At 3:06 PM -0400 10/23/07, Sam Hartman wrote:
Let me suggest starting with a lesser goal. Try to build a consensus
that unless there is a good reason to do otherwise, it needs to be
possible to write an open-source implementation of a standard and that
the absence of such an implementation should be considered a red flag
when advancing beyond proposed.
I think you have to be careful here, as "open-source" covers a variety
of licenses. Having a diverse set of implementations is clearly
a good sign that a standard's specification is clear enough to implement
and useful enough that folks have chosen to spend the time. Those
ought to be critical aspects of our thinking when we look at how
to revive the standards track's upper reaches. But reviving it
will get more difficult, in my opinion, if we set tests like "must
show at least one implementation subject to the GPL", as that
presumes which implementation groups are interested,
or delays forward progress until a group that does not work
in that mode produces an example implementation that meets
the test. Even if this is an informal requirement (lore vs. spec.),
this could discourage those working for advancement.
Basically I'd like to start by getting to a point where we assume that
open-source implementations are a goal and that we explicitly decide
that they are not a requirement in contexts where that makes sense.
I suspect we would run into resistance building that consensus but it
might be worth trying.
I'm a little confused as to the antecedent of "we" in the statement above.
I assume you mean you and Simon, but that you are basically speaking for
yourself. If you mean "we" in some other sense (especially if you mean it to
include the IESG, which some might infer from your role), it is not clear.
Speaking only for myself,
Ted
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf