On Sat, 30 Oct 2004, guy wrote:
What about this?
When using sc, and scope uses "+" or "-" then the current set of scope(s)
will be modified. If "+" or "-" is not used then a new set of scopes is
defined. If a "-" is used then any other scope without "+" or "-" will
assume "+" (or we could say: if any 1 scope used "+" or "-" all other scopes
must also use "+" or "-", if not, syntax error). The first set of scoped
defined can not use "+" or "-" since the set is not yet defined (or we could
just assume the default is an empty set).
Examples:
"sc=s1 m1 sc=s2 m2 m3 m4"
m1 is in s1 scope
m2, m3, m4 is in s2 scope
"sc=s1 m1 sc=-s1,s2 m2 m3 m4" ##s2 is assumed to be +s2
m1 is in s1 scope
m2, m3, m4 is in s2 scope
"sc=s1 m1 sc=+s2 m2 m3 m4"
m1 is in s1 scope
m2, m3, m4 is in s1 and s2 scopes
"sc=s1,s2 m1 sc=-s2 m2 m3 m4"
m1 is in s1 and s2 scopes
m2, m3, m4 is in s1 scope
This looks good to me. I believe it gives more expressive scoping syntax
and is syntaticallly possible to define. The only thing is that I don't
like if we we have to foce somebody to have different meaning depending on
if there was - or not. So I propose it that if there was no +/- that it
means that all old scopes are closed and new one is open. But if there is
+/- than it has meaning of modifiying existing set. So the only modified
example from above is:
"sc=s1 m1 sc=-s1,+s2 m2 m3 m4"
m1 is in s1 scope
m2, m3, m4 is in s2 scope
And this really is like a scoping parameter allowing for opting in and
opting out of various meanings. Also if somebody wants to have v=spf1
stand for all scopes (i.e. like what Meng and Microsoft want), then the
following will specifically say "no pra":
v=spf? sc=-p .... -all
---
William Leibzon
Elan Networks
william(_at_)elan(_dot_)net