Boasting a revamped development and release structure, an upcoming instalment of its desktop environment in March, and a raft of new applications for future releases, the GNOME project is pushing to ensure free and open source software earns its place in the global desktop space.
According to GNOME release team head Jeff Waugh, who gave a presentation at the Linux.conf.au 2004 conference last week, the GNOME project has undergone a major facelift over the past 12 months. This has included switching to a time-based release schedule, starting with its 2.x series, as well as redefining the project's social structure into module maintainers, a release team, and an administrative and advisory foundation board.
As a result of its more 'cut and dried' structure, the release team for GNOME is also on track to issue the next instalment of GNOME version 2.6 in March this year, Waugh said. The release, now in beta testing, will include several new features, such as the introduction of the latest graphical user interface toolkit release GTK 2.4, as well as ongoing compatibility with API (Application Programming Interface) and ABI (Application Binary Interface). GNOME 2.6 will also support Evolution Dataserver version 2.0, which provides real-time, read-write access to calendar and contacts data from the desktop.
The project's development team is also working on the prototypes of value-added packages for GNOME version 2.8 and beyond. These range from application plug-ins to a collaboration station for its desktop stack, allowing users to interact with their network and database solutions through a variety of interchangeable interfaces off the desktop.
LinuxWorld's Nadia Cameron caught up with Waugh after his presentation to ask about the team behind the GNOME project, working with the big vendors, dealing with proprietary software patents and improving the open source desktop stack for global take-up.
LinuxWorld: Looking at the patent side of things -- working on a desktop system, in particular -- do you find that software patents are a big issue?
Waugh: Yes, it's something that we have hit our heads on a few times now. There's a couple of things in our file manager that we'd really like to do, and that our users would like to see, but we can't because Apple has patents on software in their finder.
The whole point of the patent system is that they're supposed to be obvious things. But there are a lot of things in computing that are unobvious to a point… (breaks off)
Also the mono project, which is a real implementation of Microsoft's .NET. A lot of people are excited about it but are also scared because Microsoft has some really big patents which cover big chunks of code - even the stuff that MS has standardised. So no one is really sure how they're going to use those patents or whether they [MS] would use them defensively.
LinuxWorld: Sun is actually using GNOME on its Java desktop systems. Does this mean you have a lot of interaction with Sun? Do you get a lot of feedback from them?
Waugh: Heaps. With Novell, Ximian, Red Hat, Sun - all of their developers are working with us. The (GNOME) foundation also has an interface to their corporate guys who are working with planning, etc . Every day we are working with Sun. Sun have actually been really good with the interaction between their own users and the rest of the (GNOME) project. They did a big usability study on the previous GNOME version (2.4) and they came back not only with the results of the study but with a bunch of comments from their users. They recently did the same thing on the beta testers of GNOME 2.6.
It's useful for us because the kind of users on our mailing lists are that one per cent of very technical users who get involved in open source stuff. When we're trying to deliver software that is going to appeal to the 99 per cent of people who just don't care and just want it to work, getting feedback directly from systems administrators who need to administer our software through Sun really helps.
LinuxWorld: Just on the usability front: I noticed there was a recent conference about human interaction with computers (Human Computer Interaction conference held in Queensland in November 2003). It seems to be something that's coming more and more into the GNOME releases. Could you tell us a bit more about what some of these guidelines are?
Waugh: One of the biggest things is that we have a human interface guideline, which is the only document of its type in the open source world. It basically goes through the technicalities of implementing the guidelines, creating software that people with disabilities can handle, why we do things, and so on. It makes it really easy for the developers to understand what the usability people are talking about and what some of the outcomes of usability are when talking to other companies and users. These issues have really come into GNOME since 2001 - the time we were starting the 2.x process.
Then we've had the use of some of the usability studies from Sun, plus we've got people involved in the project from the companies who distribute GNOME, as well as independent studies from people who are doing research work at university. So we've got a large range of usability people to speak to: some in research, some doing it practically.
LinuxWorld: Getting back to the involvement of the bigger vendors, which seems to be a fairly recent phenomenon. Does it change the way you actually develop this software? Does it improve it, or does it make it harder?
Waugh: The biggest change is that we've now got people who are working on it on paid time, which is 40 hours or more, compared to the 10 to 20 hours a week you could put in when you were a volunteer or a student. That has upped the pace significantly. It has improved the quality of the software because they [vendors] have to deliver a product. A lot of the software development is now done under quality and assurance, and so on.
Some of it has caused problems in the past between the companies and people working on it. But the way that we've built the GNOME social structure now works. We've split off the technical but made sure that everyone works together within the defined social structure.
LinuxWorld: The way you guys have the GNOME social structure seems quite unusual. In one of the other presentations here [at the Linux.conf.au 2004 conference] the speaker was talking about the lack of roadmaps on the open source side - but obviously what you guys are doing is clearly defined.
Waugh: Well, it depends on how you define roadmap (Waugh laughs). We used to have this thing where everyone would say "I want my feature blah in the next release". And we'd sit there, and people would hack on their features, and finally there would be a new release. [In other words], the release was based on what needs to go in.
Our current roadmap is really defined by knowing the time of release. We know that there's going to be a release in two months time, so this is where we start working on defining the actual features for the 2.6 release and go back into testing.
The roadmap in terms of time is really straightforward. Every six months we do a release. In terms of features, it basically comes down to what people are currently working on, what they are committed to doing for each release and making sure they keep their deadlines for each version.
I hope this has influenced some of the other open source companies. Sun has recently switched to a quarterly release cycle, so all of their products have specific dates for release. Obviously some of that is marketing, but I think our way of handling things on a time release has affected them, too. A lot of what we do fits in with their teams and their systems.
LinuxWorld: But GNOME still takes contributions from people sitting at home, thinking "well, this is a good feature, it should be included". How do you lock off from that with the releases?
Waugh: That's the release team's job: to be incredibly ruthless. We have a point where we put a feature freeze on new modules and applications. Once we've reached that point, nothing will get added at all. Currently we set this about two and a half months prior to release date.
Now that everyone has bought into the time-based release, contributors will say "well, I've written this thing and I'm going to put it forward for the next release of GNOME in six month's time". And the good thing about our releases is that you only have to wait a maximum of six months to get something in. The speed and the satisfaction of getting software into GNOME is not a long wait.
LinuxWorld: So, a last question: where is GNOME at now? The room [for your presentation] was full, so GNOME is obviously popular - what's the next step in terms of getting it adopted by more people?
Waugh: Supporting our distributors. We've always felt our distributors are really important to us. We're not going to be getting the product out there, we're just writing the software. We've got huge deployments of GNOME in China and the UK, we've got deployments in Brazil, Spain, Peru and all these other places. The foundation is looking to support a lot of those places around the world, and tell everyone about GNOME.
In terms of the technology, we've basically got all of the desktop applications solved. Between OpenOffice.org, GNOME, Mozilla and a number of other projects, the stack of stuff people generally use on the desktop is pretty much there.
That's what I was talking about in my presentation. There are now these crazy ideas starting to come out for features for the Linux desktop because we've solved the big problems. People are embracing [open source desktop environments] as real solutions now.
- The linux.conf.au 2004 conference was held in Adelaide between 12-17 January 2004.