says something about Gnome integration!
"Government regulated/approved accessibility is no longer optional. If we don't do it and do it now, we might as well give up hope of seeing KDE accepted into government and corporate institutions."
Everyone said KDE's accessibility features are way ahead of Gnome's. So how come this? Just curious...
"We need to come up with an alternative plugin scheme to Mozilla's XP based plugins that can be used across all Linux browsers, and perhaps on other platforms too. Browser plugins are a mess."
That's not specific to KDE, plugins are generally a mess under Linux
"The Linux Desktop, according to business types, is Linux+Desktop+Mozilla+OpenOffice. We need to integrate better with OpenOffice, and we need to better educate people about Konqueror."
There's Mozilla, Firebird and some other Browser that are really good. So why use konqueror? I also use KDE but it never came to my mind to use konqueror when having the above mentioned alternatives. Why trying people eduacate about konqueror when most if them even recognize Mozilla/Firebird (under Win)
"The real target for most businesses (IBM and others) is thin clients and hybrid configurations. Ironically KDE is best suited to this, especially with Kiosk mode (something people were demanding at the conference, and apparently NovellXimianSuse will provide a proprietary Ximian desktop solution for)"
Evolution may possibly be not better than kmail but it's way more "business ready"
"Businesses want a corporate partner to buy their free desktop from. Go figure."
Hmmm... I'm thinking of RedHat/Fedora, IBM, Sun an many more others. No one I know uses SuSE/KDE in their business
"We need to be involved with the Desktop Linux Consortium. We need help from everyone to speak about KDE, and additionally about Linux on the desktop in general."
Already on the way AFAIK!
"Let's get involved with Fedora. They want our support. There is mutual benefit to be had."
Why not a 2 years ago? RedHat pi**ed on KDE as we all know, so why didn't the KDE team built *working* packages for RedHat's distro?
"Linux on the desktop is a 4-6 million machine pie right now. That's 1-1.5% of the market. There's only one direction to go from here. We need to make sure that KDE remains competitive and retains our majority portion of that pie."
Business ready? OK, stop implementing useless features, stop the bloatware, make KDE rock solid! Implement features needed in everydays business, not gimmicks for geeks! That's the way to succeed IMHO!
Maybe you guessed it, I am a Gnome user, but until 3.1 I would have died for KDE. But now I miss some kind of professionality, ease of use. The main goal of KDE now seems to be "more configuration options" and not "business orientation".
Anyway, KDE overall is a nice piece of work
> It's not KDE's PR! The only one of the big player (really) supporting KDE was/is SuSE. Everyone else was using GTK/Gnome.
Hmm.. really? SUSE is the major distro in Europe- they use KDE. Connectiva is the major distro in South America - they use KDE. Turbolinux is the major distro in Asia - they use KDE. I really don't know about Africa or Austrialia, but that leaves the other major continent: North America. RedHat is the major distro in North America, and they use GNOME.
Thankfully, Fedora is a community project, and they have seemed to be much more opened up about KDE than RedHat was. Notice that KDE support in other community-based distros is *very good* (Debian and Gentoo).. hell, you even have KDE developers maintaining the packages for these community distros (Ralf Nolden of kdevelop fame, for Debian, and Caleb Tennis, also of kdevelop fame, for Gentoo)
Remember Lycoris, Lindows and Xandros all North American and all target the desktop unlike Redhat. Guess what they run? KDE...
Don't forget Mandrake and Knoppix, which use KDE by default I think and are still rated #1 and #3 on distrowatch.
RedFlag, the leader distro of China, uses only KDE as its desktop environment. It shipped more than 1 million desktop copies one year and occupied about 30% of Linux desktop market.
I only browsed through it, but something out of your stupid rant really stood out:
"Maybe you guessed it, I am a Gnome user, but until 3.1 I would have died for KDE. But now I miss some kind of professionality, ease of use. The main goal of KDE now seems to be "more configuration options" and not "business orientation"."
