opensocial-samplecontainer

[Kevin Brown <et...@apache.org>]

Hi Cassie,

I'm a little unclear on why this particular bit of code was made into a
feature. Isn't this the actual container javascript, or is this simply the
abstract mapping of a "container" that opensocial needs to function? In the
former case it shouldn't be a feature, but in the latter it should. Also in
the latter case, it seems like this should be <container> code and not
<gadget> code if it is intended to be a feature. Am I missing something
here? This is based mostly on reading the comments rather than any thorough
examination of what the code is doing, so if this is the intended behavior
than perhaps we need to modify the comments.

~Kevin

Re: opensocial-samplecontainer

[Kevin Brown <et...@google.com>]

Ok, that does sound like it's correct; the terminology is a little awkward
because generally when referring to the "container" we were referring to the
site generating the iframes (igoogle, orkut, etc.)

~Kevin

On Jan 11, 2008 9:49 AM, Cassie <do...@apache.org> wrote:

> So how opensocial works today is that each instance of a gadget gets its
> own
> opensocial container. This is confusing when added to the gadgets
> terminology, but opensocial came up with its container spi first :)
>
> In addition to the "container" code that is stuffed inside the iframe,
> there
> is also code that goes in the parent iframe (the other half of some ifpc
> calls). The sample container does not have any of this right now. However,
> I
> am working on something at the moment which will add js to the outer
> iframe.
> This code is simply modeled like one of the sampleX.html files.
>
> So I think I got it all right... hope that clears things up!
>
> - Cassie
>
>
> On Jan 10, 2008 7:21 PM, Kevin Brown <et...@apache.org> wrote:
>
> > Hi Cassie,
> >
> > I'm a little unclear on why this particular bit of code was made into a
> > feature. Isn't this the actual container javascript, or is this simply
> the
> > abstract mapping of a "container" that opensocial needs to function? In
> > the
> > former case it shouldn't be a feature, but in the latter it should. Also
> > in
> > the latter case, it seems like this should be <container> code and not
> > <gadget> code if it is intended to be a feature. Am I missing something
> > here? This is based mostly on reading the comments rather than any
> > thorough
> > examination of what the code is doing, so if this is the intended
> behavior
> > than perhaps we need to modify the comments.
> >
> > ~Kevin
> >
>

Re: opensocial-samplecontainer

[Cassie <do...@apache.org>]

So how opensocial works today is that each instance of a gadget gets its own
opensocial container. This is confusing when added to the gadgets
terminology, but opensocial came up with its container spi first :)

In addition to the "container" code that is stuffed inside the iframe, there
is also code that goes in the parent iframe (the other half of some ifpc
calls). The sample container does not have any of this right now. However, I
am working on something at the moment which will add js to the outer iframe.
This code is simply modeled like one of the sampleX.html files.

So I think I got it all right... hope that clears things up!

- Cassie


On Jan 10, 2008 7:21 PM, Kevin Brown <et...@apache.org> wrote:

> Hi Cassie,
>
> I'm a little unclear on why this particular bit of code was made into a
> feature. Isn't this the actual container javascript, or is this simply the
> abstract mapping of a "container" that opensocial needs to function? In
> the
> former case it shouldn't be a feature, but in the latter it should. Also
> in
> the latter case, it seems like this should be <container> code and not
> <gadget> code if it is intended to be a feature. Am I missing something
> here? This is based mostly on reading the comments rather than any
> thorough
> examination of what the code is doing, so if this is the intended behavior
> than perhaps we need to modify the comments.
>
> ~Kevin
>