|
Two bits of open source news and (not so) witty commentary |
|
|
|
Wednesday, 23 July 2003 |
Mozilla Firebird
This release harkens back to the compiler slinging days of the browser wars. This time around, I think it's going to be one of those wars where nobody shows up. Even if Mozilla is better on the margin, most people just don't care. IE is already there and that's what they're going to use. Mozilla may gain market share, but this is a skirmish, not a war. It's just like linux on the desktop: people are not going to bother to install new operating systems over ones they already have. It's hard enough to get people to upgrade Windows or IE, let alone replace them with something else.
The only hope Mozilla has is bundling (not tying - a subtle but important legal difference, one is a violation of anti-trust laws and one isn't). Maybe it could come with Windows if there's a court order, or more realistically, AOL. Master Blogger Joel was right on with this one.
Mozilla firebird is now my default browser (click here to get yours). The fact that the most exciting features are so subtle shows just how low expectations are. Don't get me wrong, I really love having a non-bloated browser and tabbed browsing is great. But shouldn't I already have that? Let's face it, IE hasn't changed very much since version 4 when the browser war wrapped up with Microsoft the victor.
I am also encouraged by Firebird because it shows that somebody at Mozilla finally realized features are more important than re-writing old code.
The RedHat Project
RedHat restructured the way they create their distribution of linux. Various open source projects have different models for control and contribution, I think RedHat has found the sweet-spot for a large scale linux distro.
BSD is ruled by commiters -- maximum contribution but less coordination and control. The linux kernel is ruled by Linus and his round table -- maximum control but less contribution and innovation. RedHat used to follow the linux kernel model. They took what they wanted and built their distro almost entirely using the labor of their own developers.
Under the new system, RedHat developers are like the editors of a magazine with the current draft of the latest issue always available for everyone to see. Periodically, RedHat skims off the latest stable build, makes some editorial deletions and additions, then puts it on newstands as a shrink wrapped and supported product.
This is a great model. They get more innovation and contribution (some of it from outside RedHat which saves them money) but can still exert control over the supported builds. I haven't seen very much press about this change, but I think could signal a significant change in business models many companies using open source software follow.
In the short term, the number and quality of packages included in RedHat may suffer (they already cut some mostly for licensing reasons, see the Package Changes section of the Release Notes). But in the long run tipping the balance toward innovation will lead to more and better features in their supported releases and greater flexibility in the unsupported version.
If Apple ever decides to commercialize OSX on Intel hardware this would be a great lead to follow.
Ahoy,
Jason
Powered by AkoComment 2.0! and SecurityImage 3.0.4 |