Where the fuck did you get this from? All signs point the opposite way, have you paid any attention to KDE development since 3.1? They have been working hard to make it easier to sue, removed a lot of context menu bloat and many features aimed for businesses mostly.
Don't just spit out the propaganda that the GNOME camp throws at its cluelss users. Think. Try. Say something smart.
The easier to sue part is only for SCO in particular.
> "The Linux Desktop, according to business types, is Linux+Desktop+Mozilla
> +OpenOffice. We need to integrate better with OpenOffice, and we need to
> better educate people about Konqueror."
> There's Mozilla, Firebird and some other Browser that are really good. So
> why use konqueror? I also use KDE but it never came to my mind to use
> konqueror when having the above mentioned alternatives. Why trying people
> eduacate about konqueror when most if them even recognize Mozilla/Firebird
> (under Win)
Why educate people about Konqueror?
To show them an alternative to Mozilla. If they still prefer Mozilla, fine. But they should at least know that there's an alternative which probably much better suits there needs.
Why use Konqueror?
Because it starts much faster than Mozilla. Because it renders faster than Mozilla. Because it's highly integrated with KDE. That's why a KDE user should use Konqueror.
BTW, people who come from Windows don't recognize Mozilla/Firebird. They recognize IE.
It's a good idea to let people know that Konqueror exists. It's a bad idea to make them feel like they need to use Konqueror if they're using KDE.
It's the nature of Linux for things to be mixed and matched and for a variety of configurations to exist. That enables a healthy ecosystem.
I prefer KDE as a desktop environment and use many of the included programs. But there are also programs that I prefer using over the KDE equivalents, such as:
My point is that this diversity is good and necessary. We don't want to reduce ourselves to using an impersonal integrated, homogenous environment like Windows and Mac users do. Ditto about standardizing a desktop environment.
What is wrong with KVim?
I hope you know yourself that what you were writing there is pretty funny. The more audible audience nearly always complains about the lack of consistency and integration, ignoring the fact that with KDE they'd get a pretty complete desktop environment where consistency and integration is no issue at all. Instead they do make excessive use of diversity and then keep complaining about the lack of consistency and integration as if it isn't available or it's others' fault. *shrugs*
> "Government regulated/approved accessibility is no longer optional. If we don't >do it and do it now, we might as well give up hope of seeing KDE accepted into >government and corporate institutions."
>Everyone said KDE's accessibility features are way ahead of Gnome's. So how come >this? Just curious...
Not sure who you've been talking to ;-) Or perhaps you accidentally inverted the comparison above? I would be interested in hearing more details about the accessibility requirements, i.e. who's saying it and why. I do agree that this is becoming a very important requirement.
KDE accessibility has come a long way in the past few months, but it mostly provides a few utilities that are helpful for certain limited disabilities. That's a long way from true platform-level support for theming, low vision, and assistive technologies. ( http://developer.gnome.org/projects/gap/AT )
True accessibility means that blind users, users who cannot use a keyboard at all, and users with other serious disabilities can use the complete desktop; KDE still doesn't have any support for this level of accessibility (though it's on the roadmap). The plan for KDE accessibility's assistive technology support is to support and interoperate with the 'GNOME' accessibility APIs, but this KDE/Qt work hasn't begin yet.
On the other hand GNOME has been adding such support since its 2.0 release, and now includes bundled assistive technologies which work with an ever-increasing number of core GNOME applications. This isn't meant as a flame in any way, it's just that GNOME started work on platform accessibility well in advance of the KDE accessibility effort. However with concerted effort and focussed work I don't see why KDE can't leverage and reuse much of the GNOME accessibility work; there's really no reason IMO why an assistive technology needs to be tied to one plaform or the other, provided the same services and interfaces exist regardless of which 'desktop' the user is running or toolkit the developer is using.
Architect and Technical Lead
GNOME Accessibility Project
Sounds right. I've done some digging around and ATK support is considered a MUST-HAVE for Qt4 by Trolltech and accessibility support for KDE4 is considered VITAL by some important KDE developers.
