The Open Source Business Resource is an academically sponsored body of work published each month out of Carleton University in Ottawa, Canada. I've an article in this month's issue. Regular readers will recognize a lot of themes they've read here or heard me discuss and present. The focus is on the coming mobile Internet and open source software. I would encourage people that want to comment to hold the discussion on the article's website.
The Arrival of the Mobile Internet Thanks to the Economics of Open Source Software
Walli, S. 2009 Jan 5. The Arrival of the Mobile Internet Thanks to the Economics of Open Source Software. Open Source Business Resource [Online] 0:0. Available: http://www.osbr.ca/ojs/index.php/osbr/article/view/818/790
Nice article which is a good general summary. I would slightly disagree with one aspect; I don't think handset manufacturers forked the Linux kernel (and pieces higher in the stack) because those projects weren't sensitive to their critical issues such as battery life, memory overhead, etc. The real reason IMHO is that the handset industry is so proprietary, reactionary, and paranoid that it has struggled to come to terms with standard practice in the various communities - they are attracted by the lack of royalties, but haven't got beyond that to participate directly and collaborate in the upstream communities.
Even simple things like using parts that have upstream drivers (or even open specs) are still not common place, e.g. why no 3rd party alternate distros for the 770 or n8x0 devices? Because important parts of the devices don't have open drivers - and Nokia are a shining example of the correct way to work with upstream communities (primarily the kernel and GNOME Mobile), so if they are struggling with component price pressure, and general lack of choice, then you know less scrupulous OEMs will just fork the kernel and live with the maintenance pain.
In that light, I see things like LiMO as the handset industry (and carrier industry to a certain extent) trying to make Linux and OSS more like what they're used to - a black box solution that they get to build apps and services ontop of, with no per device royalty. Also one that sits still for a bit and doesn't answer back :-)
Similarly I think part of the attraction of Android for many handset makers is that Google takes away all that horrible OSS chaos, isolates all that scary GPL stuff to the kernel and a few other lower level middleware bits, and presents a nice APL face to the app developer, with nice IDE and tools. All the bits they're really interesting in are in the APL layers so they get to fork without the tedious requirement of releasing their code (let alone work upstream).
I think the danger for handset manufacturers is the same with Android and Windows Mobile, where the value moves from the commodity hardware to the OS, except now the OS is commodity and the value moves again to the adjacent (ad supported Google) services - Nokia clearly think this way, with the development of the Ovi brand of services, and acquisition of Navteq.
Posted by: Paul Cooper | 09 January 2009 at 07:56
Hi Paul: Thanks for excellent commentary. I agree the handset manufacturers have lived in a historical world of hardware secrecy making it difficult to transparently share. I certainly agree with your comments on LiMo. Travel the web site -- there's absolutely no obvious way to participate, learn, contribute. It will be fascinating to watch this unfold over the coming year.
Posted by: Stephen Walli | 09 January 2009 at 16:25