"Today telcos behave much like time sharing computing in the 60s" (bill.st.arnaud)

for grid computing you need a network. a fast network. therefore the canarie project is trying to push something like: purchase networks like computers. that means: buy your own wavelength.

it's good for telcos, arnaud says: they finally get some money back for their digging, and finally more than 1% (some say 2) of the fibre network would get lit up and be in use. somehow they are promoting the line: let the customer/private community solve the problem with the last mile, which can be an opportunity. it is good for those who want to force the development of bandwith intens applications: since years movie is the only one out there. and as people want to push grid computing for research and teaching, well you wont get it if schools, libraries and universities have to pay for time and traffic.

"Extending the Internet model of autonomous peering networks to the telecom world"

the catchline for politicians and moneylender in the outline of the project canarie sounds as always really pathetical:

Canada could be poised to be a world leader in the 'optical age'. Silicon Valley was the capital of the 'silicon age'; Britain was the capital of the industrial age. (community condo fiber networks)

still, everything sounds a little bit more advanced as the "access for all" period some years ago.