Yep, the WG envisioned this, the final spec did not. :)

On Saturday, October 24, 2009, Bob Wyman <[email protected]> wrote:
> Actually, the PSHB use case *was* frequently discussed in the Atom WG... The 
> reason for this is that what PSHB does is provide pretty much what FeedMesh 
> was intended to provide - a topic based subset of the content-based system 
> that Pubsub.com provided.
> An atom:source, if present in an entry, should never be modified or 
> overwritten by an aggregator. An aggregator should only add atom:source if it 
> isn't already present.
> If you want to show provenence, you need to add an extension element. As 
> pointed out in an earlier message, the WG was aware that provenence was 
> useful but we couldn't get consensus on how to record it.
> Please, if you define a provenence element, include in the definition a 
> requirement to remove that element prior to canonicalisation when verifying 
> signatures...
> bob wyman
> On Oct 24, 2009 12:45 PM, "John Panzer" <[email protected]> wrote:
>
> Agreed to keeping the tone civil -- assume good intentions until proven 
> otherwise; it can only help.
>
> Mea culpa for not noticing this issue with PubSubHubbub.  The Atom spec 
> didn't envision this use case and so atom:source is almost, but not quite, 
> what's needed -- thus the confusion is understandable.
>
>
>
> The right thing to do IMHO is to add something like a psh:provenance element 
> that is just like atom:source but tracks the most recent context of the entry.
> --
> John Panzer / Google
> [email protected] / abstractioneer.org <http://www.abstractioneer.org/> / 
> @jpanzer
>
>
>
> On Sat, Oct 24, 2009 at 7:24 AM, Pádraic Brady <[email protected]> 
> wrote:
>>
>> Whoa ;). Ton...
>
>

-- 
--
John Panzer / Google
[email protected] / abstractioneer.org / @jpanzer

Reply via email to