No. They're free to pull the plug on whatever. I will just make all future decisions knowing that this will probably happen. i.e. I won't rely on anything Google releases unless I know it makes them money.

Google can do what they like -- but at this point, without such a guarantee, their history shows that it would be unwise to take the time to integrate their side projects into one's life or workflow.

> Should Google only work on things that it could guarantee to run for at least a few years no matter how it is performing?

Of course not, but it's gradually becoming self-fulfilling.

Early adopters in the 'tech community' are becoming wary of investing time in new Google experiments, which in turn affects promotion of the service to the mainstream community.

What might help is if they moved this sort of product back into the domain of Google Labs and gave it a minimum lifetime indication. I think people would be happier to give things a try if they knew it was experimental and might only last for 12 months.