Bill, since you have a nice setup over at Sun Microsystems, would you guys be willing to help us test the code, etc? I don't imagine it will be easy for our developers to do so without proper hardware support... most of us just don't have those special resources.
With Sun sponsorship, a big hurdle would be removed.
actually we don't have a lot of hardware either, in our main development site for accessibility. Fortunately it's not required, since the gnopernicus screen reader provides a "braille monitor" which is a software braille display, and everything else used by gnopernicus is software-based.
The hardware required for testing GOK is also fairly straightforward, in that most GOK features can easily be tested with a USB trackball or similar device. It's worth having a head-pointing device (Tracker USB) which costs about $1000, but not essential to testing.
> KDE accessibility has come a long way in the past few months, but it mostly provides a few utilities that are helpful for certain limited disabilities. That's a long way from true platform-level support for theming, low vision, and assistive technologies.
Well said. There are many improvements in KDE 3.2, including improved support for people with low vision. But most things still need to be done, e.g. support for blind users.
> This isn't meant as a flame in any way, it's just that GNOME started work on platform accessibility well in advance of the KDE accessibility effort.
BTW, Bill Haneman attended our first IRC meeting to help us getting started last year.
> However with concerted effort and focussed work I don't see why KDE can't leverage and reuse much of the GNOME accessibility work; there's really no reason IMO why an assistive technology needs to be tied to one plaform or the other, provided the same services and interfaces exist regardless of which 'desktop' the user is running or toolkit the developer is using.
The current ATK implementation has several ties with GNOME, but this can easiliy be changed, as Bill and other involved people affirmed.
Olaf (Co-maintainer of the KDE Accessibility Project)
> Everyone said KDE's accessibility features are way ahead of Gnome's. So how come this? Just curious...
Maybe you mean usability, not accessibility?
The KDE Accessibility Project is still very young - about a year old by now.
We soon released some small utilities, but the really crucial work is still far ahead. Fortunately, Trolltech promised to support ATK in the next version of Qt. GNOME's ATK is currently supported by Java and GTK2, so real interoperability is possible. And by reusing existing solutions, we hopefully can catch up quickly.
Olaf (Co-maintainer of the KDE Accessibility Project)
I am not sure why OpenOffice has to be integrated into KDE. It seems that if OpenOffice supports GNOME it should automatically plug right in to the KDE desktop - at least this is what I understand the goals of FreeDesktop.org are.
Am I wrong in this?
Yep. Open Office does not support GNOME.
Sun is pushing for better platform integration for OpenOffice in the roadmap. In Sun's case that means better integration with GNOME, but they are open for KDE support as well, that is, if someone is going to write the code for that.
Cuckooo is an attempt to do just that, see http://artax.karlin.mff.cuni.cz/~kendy/cuckooo/ There was some talk about it becoming an OpenOffice incubator project but I'm not sure what the status is of that.
If you are interested in helping out with this you may want to drop a mail.
George, thanks a lot for your work. I think you're filling a gap right now, and you're doing it with an apparent enthusiasm that's truely inspiring. Rock on, KDE!
George Staikos is one of the people who wait a long time to start working on something, but then do it brilliantly. I will except kwintv, I am really not aware why that is so hard to do.
There are many great and probably not *very* time consuming things that get postponed to the "next release". I figure most KDE developers have jobs, and even though when their skills are applied tremendous results are produced, it also depends on their personal interest. And there are great features that wait a long time to complete, while they have taken shorter in other projects.
Still, KDE has an extremely mature development. The hardcore base is never overlooked and that is what gives the overall sense of stability and cohesiveness.
Unfortunately (or not so), those with a lot of time in their hands, choose GNOME.
