Linux Online
[ Register ]

[ Applications ]
[ Documentation ]
[ Distributions ]
[ Download Info ]
[ General Info ]
[ Book Store ]

Advertisement

[ Courses ]
[ News ]
[ People ]
[ Hardware ]
[ Vendors ]
[ Projects ]
[ Events ]
[ User Groups ]
[ User Area ]

Linux Kernel Development

[ About Us ]
[ Home Page ]
[ Advertise ]

Linux Online Interviews

This week Linux Online interviews David Faure. David is a developer for KDE. He is the maintainer for apps such as kfm and Konqueror. Konquerer is the promising new web browser designed by the folks at KDE. He has also given lectures on the Linux circuit around the world. Linux Online interviewed him about himself, his work for KDE and Linux and the future of KDE.

Linux Online: You've just released KDE 2. If you've been using the KDE 1+ builds, what important improvements are built into KDE 2?

David Faure: KDE 2 has been under development for 18 months. It's a major release, very different from KDE 1.1, so there are too many improvements to list them all. The most important improvement for the user is probably the full-featured web browser, Konqueror, which has reached a level where it can honestly compete with the other major Web browsers out there. The first release of KOffice is a major step forward too, although it may not yet be as fully-featured as the well-known commercial office suites. The KDE 2 desktop is also much more configurable than KDE 1 used to be, and it's also less memory-hungry, thanks to a better design.

But I think that the improvements are even greater for the developers of KDE applications. The KDE libraries provide much more than they used to, and with a much better design. This leads to better applications for the user in the end, of course :)

Linux Online: Is there any specific application or feature that the KDE people are particularly proud of?

David Faure: I think Konqueror is what most of us are proud of. You have to realize that konqueror is more than one application, it's the combination of several applications (file manager, web browser, generic viewer), and several helper applications, those that support the various protocols supported by konqueror (HTTP, FTP, SMB, etc.). This means, there are many KDE developers that contribute indirectly to konqueror, and their work is also used by other applications (KMail uses the same HTML engine, all applications gain network transparency the same way as konqueror does...).

Well, this is my view on this, I suppose other developers may be proud of another application. I am a bit biased, being the maintainer for Konqueror.

Linux Online: You're known as the 'bug-fixer' at KDE, or as you say yourself, the one that does the 'dirty jobs'. When you track down and fix a particularly nasty and elusive bug, what's the first thing you do: Cry out 'Eureka!'? Crack open a nice cool beer?

David Faure: That would probably be "cry out Eureka", and send the fix as soon as possible to the source repository. I don't drink alcohol, so the second option doesn't apply :) And the nicest thing after the fix is sent, is that usually other developers will notice the change, and congratulate the one who fixed it. That's the very good spirit of open source, and the reason why it's so much fun !

Linux Online: They say that no software can be 100% bug-free. Are you one of these developers who takes the 'Canadian Mounties' approach: 'I always get my bug? Or at least give it your best shot?

David Faure: You mean, do I drop the matter when a bug turns out to be too hard for me ? Well, it depends. If the bug is really important, I can spend a huge amount of time tracking it down. I add lots of debugging output everywhere until I figure out what's happening, so it can take a very long time. On the other hand, it happens that I drop an investigation, when the bug turns out to be outside my field of competence. At that point, if the matter is important or urgent, I try to find someone who can have a look at the problem.

Linux Online: Did you just sit down one day and say: "I'm going to work on KDE?" How did you start developing for KDE?

David Faure: I gave the long answer to that in my interview on the KDE web site, so let me give the short answer this time : I discovered KDE when I was looking for a way to give a dialog box to my talk daemon, which has now become "ktalkd". But I really got involved when I took over the maintainance of kfm, konqueror's ancestor. For the record, my first contribution to the open source world was some code in Wine [the windows emulator], to handle non-US keyboards. But after that, I didn't know enough about Windows to carry on with Wine, whereas helping to improve KDE is a much easier job.

Linux Online: When you're programming, is there any music in particular that you listen to that helps the process along?

