I'd like to have a discussion of Wave at a higher level.  I'd like for a
moment to ignore the details of what Wave is and what it does.  To have
a high-level discussion, we only need to know that Wave will be
implemented with software and will be a network service.  Those two are
trivially true, so we can figure out a few things about how to approach
it before getting bogged down in details.

First, I think Wave is important primarily because anything that Google
does will almost surely become very popular quickly simply due to their
marketing power.  It doesn't matter what it does; it's going to be
popular.

We also have to assume that Google will try to keep as much as they can
focused around lock-in with their services, and that they'll release
relatively FaiF clients and keep most server-side stuff proprietary.
This would be what their pattern of behavior seems to predict.

Like Android, they will probably release the FaiF parts late enough in
the process that it will be difficult for community-driven development
to pick up the code and catch up.  Only in recent months are people
finally putting together freedom-respecting Android-based distributions,
and the next generation of the product is already due out.

I think the benefit we have with Wave is that all that is out there is a
protocol.  I think it's worthwhile for us to give a serious go at
implementing to protocol in a netservice-FaiF way, and advocating for
changes to the protocol when it seems to block a federated, autonomous
implementation.

I don't know how we're going to get the time to code this, but I have a
feeling if we make a start, we might get others interested, due to the
buzz around Wave.

I think this is a top priority for people who support FaiF netservices.
After all, it's always easier to find violations of the Franklin Street
Statement when you are in the trenches implementing and testing a
protocol implementation.  Plus, if we make AGPLv3'd implementations of
the protocol early, we have a chance of competing with Google's
proprietary implementations.

What does everyone else think on these points?
-- 

   -- bkuhn
_______________________________________________
Discuss mailing list
[email protected]
http://lists.autonomo.us/mailman/listinfo/discuss

Reply via email to