Linked by Eugenia Loli on Wed 23rd Nov 2005 18:54 UTC
GTK+ The GIMP toolkit (GTK+) matures to include yet another platform. This project aims to bring native GTK+ support to the Apple Mac OS X platform. The basic implementation is sponsored by Imendio AB and the work so far has been done by Anders Carlsson (of Imendio).
Permalink for comment 64535
To read all comments associated with this story, please click here.
RE[5]: a few points
by segedunum on Fri 25th Nov 2005 14:32 UTC in reply to "RE[4]: a few points"
Member since:

Let's make this more positive. What do you believe works in the open source world and which business models work?

Thank you very much.

It really depends on the kind of software - and that's a large part of the problem. Will we always have LGPL software, BSD and GPL? Yes, of course we will and we need them. However, the notion that some people have of a BSD/LGPL/develop-for-nothing panacea is just wrong.

It depends on the complexity and type of the software involved. The desktop, for example, is a very complex area because it depends on the development tools and desktop development infrastructure, which takes a great deal of time and effort to get up to the required quality and for it to continue to be maintained. You've then got to develop applications on top of that development infrastructure that are of good enough quality for people to use, which is complex in itself, and then there are other peripheral issues like usability. Open source software by and large has succeeded in quite a few areas, but it's what will take us over the tipping point in terms of quality and useful software that will replace much of the proprietary stuff we have (but won't wipe out proprietary software). Remember that the open source desktop is user-facing which server software, the place where open source software whatever the license has always been successful, is not.

However, look at the Linux kernel. It would never have reached the quality it has without people pushing more code into it, which the GPL compels developers to do. If people were allowed to put proprietary extensions on that's exactly what most companies would do and you would never have the number of open source drivers you do now. Actual kernel development would be an absolute fraction of what it is.

In the case of Red Hat I think they have it right for Java because there is a market and a very large community of existing software and support. You have IBM and a multitude of others putting huge developer resources into stuff like Eclipse and SWT. What they're doing with gcj and classpath is just about doable because Red Hat have paying customers whatever licenses the software that they use have. However, there's no getting away from the fact that this is a huge amount of effort and it remains to be seen whether it is doable from a desktop point of view. Java server software is already well established there though, and this is really what Red Hat is aiming at - at least first.

In short, the past few years of the open source desktop's failure to really go places should have taught us something about the nature of the software required there.

Edited 2005-11-25 14:36

Reply Parent Score: 1