- From: Henrik Frystyk Nielsen <henrikn@microsoft.com>
- Date: Fri, 16 Nov 2001 06:29:04 -0800
- To: "Martin Gudgin" <marting@develop.com>, <Noah_Mendelsohn@lotus.com>, "Doug Davis" <dug@us.ibm.com>
- Cc: "Mark Baker" <distobj@acm.org>, <skw@hplb.hpl.hp.com>, <xml-dist-app@w3.org>
Only as a mandatory extension and only by effectively redeploying *all* existing SOAP nodes. Without a targeting mechanism, it furthermore becomes very difficult to not just have passive forwarders. This all has a *very* high cost and will be fragile as a result. We had this discussion a long time ago - I would strongly recommend focusing on solving the outstanding issues of which there are plenty rather than going back in circles. Henrik Frystyk Nielsen mailto:henrikn@microsoft.com >+1, in fact, +asmanyasIcangetawaywith > >I would be heavily in favour of removing actor from Part 1 and >defining an extension for whatever 'actorlike' features we >decide we need. I would also strongly suggest that we define >said extension outside of our current path to REC. Many of the >current issues we have to resolve are to do with actor and/or >intermediaries. Pushing this out into an extension that is >defined post 1.2 seems like an ideal way of keeping ( getting >us back? ) on schedule. > >I've already done a *lot* of thinking about actor and I'm >fairly convinced that it *can* be done as an extension. >Hopefully I'll get chance to post some of my thoughts/designs >in the next week.
Received on Friday, 16 November 2001 09:30:03 UTC