Thanks guys :)
Regarding KWinTV, we had some parts working quite well, but it's quite a huge task and I have much more important things to deal with at the moment. There are too many major areas of KDE that need work and KWinTV is not what I consider critical. Also I haven't had a working TV tuner here in months so it's difficult to work on the rewrite. :-) Dirk Ziegelmeier is working very hard on kwintv now and has joined maintenance and lead development. There are only a few major bugs left and we can start to ship stable releases. Unfortunately one of them is quite huge (lacking a usable V4L driver), so it will take some time to sort out.
"Businesses want a corporate partner..."
Well, yes, they want a stable product line that they can trust, won't disappear tomorrow, and someone to turn to when problems occur. Now actually KDE provides a pretty stable product and a steady flow of updates which is good but there's not even a basic user's forum - Where can you post questions? (and I don't mean developer's mailing lists) Who do you turn to when problems arise? Some KDE newsgroup forums and FAQ's should be high on the list. Hopefully the KDE team leaders can get some major corporate partners to sponsor and lend corporate credibility to KDE.
"Fedora.." - There should be a packager for all KDE stuff for Fedora unless of course, all of us previous RH users keep switching to Debian, in which case make sure all packages get into Debian and get to Testing asap.
Where are the Developer Tools? People seem to overlook the significance of MS putting a big emphasis on having excellent developer tools for Windows. How does the corporate world move to KDE without great developer tools. KDevelop3 has great promise but has been in alpha for a long, long time and still is not available for RH, Fedora, etc What does leave for developers?
> but there's not even a basic user's forum - Where can you post questions?
http://kde-forum.org, this shows that you never looked for it eg at http://kde.org
> "Fedora.." - There should be a packager for all KDE stuff for Fedora
> Where are the Developer Tools?
Modules kdesdk and kdevelop
just my 2 cent:
the admins in corporate networks *hate* "a steady flow of updates". in a reasonably large installation, there are just too many people who have their head full of stuff they call work, and juggling rpms is not notmally part of that scenario.
when you want to sell, remeber how you want to buy...
> Where can you post questions? (and I don't mean developer's mailing lists)
> Where can you post questions?
> Who do you turn to when problems arise?
Search no more: http://www.kde-forum.org/
Okay, I stand corrected - there is a rather extensive forum. And it's my fault for not finding it BUT as a long time KDE user I didn't know about it and roughly judging from the number of posts (compared to many other linux newsgroups), I'm apparently not the only one, soooo, what that means it might simply need to be better publicized etc. Maybe we need to encourage more of us users to write up stories and how-tos about how we make use of the many features throughout the KDE project and get them published in popular sites such as user groups, distribution install how-tos, pclinuxonline, linux journal, linux forum (which does have many excellent how-tos such as GIMP) etc. btw we are one of those unheard of US KDE users :)
re steady stream - sure, admins want to minimize updates, but what I meant was that KDE is active, constantly improving etc - if there's a major problem, it's bound to be fixed shortly. It's much better to choose when to update, than not to get updates. In short, you can count on it for your business - and we do! - We depend on the desktop, Konqueror browser, KMail, etc every day and it has worked well (with the exception of KPilot which completely screwed up our data but I think that has gotten fixed in 3.2) And I think the lastest KDE is pretty well supported in most of the Distro releases. But if a sys admin has to go before the exec committee, they need to explain who is behind the product they want to use and what guarantee of support are they going to get.
I didn't know about the forums until now, either.
And I've been wishing for something like that.
Could the progress made in these points maybe put on a seperate page like the 3.2 features and 3.2 status list, so it doesn't get forgotten next week. It would give a good overview of the direction KDE develops and how it succeeds.
This is NOT a flame, please bear with me here. I need to know why Kiosk Mode would be important to businesses. I thought locking down a computer in business meant iinstalling only the required software, and making sure no other software can be installed. since a computer is usually assigned to one person, is ther really a need to lock down computers much further. PLease educate me? :)
valid question. a particular company may want consistency across desktops.it might require everyone to have the same wallpaper for example and not waste time tweaking stuff
One simple example: You want people at a conference to be able to browse the web, but not to close Konqueror. Without the kiosk code this is not possible.
