Some optimistics thoughts about our chances to use Brazil one day:

The Plan 9 user community is small; it lacks coordination. So, things
are moving in Plan 9, but slowly, and in a non-coordinated way that
requires a not small energy to keep your own system in sync with what
is done at many sites by a dozen of persons.  Conclusion: you won't
get very far with Plan 9, it's too hard for a newby to follow two or
three years of 9fans archive to (for example) choose a video card or
compile ssh.

When someone on the mailing list wrote that the Inferno source license
cost 1M$, my opinion was that inferno was not something for me. What's
more, I was convinced that the (supposed) marketing approach of Lucent
(make it something expensive, used to solve hard problems; don't try
to spread it), implied that Brazil would never get out of the Bell
Labs.  I stopped to promote Plan 9 around me, only using it for myself
and marginally using it to convince the students that some things they
thought difficult are actually badly formulated problems.

I was reinforced in the opinion that we would never use Brazil by what
happens on the 9fans mailing list: every six month, someone asks about
the probability of seing Brazil, and doesn't get an answer (at least,
no answer on the mailing list).

My mind changed this summer, when I discovered Lucent's Inferno
university partnership program. I hope that it is a sign that someone
at Lucent realized that Inferno would be more strong on the marketing
side if it disposed of a strong user community.

If I am right, regarding the preceding point, I think we can have some
hope to use Brazil. Someday, someone will realize that distributing
Brazil will only boost Inferno sales. The OS market being what it is
(i.e. microsoft vs free unixes), it won't be too expensive for me.