What do you feel about Ian Murdock's suggestion that there be a "benevolent dictator" in a project like Debian so that decisions are taken faster?

I think that model works with 50 people, but not 1000. However benevolent and dictatorial you might be, you cannot force people to do something they do not want to do, especially in Debian.

Maybe my opinion on this will change with time, but I currently don't think the DPL has his hands tied by the constitution not giving him enough power: I am allowed to get rid of all delegates and replace them with other people in a flash! A DPL who manages to solve our issues of knowledge retention, poor communication and concentration of powers will not need to be a dictator to be listened to.

What is your opinion of the GPLv3? Do you think Debian should think of moving towards where possible or do you favour staying with GPLv2 as is the case with most of the packages now?

Most GPLv2 software in Debian is already GPLv3-compatible due to the "GPL version 2 or later" chunk that the FSF suggests, or some may say lures (one) into, using in the code. And I cannot imagine what special, highly strategic issue could possibly make us want to make a piece of software GPLv3-only.

Debian does its best to respect licence terms and licence incompatibilities have always been a major pain. KDE wasn't in Debian for a long time until Qt eventually became GPL. We are still today dealing with GPL software that uses the OpenSSL library. Deliberately adding more incompatibilities by using the GPLv3 would be terribly unwise.
That said, developers still do what they wish. Debian does not own their code.

In the past there has an idea floated - I think it was shortly after Sarge was released - that some of the supported architectures be dropped in order that release times could be shortened. Your view?

You are probably referring to the Vancouver proposal document, issued shortly before Sarge was released. It was later used to fix rules for the inclusion of a given architecture in the Debian development process and in the Debian release. This is why Etch was released without m68k support, for instance, because it failed to fulfill some of the requirements, but the m68k port is still available in our developmentbranch (nicknamed "unstable" or "sid").

m68k and a few other architectures such as Alpha are not evolving as fast as the mainstream platforms in terms of hardware performance, which means our build infrastructure doesn't scale terribly well and might cause the removal of more architectures. There have been attempts at using emulators and cross-compilers to save these architecture, but they were met with mixed feelings. I personally think emulators and cross-compilers are an excellent solution. This is already what happens in most of the embedded platform development.

The more architectures we have, the closer to a "universal operating system" Debian becomes, and the happier I am. We have two ongoing ports, not even using the Linux kernel but the FreeBSD kernel, that are doing tremedous progress and will most certainly be part of Lenny, our next release.

Some people are already writing off Debian as a project whose time has passed and predicting that Fedora or OpenSuse will be the number one community Linux soon. What is your response to such soothsayers?

Rumours of Debian's death have been greatly exaggerated. Seriously, all these projects are young. Of course, they look shiny and new, dynamic, have healthy structures behind them, but aren't they all like that at the beginning? Debian is still bigger in terms of people, software, ports, derivative projects, QA efforts. I remember when Gentoo was the next big thing; it's still doing good, but it did not exactly bury Debian. Let's see how Fedora and OpenSuse are doing hen they're 8 years old.

A question that every Debian leader is asked nowadays - how do you plan to improve relations with the Ubuntu project?

I have no specific plans about that yet. I don't even know what really needs to be improved, but I feel that our biggest hope is to make Ubuntu developers want to contribute to Debian because they have some interest in doing so, such as being more productive.

For instance, a few Ubuntu developers and packaging teams are already using our Alioth collaborative development platform, making it easy to share patches between Ubuntu and Debian versions of the same package.

Would you consider standing for a second term as leader only if you made good running this time? Or would it happen anyway?

I have no such plans. As I explained it when I was elected, I cared far more about my platform than the guy who would try to apply it, and I hope people will run with even better platforms next year. And anyway, if some of my projects are not finished by then, I don't think I'll need a position of power to carry on with them. 

{moscomment}

---

Get stories like this delivered daily - FREE - subscribe now
When you subscribe get a 12 months license for LiveProject
Valued at $99 USD


Download the FREE iTWire desktop alert widget

<< Start < Prev 1 2 3 Next > End >>