I've kind of compromise idea for the "licensing issue". While we wait for
Microsoft to respond on what exactly does their patent cover, I've
strong feeling that the only thing it might potentially cover is algorithm
to find which is the address to use for PRA from other mail headers. But
we also have draft that defines SUBMITTER for email envelope parameter.
So my proposal is to rewrite some core documents and separate them as follows:
1. Core document that defines what MARID is and says that when "PRA"
address is to be used, it is taken from SUBMITTER parameter Mail From. If
abcent "other" algorithms maybe used to find what submitter is. This
document should be free from any patent issues.
2. Document that defines SUBMITTER extension to Mail-From. Same as now and
it also should be free from patent issues. I
3. Document that defines "caller id" algorithm for determining PRA address
if SUBMITTER was not present This may have patent issues.
Now as far as last document we can say that SMTP vendors are not required
to implement it for MARID and if they did not they may use other PRA
determinining algorithms or may simply not check PRA is submitter is not
present. Additionally as mentioned we we can work futher to come up with
"good enough" altorithm for determining PRA that is not covered by any
patents. And even if we don't come up with such algorithm, the patent
issue with current algorithm would be interim anyway until MTA are
upgraded to use SUBMITTER and as such it does not cover MARID work long
term and the core remains patent free.
I do realize that separating PRA algorithm into completely separate
document and making sure that its the only one that has patent claims
means the current drafts will all have to be rewritten and this means
additionally delays, but this small delay is worth it.
--
William Leibzon
Elan Networks
william(_at_)elan(_dot_)net