August 16, 2004

Sneak preview of KDE 3.3: Q&A with developer George Staikos

Author: John Knight

With KDE 3.2 reaching a status of stable and mature, a lot of work has been going into the next and final release of the 3.x series, 3.3. Because 3.3 recently hit beta status, we caught up with KDE developer George Staikos to talk about all things KDE.

NewsForge: With the previous KDE release (3.2), there were lots of CPU optimizations that brought the minimum hardware requirements down. What kind of minimum and recommended specs are we looking at for 3.3?

Staikos: I don't think anything has changed in 3.3, and I wouldn't really expect it to. KDE works well on most of the hardware out there, and changing that would likely require us to remove many of the features that people have strongly requested over the past few years. The minimum hardware requirements are actually hard to define, given all the different types of hardware that KDE runs on, and the huge variety of hardware in existence. Chances are if you have a 300MHz or better processor and 128MB of memory, KDE will run quite nicely. You may not be able to take advantage of all the fancy features such as transparency, high resolution, and high-color images, and anti-aliased fonts, but it should certainly be quite usable. KDE can run with less memory as well, but I don't recommend it.

NewsForge: What is the driving force behind this release: features or speed?

Staikos: Actually KDE PIM (Personal Information Management) was one of the big focal points of this release. An incredible amount of work has been done on all of the PIM components -- KOrganizer, KAddressBook, KMail, Kontact, groupware, resources, and more. We have definitely seen speed improvements, too, especially in Konqueror file browsing, KMail, and the IMAP I/O slave. Optimization work for 3.3 is still ongoing, and I expect to see more.

Other big features for this release include a Google search entry field in the Konqueror toolbar, search-as-you-type in KHTML, a news ticker sidebar, some new edutainment applications, and a big list of Kopete improvements.

NewsForge: Do you think KDE has a strong commercial future?

Staikos: KDE is really a non-commercial project. Almost all development is done in a non-commercial manner. Ironically it seems to be the most popular choice for commercial Linux vendors. I think most use it as their default desktop environment, and virtually all of them ship it as least as an alternative. So in that sense, it does have a strong commercial future -- as a part of a platform.

NewsForge: Recently Konqueror has been getting many improvements, especially with the help of Apple with its Safari enhancements. Is Konqueror gaining in popularity, and is it being favored over other heavyweights, like Mozilla?

Staikos: I think Konqueror is maintaining steady growth in use but probably only parallel to Linux desktop growth. Mozilla is the one experiencing the real popularity growth, as it moves into the Win32 userbase and saves people from Internet Explorer viruses and worms. Safari is also a very popular choice for most Mac users. They have implemented some new features in their latest Webcore that might find their way into KHTML sometime -- canvas, drag and drop, and an upcoming new plug-in spec, to name a few.

Screen shot of KDE 3.3.

NewsForge: KDE 3.2 introduced the Wallet system. How has feedback been so far, and have many changes had to be made for 3.3?

Staikos: There have been some bug fixes and a few new features. The feedback is definitely plentiful, but most of the real work has to be done in application integration. Only a few places have KWallet integration right now: KHTML, KIO, KNode, and Kopete come to mind. Hopefully we will see improvements in each of those apps, and maybe even KMail integration soon.

NewsForge: The current trend with most distros seems to be to modify the normal desktop and make it like something else, perhaps Windows. Is it disheartening to have such heavy modifications on the desktop from a large proportion of distributions? Is this having any effect on KDE's creative direction?

Staikos: No, definitely not. We go to great lengths to make this possible in KDE! We don't want to dictate how you use your computer. This is up to you. We present to you a set of defaults that we like. If you don't want to customize, you don't have to; we respect your right to choose. I think it's particularly interesting to see just how far KDE can be pushed, bent, styled, and reworked to accommodate different requirements. The only way this impacts KDE is if there is a need for more flexibility in an area, but I think this is rather rare these days. Creative direction is still very much driven by the developers (coders, artists, etc.) of KDE.

NewsForge: One of the big things among desktops these days is freedesktop.org standards. Are there any further changes being made? Or is KDE fully freedesktop.org-compatible? If not, is it a project being worked on?

Staikos: Freedesktop.org has a huge number of projects under way. It's a bit overwhelming. I think most of them remain unimplemented or unused by most projects so far. These things take time. KDE attempts to implement support for these things where technically feasible, when there is motivation by developers, and where it makes sense. We do actively contribute to many of the specs and standards there, and we have implemented several already. There is always more work to be done, though.

NewsForge: KDE has been integrating GNOME menu entries in recent times. Are there any other further efforts being made toward GNOME co-existence?

Staikos: Actually it's not that KDE is integrating GNOME menu entries. It's more that we have agreed on a specification for the menuing system. I think you'll agree that's a much better approach. What it means is that we both write our menus out and read them in using the XDG specifications. It consists of simple INI-like files for entries and XML files for menu structure, in a specific directory structure. See this site for more details.

There is plenty of ongoing desktop integration work at this time. KDE4 is just around the corner now -- I predict late 2005 -- and we will see much more of it then as KDE gets a chance to break binary compatibility.

NewsForge: GNOME and KDE both have office suites, and OpenOffice.org is available to users. Is there much concern about duplicating work?

Staikos: Yes there is, and we are starting to share the file formats with OpenOffice.org. We will use the Oasis file format as the native KOffice format in the next release (either 1.4 or 2.0 -- undecided at this time). KOffice also shares a hyphenation library with OpenOffice.org. I guess there is some concern about duplicating work, but for the most part they really are different apps with different design philosophies.

NewsForge: Do all three projects share code and algorithms, especially with reverse engineering? Do you think that all three will eventually become one standard?

Staikos: No, that was originally the plan, but only KDE uses the new library so far. Abiword uses libwv1 and KDE uses libwv2. However, libwv2 is unfinished, so things may change in the future. OOo has always had its own code and I don't think they intend to change. Of course, I'm not an OOo developer.

NewsForge: Do you feel that KDE could be used for developing countries? Especially considering the internationalization team, do you feel that KDE could make inroads into areas where proprietary software is forbidding?

Staikos: I think it can and that it already does. KDE actually even has some contributors from developing nations. We have also seen momentum toward Linux deployment in these regions, too. Recently HP started shipping an interesting custom solution where a single PC drove four displays and keyboards, allowing four people to share a PC simultaneously. This was a Linux solution.

I think KDE is the best choice for the user interface, if for no other reason than for the great internationalization. It's much easier to learn the basics of computing when there is no language barrier in the way. KDE is also a very user-friendly operating environment, perfect for beginners but designed with power and flexibility in mind. This means it gives newcomers room to grow. They can learn the basics quickly and easily, then break out of the beginner box and experience a powerful computing experience.

NewsForge: Where do you feel that KDE has to improve to gain more new users?

Staikos: We need to be faster and smaller, that's for sure. We are constantly improving performance, and I think we are consequently gaining users. We also need to keep up with the new features in the other platforms -- especially Windows and Mac OS.

NewsForge: Where do you feel KDE has strong advantages over Windows or Mac OS?

Staikos: Network transparency and integration, for sure! Those other platforms have a long way to go to compare with what we have in KDE. To me, it's downright irritating to use a platform that doesn't allow access to all your network protocols from any place that files are referenced. The component model is great, too. Everything just integrates so cleanly in KDE -- if you have a KPart handler for your data, such as a spreadsheet, it can be easily opened or worked with from wherever you are on your desktop.

NewsForge: Thanks for your time, George.

Staikos: Thank you!

Click Here!