On Tue, 27 May 2008, Paul Smith wrote: > Tony Finch wrote: > > > > http://www-uxsup.csx.cam.ac.uk/~fanf2/hermes/doc/qsmtp/draft-fanf-smtp-rfc1845bis.html > > OK. Does anyone support that though?
Not as far as I know :-) > It does seem overly complicated. I'd be reluctant to put the effort in to > support that (unless it became very widely supported) For instance, I'd have > thought that having to resume partial DATA sessions is pretty much unnecessary > now. Servers don't have to implement suport for partial data retransmission. Part of the motivation behind the draft was mobile users who suffer from mobility-related dropouts, in which situation partial data retransmission would be helpful. I suppose it might be reasonable to allow clients to use TRANSOFF=0 instead of echoing the server's offset, which would make partial retransmission optional for clients too. Tony. -- f.anthony.n.finch <[EMAIL PROTECTED]> http://dotat.at/ SOUTH UTSIRE: EASTERLY 4 OR 5, OCCASIONALLY 6. SLIGHT OR MODERATE. FAIR. GOOD.
