When is it OpenSocial, and when it is just a gadget?

[Alejandro Rivero]

After a couple months, there are already some repositories of
"opensocial applications" listing a good bunch:

http://www.opensocialapplist.com/
http://opensocialdirectory.org/wiki/Main_Page
http://www.google-info.net/osocial/directory/index.php

my surprise is at least one half of the listed apps do not use the
opensocial javascript object at all. It seems all they require is a
way to store defaults: they are just "igoogle" apps, not opensocial.

I wonder, could the people uploading these applications to indicate
somewhere (in the code comments, or in the listing entry of the
repositories, or both) what parts of opensocial API do they use and
what parts do they intend to use? Think that container developers will
be trying your apps in their containers; it is not sense to attempt as
a test an application that does not use the part of the API you are
testing.

Also it could be good coding style to mark lines which are now done
with your own data format but you expect to evolve into specific
OpenSocial Data API formats.

Thanks

Alejandro
(who is trying a container for BOINC based websites).

[Didier DURAND]

Hello Alejandro,

I would agree with you: I thing that the current confusion comes from
the fact that a pre-requirement to opensocial isfor the containing
social network to also implement part of Google Gadget.

As such, it gives some "pure" (not using opensocial per se) gadgets to
also run in new containers (Hi5, ning, etc...) .

So, some people whose gadget now works elsewhere then igoogle claim
that it is an opensocial application

It should clarify over months
regards
didier

On Dec 3, 4:22 pm, Alejandro Rivero <Al.Riv...@gmail.com> wrote:
> After a couple months, there are already some repositories of
> "opensocial applications" listing a good bunch:
>

> http://www.opensocialapplist.com/http://opensocialdirectory.org/wiki/Main_Pagehttp://www.google-info.net/osocial/directory/index.php

[Alejandro Rivero]

I hope it will clarify. I suppose that catalog managers can not decide
neither, because some of these pages could be really intending to
introduce opensocial features even if very minor. For instance a
whiteboard app installed in the homepage of some orkut user could want
to be able, in future versions, to log the id of the user doing the
last modification to the whiteboard. Or similarly simple things. And
persistence, etc.

So one can not rule out applications on the simple fact that the
current version is not using any opensocial.xxx function. Worse, they
could be using only the Data APIs without the Javascript API (not very
sensible thing to do, is it?).

A possibility, instead of yes/no, is to introduce in these app
catalogs some metric of "OpenSocialness", perhaps by counting how many
different calls to opensocial.xxx does a gadget contain.

On Dec 3, 6:29 pm, Didier DURAND <durand.did...@gmail.com> wrote:
> Hello Alejandro,
>
> I would agree with you: I thing that the current confusion comes from
> the fact that a pre-requirement to opensocial isfor the containing
> social network to also implement part of Google Gadget.
>
> As such, it gives some "pure" (not using opensocial per se) gadgets to
> also run in new containers (Hi5, ning, etc...) .
>
> So, some people whose gadget now works elsewhere then igoogle claim
> that it is an opensocial application
>
> It should clarify over months
> regards
> didier
>
> On Dec 3, 4:22 pm, Alejandro Rivero <Al.Riv...@gmail.com> wrote:
>
> > After a couple months, there are already some repositories of
> > "opensocial applications" listing a good bunch:
>

> >http://www.opensocialapplist.com/http://opensocialdirectory.org/wiki/...