Live from IIW - Discovery

[Monica Keller]

How to do Discovery of Activity Stream Extensions

Link header ?

Standarize rel values across protocols: HTML, Atom

Use long uris instead of short rels

PubSubHubBub wants link rel="hub"
But its generic so they should not take the good short one.


Therefore we should use

link rel="http://activitystrea.ms"

hmmmm. Eran says this is normal we would like a decision so we can
expose the stream.

What do you guys think ?

[Kevin Marks]

As I said, contention over rel values has not been a problem in the HTML space at all - microformats have extended these, a few have been incorporated into HTML5 and WhatWG maintains a wiki list of proposed ones.

Requiring a URI is thus objectively unnecessary. If this is purely for use in an XRD document, where practice allows URIs, fair enough, but if we are expecting the rel to be used in HTML link or a elements we should use a recognisable word. In that context "hub" is fine

--

You received this message because you are subscribed to the Google Groups "Activity Streams" group.
To post to this group, send email to activity...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/activity-streams?hl=en.

[Martin Atkins]

As I've noted before, my opinion here is that we don't need any
additional rel value because Atom autodiscovery ought to already be
sufficient. Any Atom feed is, via fallback behavior defined in the spec,
a valid activity stream.

For those that are unfamiliar, this would consist of a link on the
activity stream HTML page with rel="alternate",
type="application/atom+xml"... both generic Atom processors and Activity
Streams processors can and should follow this link.

Other mechanisms may also reveal activity streams URLs, of course. For
example, hopefully once OpenSocial has been extended to support the data
model for activity streams it will have an endpoint at which a list of
Activity Streams-shaped entries can be retrieved whose URL and
authentication mechanisms would be determined via the conventions
described in the OpenSocial specifications.

[Monica Keller]

What do you guys think about this

<link rel="alternate" type="application/atom+xml" title="Monica Keller
- Stream Atom" href="http://www.myspace.com/ciberch/stream/atom.xml" /
>

on the profile page ? In particular is there a good standard for a
link to a user's lifestreeam ?

On Nov 4, 5:36 pm, Martin Atkins <m...@degeneration.co.uk> wrote:
> As I've noted before, my opinion here is that we don't need any
> additional rel value because Atom autodiscovery ought to already be
> sufficient. Any Atom feed is, via fallback behavior defined in the spec,
> a valid activity stream.
>
> For those that are unfamiliar, this would consist of a link on the
> activity stream HTML page with rel="alternate",
> type="application/atom+xml"... both generic Atom processors and Activity
> Streams processors can and should follow this link.
>
> Other mechanisms may also reveal activity streams URLs, of course. For
> example, hopefully once OpenSocial has been extended to support the data
> model for activity streams it will have an endpoint at which a list of
> Activity Streams-shaped entries can be retrieved whose URL and
> authentication mechanisms would be determined via the conventions
> described in the OpenSocial specifications.
>
>
>
> Monica Keller wrote:

> > How to doDiscoveryof Activity Stream Extensions

>
> > Link header ?
>
> > Standarize rel values across protocols: HTML, Atom
>
> > Use long uris instead of short rels
>
> > PubSubHubBub wants link rel="hub"
> > But its generic so they should not take the good short one.
>
> > Therefore we should use
>
> > link rel="http://activitystrea.ms"
>
> > hmmmm. Eran says this is normal we would like a decision so we can
> > expose the stream.
>
> > What do you guys think ?
>
> > --
>
> > You received this message because you are subscribed to the Google Groups "Activity Streams" group.
> > To post to this group, send email to activity...@googlegroups.com.

> > For more options, visit this group athttp://groups.google.com/group/activity-streams?hl=en.- Hide quoted text -
>
> - Show quoted text -

[John Panzer]

That could be appropriate it the page itself contained the lifestream
in HTML format. Or a sliding window into it of course.

> For more options, visit this group at http://groups.google.com/group/activity-streams?hl=.
>
>
>

--
--
John Panzer / Google
jpa...@google.com / abstractioneer.org / @jpanzer

[Monica Keller]

Yes the stream is on the user's profile

I argued for this being the top level atom feed fro a user

www.myspace.com/ciberch/atom.xml

But there was a good point made in that we are missing the word stream
so it may not be as clear

Also do you guys know if there is any kind of standard for the
extensions atom feeds should ahve if any ?

> >> > For more options, visit this group athttp://groups.google.com/group/activity-streams?hl=en.-Hide quoted text -

>
> >> - Show quoted text -
>
> > --
>
> > You received this message because you are subscribed to the Google Groups "Activity Streams" group.
> > To post to this group, send email to activity...@googlegroups.com.

> > For more options, visit this group athttp://groups.google.com/group/activity-streams?hl=.

>
> --
> --
> John Panzer / Google

> jpan...@google.com / abstractioneer.org / @jpanzer- Hide quoted text -