Its to reduce service-effort.
Most pwople are not as familiar to computers as you might be - if you allow them to tweak theyr desktop they might end up in a mess, helpless to fix their ill settings. thats a problem companys want to avoid.
Kiosk mode is very important to two kinds of business use. The first and most obvious use is when a business makes a terminal available to the public as in a mall or an airport. In this case, the terminal's user needs to be restricted to do only the usually single function that the business wants to offer the public. It wouldn't do to have a kiosk that prints maps at a car rental agency reconfigured to get email and play frozen bubble by some twit with too much time on their hands, there by preventing the next person from printing their map.
The second business use is to control the user's desktop and reduce support costs. Companies spend thousands of dollars per year per user on support costs and many of the problems are self inflicted by the user. Countless hours are wasted explaining to users how to get back the icons that they lost (moved off the edge of the screen) or reconfiguring their themes so they no longer have black text on a black background making it impossible to read.
There are liability issues too. If you look at http://www.kde-look.org you will find a wide variety of themes and wallpapers that people enjoy on their desktop. But, many of those themes are inappropriate for business use. A company can be slapped with a sexual harrassment suit if you display some naked anime as your destop wallpaper, costing the company thousands or millions of dollars.
There are also issues with people installing their own software on company systems. This can cause all sorts of problems such as licensing, system instability, viruses/trojans/worms and also wasting company time. They aren't happy to pay you to play frozen bubble. It costs them money when you download some app off the web that clobbers a library and now the corporate app won't run and support spends half the day fixing it. Businesses and support departments greatly prefer to lock down the desktop so that the user can change nothing at all and can do only what their business function is with the computer.
Indeed Kiosk mode is very important to business and managing it centrally for the entire enterprise is also critical. This is what Microsoft's Group Policy Objects does and also what Novell ZenWorks does plus a whole lot more. And, almost every company with more than 75 desktops is unwilling to run their enterprise without these features.
Perhaps SOT Office may become a KDE OpenOffice as they didn't define their market nice. Waht makes this finnish package so special. I would like to use my KDe buttons in OO as well as the OO GUI looks outdated and broken. I would like to see OO as a Kpart...
Same applies to Mozilla Composer, now NVU.
Imagine you have a company. You want to install linux because you don't want to be forced to follow Ms decisions, you want to spend less in licenses and try to minimice virus-related troubles.
You have legacy hardware which does not have a linux proper driver, and legacy software, difficult or impossible to run on Wine.
So, you decide to make a partial migration:
- Some servers, work on linux.
- Most desktops, migrate to Mozilla and Open Office on Windows.
- Work on legacy apps to do them multiplataform, replace them or make them run on linux by some kind of emulation. Same thing on legacy hardware.
So you have a mixed environment, and open source multiplataform apps are key. In conclusion, doesn't matter if KOffice-Konqueror is better in Linux, the important thing is that they are not as good on Windows.
That's why integration and work with them is important: they (Mozilla+OO) are the door througth which users came to Linux + KDE on desktop.
No, it is more a "let's try" thing. If it works, fine. If not, you will immidiately switch back,
- key applications for a good first impression (birthday card printer :-)
- plattform unification
- "stable plattform", rpm for all distros. You can run a windows programm from 95 on WinXP, but you shouldn't install a suse7.3 rpm under suse 8.2. And there is one package for each distribution. this may make sense for system software but not for applications.
It appears to me that the question is whether the Fedora project wants to support KDE.
The current Fedora disto is basically just a retread of RedHat 9.0 with some updates.
If they are going to work with us, they are going to have to be willing to have a stock KDE distro rather than the crippled one the RedHat was using.
The RPMs on the Fedora site are new and they look like a proper set of KDE RPMs but I don't know what is in them -- at least they aren't sliced and diced like the ones for RedHat 8.0.
Just found this (don't know, if anyone else post this)
IBMs Desktop strategy: