Oops! I forgot that full and filtered cost maps return the same media-type. Sorry about that.
Yes, media-type==alto-costmap+json && accepts == "" => cost-type-names has one entry. media-type==alto-costmap+json && accepts == "alto-costmapfilter+json" => cost-type-names has one or more entries. Again, profuse apologies! - Wendy From: "Y. Richard Yang" <[email protected]> Date: Wed, July 24, 2013 10:37 To: Wendy Roome <[email protected]> Cc: "Y. Richard Yang" <[email protected]>, Richard Alimi <[email protected]>, alto <[email protected]> Subject: Re: [alto] Cost-type names Hi Wendy, On Wednesday, July 24, 2013, Wendy Roome wrote: > Fine with me! > > Although I'd go one small step farther and say that if an entry has media-type > alto-costmap+json, then it must not have an "accepts" entry. Only specifying the predicate: media-type==alto-costmap+json && no accepts => #cost-type-names has one entry is safe. Specifying media-type==alto-costmap+ json => no accepts will make it not possible to list filtered cost maps. Right? Hence, we limit to the first predicate. What do you think? > Similarly, a resource with media-type alto-networkmap+json cannot have > "accepts". > > Note that a server can provide full and filtered cost map services with the > same uri, via GET and POST respectively. The IRD just needs two different > entries, with the same uri but different media-types. Nothing says uris must > be unique. Yes. This is good comment and worthy of an explicit sentence to point it out. Thanks! Richard > > - Wendy Roome > > From: "Y. Richard Yang" <[email protected] <javascript:_e({}, 'cvml', > '[email protected]');> > > Date: Wed, July 24, 2013 10:12 > Subject: Re: [alto] Cost-type names > > Hi Wendy, > > So, the rule is: > > If media-type is alto-costmap+json, and no accepts (I.e., unfiltered map), > then the cost-type-names in its capabilities can include only one entry. I > believe that this is what you propose. If others do not see problems, we say > that we reach consensus. > > Richard
_______________________________________________ alto mailing list [email protected] https://www.ietf.org/mailman/listinfo/alto
