[jdev] Message Read Receipts

[Spencer MacDonald]

So my suggestion is "Chat Markers", where you mark a given message with a
Chat Maker.

So you could retrieve your chat makers between you and a given jid:

<iq type='get' id='chat-marker-1'>
<query xmlns='urn:xmpp:chat-marker:tmp'>
<with>juliet at example.com</with>
</query>
</iq>

You would then receive the 'read' and 'received' chat marker between the
you and the other jid.

<iq to='romeo at example.com/phone' type='result' id='chat-marker-1'>
<query xmlns='urn:xmpp:chat-marker:tmp'>
<read to='juliet at example.net' id='message-3'
xmlns="urn:xmpp:chat-marker:tmp"/>
<received to='juliet at example.net' id='message-3'
xmlns="urn:xmpp:chat-marker:tmp"/>
 <read from='juliet at example.net' id='message-1'
xmlns="urn:xmpp:chat-marker:tmp"/>
<received from='juliet at example.net' id='message-2'
xmlns="urn:xmpp:chat-marker:tmp"/>
</query>
</iq>

You can update you Chat Markers like so:

 <iq type='set' id='chat-marker-2'>
  <chat-markers xmlns='urn:xmpp:chat-marker:tmp'>
 <with>juliet at example.net</with>
<read>message-4</read>
<received>message-4</read>
  </chat-markers>
 </iq>

These would be pushed to your other devices

<iq to='romeo at example.com/tablet' type='set' id='chat-marker-3'>
<query xmlns='urn:xmpp:chat-marker:tmp'>
<read to='juliet at example.net' id='message-4'
xmlns="urn:xmpp:chat-marker:tmp"/>
<received to='juliet at example.net' id='message-4'
xmlns="urn:xmpp:chat-marker:tmp"/>
</query>
</iq>

and to other user's devices

<iq to='juliet at example.com/phone' type='set' id='chat-marker-3'>
<query xmlns='urn:xmpp:chat-marker:tmp'>
<read from='romeo at example.net' id='message-4'
xmlns="urn:xmpp:chat-marker:tmp"/>
<received from='romeo at example.net' id='message-4'
xmlns="urn:xmpp:chat-marker:tmp"/>
</query>
</iq>

Regards

Spencer




On Thu, May 16, 2013 at 9:59 AM, Jon Doyle <jdoyle at communigate.com> wrote:

> Hello;
>
> *[JH] I do think another interesting case is where the message is in fact
> “read” by a agent/bot/program (e.g. service bots or even M2M messaging),  a
> timestamped receipt request is in fact helpful as there is a temporal
> difference between the reception, processing and response to a message –
> would be a real-world case 1 example.*
>
>
> I can see the value, and would propose that really means some "tangible"
> action must occur for it to be a valid "receipt". No different than in the
> human world where I sign for an envelope at the front door. I say this
> because what you are saying is that the "value" here is to show the
> "forensic" data, that "time-stamp" you mention. It should in fact be the
> record of when the button was pushed, or the app "signed off" on receipt.
> It could include a variety of things, the other side can see a display
> (client side) of "read", but the data around that transaction could be
> stored as a "receipt" which could be used in a lot of business or forensic
> applications. All of which in theory could have tremendous value in a
> variety of usage model cases.
>
> I think all of us that have used Skype or iMessage know about the
> challenges of "message receipt" in a multi-cleint scenario.
>
> Regards,
>
> Jon
>
>
> _______________________________________________
> JDev mailing list
> Info: http://mail.jabber.org/mailman/listinfo/jdev
> Unsubscribe: JDev-unsubscribe at jabber.org
> _______________________________________________
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.jabber.org/jdev/attachments/20130516/9cf919ff/attachment-0001.html>