I agree with vicmarcal and jgwright. This Plan9 project looks more as an exception than a rule. Just looking at the (random?) 5 first projects on the gsoc2009 list gives:vicmarcal wrote:@JPLR:
"Yes there are counter examples (I mentioned it in my post) but less than you might think"
- Improving AbiWord’s OpenXML Support
AbiWord with GetText; AbiText?
Speeding up AbiWord's table layout algorithm
Port AbiWord for Windows to Unicode
Writing /proc and /dev File Systems
I also agree that ROS is not a good PR. On one hand it is not mature enough to generate strong positive "public" response, but "controversial" enough to start bad hype a.k.a. "Google is sponsoring Windows replacement".
Which makes me think about the idea expressed in another topic about spinning off sub-projects for ROS components. Surly things like UniATA, USB or ACPI are interesting learning material, especially if they work for all NT-based systems.