[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [xri] XRDS media type
In the new XRD world HTTP clients should be able to do "Normal" resolution on a proxy server and get back a XRD that they can digest. The existing XRI 2.0 behavior needs to remain unchanged. The only new thing would be asking the proxy for a XRDS and that doesn't require a mime type. John B. On 2009-09-16, at 4:54 PM, Drummond Reed wrote: > So John, you are thinking with XRI Resolution 3.0 we can get by with > just > registering the application/xrd+xml media type? > > I'm cool with that if we can. I figured we had to deal with the > versioning > issue for both XRD and XRDS no matter what, because XRI Resolution > 2.0 had > both mime types (and both mime type applications) as an appendix. > > But, as Eran and Will and I discussed yesterday, the XRI TC never > followed > through with the actual registration because we erroneously believed > that we > needed to complete OASIS Standard voting first. > > Let's not make that mistake again - I think we should register as > soon as > we've had the Committee Draft vote, and in any case no later than the > Committee Specification vote. > > =Drummond > >> -----Original Message----- >> From: John Bradley [mailto:jbradley@mac.com] >> Sent: Wednesday, September 16, 2009 1:34 PM >> To: Eran Hammer-Lahav >> Cc: Drummond Reed; 'XRI TC' >> Subject: Re: [xri] XRDS media type >> >> For the proxy resolver the mime types are passed as a query string, >> and not by content negotiation. >> >> I think for the registration we can concentrate on the current use. >> >> We have to deal with versioning on the proxy server anyway. >> >> John B. >> On 2009-09-16, at 3:42 PM, Eran Hammer-Lahav wrote: >> >>> You can always use something else other than Accept header to make >>> this query different. >>> >>> But the issue here is about registration. If we register both, we >>> need to deal with the existing (and incompatible) use of >>> application/ >>> xrds+xml. We need to list all the existing applications using it, >>> detail interoperability issues, etc. This is not impossible but it >>> is not fun either. The current deployed expectation for application/ >>> xrds+xml is to get an XRDS (2.0) document. We will need to address >>> it. >>> >>> Since we are using the same namespace, it would be perfectly legal >>> to return an XRDS (XRD 1.0) document for the xrd+xml mime type. My >>> question is if that is enough? If the only use case is the one >>> provided below, I think the right solution is to change the resolver >>> interface and not register the xrds+xml type. >>> >>> I don't feel strongly about this. >>> >>> EHL >>> >>>> -----Original Message----- >>>> From: Drummond Reed [mailto:drummond.reed@cordance.net] >>>> Sent: Wednesday, September 16, 2009 12:35 PM >>>> To: 'XRI TC' >>>> Cc: 'John Bradley' >>>> Subject: RE: [xri] XRDS media type >>>> >>>> I missed the start of this thread, but here's the specific use >>>> case I >>>> can >>>> speak to for XRI Resolution 3.0: >>>> >>>> In some cases an XRD consumer (acting as an XRI 3.0 resolver) will >>>> want >>>> to >>>> request JUST an XRD from an XRD provider (acting as an XRI 3.0 >>>> authority). >>>> That means that no matter how many XRI 3.0 resolution steps (linked >>>> XRDs) >>>> the provider needs to request, the consumer only wants the final >>>> XRD >>>> back. >>>> >>>> In this case the XRD consumer can make a request asking for >>>> content- >>>> type >>>> application/xrd+xml. >>>> >>>> In other cases the XRD consumer may want the entire sequence of >>>> XRDs >>>> retrieved by the XRD provider. In this case the XRD consumer can >>>> make a >>>> request asking for content-type application/xrds+xml. >>>> >>>> So, unless I'm missing something, we actually need both mime types. >>>> >>>> =Drummond >>>> >>>>> -----Original Message----- >>>>> From: Eran Hammer-Lahav [mailto:eran@hueniverse.com] >>>>> Sent: Tuesday, September 15, 2009 6:58 PM >>>>> To: Scott Cantor; 'XRI TC' >>>>> Subject: RE: [xri] XRDS media type >>>>> >>>>> Yes, there was which is actually used already (application/xrds >>>>> +xml). >>>> This >>>>> is why I am asking because if we register both, we will need to >>>>> deal >>>> with >>>>> the existing deployment which is not compliant. So I rather just >>>> register >>>>> one. >>>>> >>>>> EHL >>>>> >>>>>> -----Original Message----- >>>>>> From: Scott Cantor [mailto:cantor.2@osu.edu] >>>>>> Sent: Tuesday, September 15, 2009 6:54 PM >>>>>> To: Eran Hammer-Lahav; 'XRI TC' >>>>>> Subject: RE: [xri] XRDS media type >>>>>> >>>>>> Eran Hammer-Lahav wrote on 2009-09-15: >>>>>>> I am assuming that a document with a root of <XRDS> but using >>>>>>> the >>>> XRD >>>>>> XML >>>>>>> namespace is considered application/xrd+xml and not >>>>>> application/xrds+xml... >>>>>>> >>>>>>> Since they both use the same schema. >>>>>> >>>>>> They don't have to, but it's typical to do that. >>>>>> >>>>>> Was there a media type for the original XRDS? >>>>>> >>>>>> Either way it's confusing, I guess. >>>>>> >>>>>> -- Scott >>>>>> >>>>> >>>>> >>>>> --------------------------------------------------------------------- >>>>> To unsubscribe from this mail list, you must leave the OASIS TC >>>>> that >>>>> generates this mail. Follow this link to all your TCs in OASIS >>>>> at: >>>>> https://www.oasis- >>>> open.org/apps/org/workgroup/portal/my_workgroups.php >>>> >>>> >>>> >>>> --------------------------------------------------------------------- >>>> To unsubscribe from this mail list, you must leave the OASIS TC >>>> that >>>> generates this mail. Follow this link to all your TCs in OASIS at: >>>> https://www.oasis-open.org/apps/org/workgroup/portal/ >>>> my_workgroups.php >>> > >
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]