Wendy, how would you encode a cost map where most of the entries are "I don't know", some are "you should try these" and some are "better avoid this PID"? Reducing the number of PIDs so that I have only one (or very few) "I don't know"-PIDs is not an option because I am planning other cost maps with many details, too.
Thanks Sebastian On Mon, Mar 25, 2013 at 04:22:21PM -0400, Wendy Roome wrote: > I don't think the "default cost" idea is worth it. > > The only advantage I see is sending fewer bytes over the wire when a lot > of the costs are exactly the same. That strikes me as very unlikely. > Outside of the artificial cost maps we invented for our interop tests, of > course! > > And if a server does use a default cost, then the server cannot indicate > "unknown cost" for a src/dest pair. > > - Wendy Roome > > >Date: Mon, 25 Mar 2013 10:53:49 +0100 > >From: Sebastian Kiesel <[email protected]> > >Subject: Re: [alto] Cost map: undefined vs. default value > > > >Ben, all, > > > >On Mon, Mar 25, 2013 at 09:06:18AM +0000, Ben Niven-Jenkins wrote: > >>Sebastian, > >>On 25 Mar 2013, at 08:31, Sebastian Kiesel wrote: > >>> Dear all, > >>> > >>> draft-ietf-alto-protocol-14, section 9.1.2.5. says: "An ALTO Server MAY > >>> omit entries for which a Path Cost is not defined (e.g., both the > >>>Source > >>> and Destination PIDs contain addresses outside of the Network > >>>Provider's > >>> administrative domain)." > >>> > >>> I think it would be beneficial to have an (optional) method of > >>>specifying > >>> a default value in the header of the map. > >>Are you suggesting the default cost replace the unknown cost semantic > >>or that we keep the unknown cost semantic and add an additional > >>default cost semantic? > > > >Both options would work for me as I personally do not see much use of > >the "undefined" semantic. On the other hand, adding the "cost-default" > >as an optional feature seems less disruptive wrt. existing > >implementations and then we could support both the "undefined" and the > >"default" semantics, so I think the proposal is: add it as an option, > >i.e., if "cost-default" is present all omitted entries in the cost map > >are assigned this value, otherwise they are undefined. > > > >Thanks > >Sebastian > > > _______________________________________________ > alto mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/alto _______________________________________________ alto mailing list [email protected] https://www.ietf.org/mailman/listinfo/alto
