Weird verbs: "nimbly dodged a grenade"

[Chris Messina]

Take a look:

http://wiki.activitystrea.ms/Nethernet-Verb-Examples

Curious what we tell these guys (and other more esoteric sources of activities).

Chris

--
Chris Messina
Open Web Advocate

factoryjoe.com // diso-project.org // openid.net // vidoop.com
This email is:   [ ] bloggable    [X] ask first   [ ] private

[David Cramer]

This is always one of the big concerns I had with the way things are "standardized". Keep in mind I still haven't gotten the time to read over the specs (only a quick glimpse), but verbs/nouns are never ending.

We ending up just taking a generic approach in our service. Everything is rendered as %user %verb %amount %noun, or %user %verb %title, and it works fairly well. The rendering (which kind of goes hand in hand w/ the noun, as photos render different from status updates) we left up to an entirely different aspect.

[Chris Messina]

Well, I do think that there should natural economic benefits of figuring out a way to shoe-horn esoteric verbs into more mainstream verb URIs, if only so that activity stream aggregators can publish content in a productive way.

That said, I do actually hope that a large number of verb and object types proliferate — especially as a way of doing interop between similar communities — even if they never gain critical mass, let's say, to make it into a mainline social network.

In the specific case of Nethernet, I suppose they could use their own namespace and invent their own set of verbs. There's nothing to say they can't or shouldn't... but perhaps it points to the need, as you're suggesting, for a more generic "acted" or "did something" catchall verb and object type...?

Chris

[Leah Culver]

Wow. I say let them have their namespace!

Jaunt!
Leah

[Sylvain Hellegouarch]

2009/4/29 Chris Messina <chris....@gmail.com>:

> Well, I do think that there should natural economic benefits of figuring out
> a way to shoe-horn esoteric verbs into more mainstream verb URIs, if only so
> that activity stream aggregators can publish content in a productive way.
> That said, I do actually hope that a large number of verb and object types
> proliferate — especially as a way of doing interop between similar
> communities — even if they never gain critical mass, let's say, to make it
> into a mainline social network.
> In the specific case of Nethernet, I suppose they could use their own
> namespace and invent their own set of verbs. There's nothing to say they
> can't or shouldn't... but perhaps it points to the need, as you're
> suggesting, for a more generic "acted" or "did something" catchall verb and
> object type...?
> Chris
>

Well ASms verbs and nouns are inviting to create your own folksnomy,
I'm not surprised some people are defining their own. The ASms schema
clearly states those that are standardized so if a service decides to
create a verb that means essentially the same thing as Post but with a
different namespace, it'd their loss in the end.

I find it healthy even if it'd be useful to have some kind of larger
IANA service that lists common verbs beyond what the spec defines.

- Sylvain

[Chris Messina]

On Wed, Apr 29, 2009 at 12:26 AM, Sylvain Hellegouarch <sylvain.he...@gmail.com> wrote:

Well ASms verbs and nouns are inviting to create your own folksnomy,
I'm not surprised some people are defining their own. The ASms schema
clearly states those that are standardized so if a service decides to
create a verb that means essentially the same thing as Post but with a
different namespace, it'd their loss in the end.

I find it healthy even if it'd be useful to have some kind of larger
IANA service that lists common verbs beyond what the spec defines.

Precisely. Over time, I would like for activitystrea.ms to become that lookup and registration resource, similar to hashtags.org.

Just need the time, energy and software to make that happen! I already mocked it up awhile back:

http://www.flickr.com/photos/factoryjoe/3244608669/

[Ari Steinberg]

hate to be a naysayer, but i still don't really see a huge amount of
value in trying to classify all these verbs. i think it's more useful
to just let the users say what they want to say and attach the
content/metadata in the appropriate way, and basically treat
everything as a "share". the reporting on background activity part of
this tends to be generally pretty uninteresting and typically often
not what users really care about sharing.
-ari

[Martin Atkins]

Chris Messina wrote:
> Take a look:
>
> http://wiki.activitystrea.ms/Nethernet-Verb-Examples
>
> Curious what we tell these guys (and other more esoteric sources of
> activities).
>

Without meaning to sound harsh, what I'd tell these guys right now is
that they're out of scope for the moment and that I hope to address
these more esoteric verbs via a template bundles mechanism that can be
built on top of the basic activity extensions elements to tell consumers
how to render verb and object type combos that they don't explicitly
support.

Of course, whether "a grenade" can sensibly be modeled as an Atom entry
is debatable. Right now the focus is on stuff that exists on the web,
and while online games are arguably "on the web" they are sufficiently
divorced from the 80% cases that we're trying to model in this first go
that I think we have no choice but to punt on them for the moment.

I'd like to echo what Ari said: we want to minimize the number of verbs
that are in common use. Whenever a new use-case presents itself I always
try to first model it as "post"; already I have a plan to replace the
"share" verb with the idea of posting a bookmark, or pointer, or
whatever we end up calling the generalized "reference to some other
object" mechanism.

[Chris Messina]

On Wed, Apr 29, 2009 at 5:34 PM, Martin Atkins <ma...@degeneration.co.uk> wrote:

Chris Messina wrote:
> Take a look:
>
> http://wiki.activitystrea.ms/Nethernet-Verb-Examples
>
> Curious what we tell these guys (and other more esoteric sources of
> activities).
>

I'd like to echo what Ari said: we want to minimize the number of verbs
that are in common use. Whenever a new use-case presents itself I always
try to first model it as "post"; already I have a plan to replace the
"share" verb with the idea of posting a bookmark, or pointer, or
whatever we end up calling the generalized "reference to some other
object" mechanism.

I think that's totally reasonable.

I do, however, imagine that we will see all manner of folks interested in expressing their verb-object combo of choice and we need to have a ready answer for them.

I think part of it is putting them through the exploration and documentation (really, just the scientific method) phases, collecting information about what they're really trying to express.

I agree that we need to start simple, hitting the 80-90% use cases. For the other 10-20%, I think we do need an answer, and the Process I think is what will save us much hemming and hawing over what's "in" and what "isn't".

http://wiki.activitystrea.ms/Process

Chris