RE: Proposal for new last call issue: Some unprocessed headers should stay

Noah, others,

my position that what you are trying to do is doable already in modules
themselves was based on my assumption that the default is that
unprocessed headers stay in the message. I don't understand how I got
that impression.

Seeing the proposal with the true default in mind I see that something
really needs to be changed in order for us to cover the presented
scenario.

In this situation, I would prefer changing the default to keep
unprocessed header entries, and deferring the relaying into modules
where I think it really belongs; but I see how this might be perceived
as a big change.

Targeting a header at a role other than 'next' is used to bypass the
nodes that *mustn't* process the header; this is not doable with the
special role. I am not at all sure that we can accept this trade-off -
either you can bypass some nodes or you can make a header sticky, not
both.

As for the proposed attribute, I agree with Henrik regarding the corner
cases.

Sorry if my mistaking the default behavior caused any confusion.


                   Jacek Kopecky

                   Senior Architect, Systinet Corporation
                   http://www.systinet.com/

Received on Wednesday, 16 October 2002 13:32:08 UTC