David Faure: Yes, I'm a fan of Jazz-Rock. Miles Davis, Weather Report, etc. and also of earliest Jazz. But I noticed that when I listen to music I can't think as fast, so I usually don't listen to any.

Linux Online: What's the most difficult part developing for KDE?

David Faure: I think the most difficult part is the continuous flow of incoming bug reports. Especially since we introduced that bug report dialog which enables users to send us feedback very easily, the number of bug reports per day has been increasing a lot. And since most of those are really bugs, what is difficult is to realize that the software will never be perfect, whatever we do. But we try !

Linux Online: What's the most rewarding and fun part?

David Faure: I think the most fun is the feedback from other developers and from users, and working with people from all around the world. The most rewarding is to know that what we do will be useful for many people (I mean a really huge number of people, nowadays), that it will be available to them for free, and that the open-source movement is reaching a level where it has definitely become part of the history of computer science. It will hopefully be remembered as "thousands of people all over the world, working together on creating what they want to use, and sharing it with everyone". That's what it is; what I want to say is that it has become important enough to be remembered even in the far future.

Linux Online: Let's talk a bit about where KDE's going. There are some that say that the day when Linux comes out on top of the OSes will be due to KDE. Do the KDE developers have that in mind when you work? Do you have the impression that there's a lot riding on what you do?

David Faure: It is definitely proven that KDE has helped immensely getting many people to try Linux. It's hard to realize when you are Unix-literate, in which case you know how to use Linux without even a graphical interface, but for sure, most people never used a command line to use a computer, and need a graphical interface. That's why KDE is major step forward for Linux, making it possible to have Linux on everyone's desktop, one day. We're still far from there, but more people know about Linux every day, and the software on Linux is improving at a great pace.

Linux Online: Along the same lines. The general consensus is that newcomers to Linux are specifically installing your product because it acts a lot like what they're used to (not mentioning any names, of course - not yet anyway!). Do you think that's true? Is that what you're aiming at?

David Faure: Yes and no. KDE aims to take the best of all the existing OSes and desktop environments, add its own improvements on top of that, and make the whole lot configurable :) What I mean is that although on the first startup, KDE may look a bit like Windows (let's name it), already improved by the virtual desktops among other things, the user has the choice between leaving it this way, or changing it to whatever he/she prefers. With KDE 2, if you select the "Platinum" style, and the mac-menubar on top of the desktop, you can really get the impression you're using a Mac. Or you can go for CDE, or NextStep.... or something completely different.

Linux Online: The corporate end user will be looking for other things. It's apparent that they've found stability with Linux itself. To finally win the corporate hearts and minds, what does KDE have to offer the business side of computing?

David Faure: I think that what corporate users are looking for, in the application domain (let's not talk about web servers and databases servers), is software such as a mailer, a web browser, and a development environment (IDE) for those who plan to do development. KMail, Konqueror and KDevelop answer those needs very well, in my opinion.

But above all, the most important is a full-featured office suite compatible with Microsoft Office. KOffice is in good shape, has a very good underlying design, but it lacks more developers to make it more feature-complete. Hopefully this will improve in the future, but more developers wouldn't hurt, that's for sure. Let's go for a bit of recruiting: if you're reading this, you know C++, and you want to be part of a fun project, contact koffice@kde.org for getting involved :)

Linux Online: Your official bio at KDE lists your favorite movie as "Back to the Future" In the movie, the main character's father, George McFly, a somewhat nerdy, geeky guy, knocks out the town bully Biff with a left cross, changing the future in his favor. Do you think we're close to seeing Linux giving that OS bully (still not mentioning names) a left cross? Or do you think Linux's triumph is going to be a full 15 round bout.

