[Tinyos-alliance] TO DO and meeting notes

[Philip Levis]

We discussed TEP 120 and agreed that accelerating its process is  
important. I attached 120 with the previous message advertisting the  
meeting. Everyone in the WG should read TEP 120 and provide comments  
by the END OF SUNDAY, JUNE 17th. I will take the comments and produce  
a new version of the document by the next meeting, at which point we  
will assign a shepherd.

Meeting notes follow.

Phil


-------------- next part --------------
Adam, Jack, Matt, Phil

TEP 120: 

Adam: We should give ourselves a short obligatory deadline and complete it.

Matt: I feel the same way. The question will become "Which document do we 
circulate about the structure of the alliance to draw in potential members."
But this isn't really written as a short statement. It might not be the
first thing to see. Once this is ratified, someone might want to write a
one page version of this. For example, all of these comparisons to ZigBee,
etc., it's confusing. It's not with a clear mission statement up front.

Phil: There is the 2-pager, but it might not be that obvious.

Adam: I agree that TEP 120 is not a good first glance. But the 2-pager can.
What we want for 120 is the complete story. What we want to make sure is that
there isn't anything we don't want set down in stone.

Matt: I just noticed that the TinyOS Alliance does not have a web presence.
I think that this is part of the documentation working group. What I would
propose is the following. There are 2 things docs wants to do: technical
documentation, mailing lists, etc. The second is to give a web presence
for the administrative structure of the Alliance. For example, to give
the Alliance a way to get information out. Policies, etc.

Phil: Would it be useful to separate TEPs and community effort?

Matt: Let's not make things too disjoint.

Adam: In addition to that, there exist links.

Matt: Going back to Siemens, you've been talking with them about being
a member.

Adam: They've said, we've put it all through the green lights, and now
we have it, and we're positive we'll join, but we want the table of fees.
Also, they need to spend the money this fiscal year. Once we are in,
it's easy to get the continuation. So they are ready to join.

Matt: So TEP 120 has a proposed fee schedule.

Adam: I wanted to understand that we don't think about changing this part.
So I can copy and paste and send to them this part. Then the second
question is then, what is our estimation of the schedule such that they
know where to send the money, you see!

Matt: My guess is that this is David's plate. He has the non-profit
formation, etc. SIEMENS could be a good forcing function.

Adam: I'll talk with David, because this could be a great way to get
things started.

Matt: It might be nice to make them a Charter Member.

Phil: Testing.

Phil: How do we encourage people to put codes in a central place.

Adam: I would like to be one place more precise. Do we need them to be
put in a central place, or linked to from a central place? I am a big
fan of links.

Matt: There are a bunch of issues. One of the reasons we don't check things
in en masse to tinyos-contrib is because we don't think it's the right way
to get visibility. I'd much rather that people search with Google and 
find a web page describing what we do, rather than find some code. We
could probably export everything into contrib. It's easier to just
put up a tarball. There's an activation energy. It's not clear what the
value is. I don't think there's much of an incentive.

Adam: Good point. What is the additional incentive. 

Matt: It's not a bad thing. It's just, is it worth the extra effort.
Now, that is a problem. There's an active development community. There's
a bit of an effort if you want to maintain a separate development tree.
We don't like releasing code that isn't ready.

Adam: That's fair -- you don't always want to release everything.

Matt: Here's another example. With our Volcano code, we made tons of tiny
hacks here and there. We copied over everything that we wanted to freeze.
What ends up happening is a enormous duplication of code.