David Faure: For the record, I like this movie because it's about time travel, not because I identify myself to George McFly :) I don't think I'm following the OS market well enough to have a valid opinion on this. I see Linux growing rapidly - compare the number of Linux Expos currently with the number there were 2 years ago, it's amazing, there is one every 15 days now ! - but it doesn't look like Linux is taking over Windows yet. About the future, I'm quite confident, since it's looking better every day, but to be sure, could Emmet Brown lend me his time machine, for a quick jump 5 years ahead ? :)

Linux Online: Redmond, Washington (OK, we've mentioned names) is obviously worried. There seems to be a pinzer movement about. On one hand Microsoft just grabbed up a nice piece of Corel and they're planning their strategy there. On the other hand, they've openly started panning Linux by running the infamous 'mutant' ad in a German magazine. Do you think this means that Microsoft is finally saying: 'We're in trouble'?

David Faure: The fact that Microsoft bought a good part of Corel, in my opinion, only means that they realized they could make money by selling software for Linux, in addition to all the rest. Just like MSOffice for the Mac didn't mean Microsoft was endorsing MacOS at all. This actually makes sense, as part of the separation between Windows (the OS) and the Windows applications. The part of Microsoft that develops applications has definitely everything to gain from supporting Linux. Whether this means they're worried for the OS, you'll have to ask them :)

Linux Online: I'll have to ask you about the KDE vs GNOME debate. I guess it's sort of a mandatory question and thanks for obliging me. What's the deal here? Is it a debate about purists vs pragmatists? Overblown hype? Somewhere in the middle?

David Faure: No, it used to be a debate about licenses - I don't think it's necessary to go into this again - but that's over now. Even the most purists about licenses don't have anything to say anymore, since Qt has been released under the GPL. This is good news for everyone, it means we can get along in peace and concentrate on making our respective desktops better and more widely known. Free software is all about choice, and there is choice at every level. For your OS you can choose between Linux and BSD, for your desktop environment you can choose between KDE and Gnome, and to read your mail you choose between pine, mutt, elm, xfmail, kmail, and whatever gnome's mail reader is :-)

Linux Online: Now that KDE 2's out there, how's work going on KDE 3? Are you guys taking a little break or is it full steam ahead?

David Faure: When you've been bug-fixing for a long time, the first thing you do after the release is done, is to start on the things you wanted to develop for a long time but couldn't because of the "feature freeze". Since the KDE 2.0 release (the internal one, not the public one), I have already re-engineered the support for compressed tar files, changed Konqueror's bookmark implementation to save the bookmarks into a single XML file instead of a myriad of small files, and I'm currently in the process of writing a bookmark editor. It's the best part of the development, when we can improve things without having to worry about release schedules. But all this work will hopefully be released soon, as KDE 2.1.

There's no plan for KDE 3 yet, fortunately. Remember that a major release usually means incompatibilities - just like KDE 1 applications need to be ported to KDE 2 (they still run using the KDE 1 libraries, of course, but they don't take advantage of the KDE 2 features). With the great design of KDE 2.0, I hope we won't have to go for a major release for a long time, and that we'll be able to deliver improvements for the KDE 2 desktop and applications, as KDE 2.1, 2.2, 2.3 etc.

Linux Online: Computer programs are pretty sophisticated now but there are things that computers still can't do. As a developer, if someone gave you an unlimited supply of money and a lot of time, what kind of new program would you develop?

David Faure: Good question.
I think the most limiting factor in computers are the input and output devices. Keyboards are convenient for typing text, but there are cases when it's still faster to use a pen - for instance, to draw diagrams, or music. Those two examples show that using keyboard and mouse isn't the best way to create non-textual documents. It takes an awful lot of time to enter a few notes on a music editor, much more than it should in my opinion. One alternative already available is to interface with the "manual" systems, such as scanning the diagrams from paper or connecting an electric music keyboard to your computer for entering music. But this means you have to correct the imperfections afterwards, so it's not ideal. I don't know what the solutions are, but I think this is the domain where there is the biggest room for improvement, both at the device level and the application level.

Well, teaching my computer how to cook would be a good idea too, it would save a lot of time :)


Once again, our thanks to David for taking time out of his development schedule to answer our questions.

You can find information about KDE at their website http://www.kde.org



Comments: feedback (at) linux.org
Advertising: banners (at) linux.org
Copyright Linux Online Inc.
Compilation ©1994-2008 Linux Online, Inc.
All rights reserved.