Proposals to push to v.Next

[Scott Seely]

I’ve signed up for a few proposals that I will not be able to get out as prototypes until the end of the month at the earliest.

·         http://wiki.opensocial.org/index.php?title=Implementing_IS_FRIENDS_WITH

·         http://wiki.opensocial.org/index.php?title=Min_age_metadata_for_ModulePrefs

·         http://wiki.opensocial.org/index.php?title=RequestShareApp_and_requestSendMessage_should_use_IdSpec

·         http://wiki.opensocial.org/index.php?title=Anonymous_Viewer

 

If anyone wants to step forward and claim one of these, you need to do so by end of day tomorrow (3-31). Otherwise, these proposals will go back into the queue for the normal process in v.Next.

 

 

Scott Seely

architect

email  sse...@myspace.com

 

 

[Evan Gilbert]

Even if these items don't make it into the spec for 0.9, they are still a good blueprint for any container implementing these features. Specifically:
- If you're returning an anonymous viewer, use -1 as the viwer ID
- If you want to implement mutual friends, follow the IS_FRIENDS_WITH spec
- If you have minimum age requirements for gadgets, use the min_age XML.

Moving forward, we'd like for many spec enhancements to emerge from actual use, but this only works if containers make an implement the same way, even if it's not yet start of the "official" spec.

Evan

[Paul Lindner]

Anonymous User support has a prototype.  Patch for Java/JS attached to the relevant Jira issue.

https://issues.apache.org/jira/browse/SHINDIG-734 

Issues encountered -- to make it easy for javascript implementations I added an isAnonymous() convenience function that tests for the Anonymous User ID.  Suggest adding isAnonymous() to the person spec.

[Lane LiaBraaten]

+1 - isAnonymous() seems like a more generic solution than the -1
convention so it's a worthwhile addition.

-Lane

> > ·        http://wiki.opensocial.org/index.php?title=Min_age_metadata_for_Modul...
>
> > ·        http://wiki.opensocial.org/index.php?title=RequestShareApp_and_reques...
>
> > ·        http://wiki.opensocial.org/index.php?title=Anonymous_Viewer

[Arne Roomann-Kurrik]

+1 to isAnonymous(), certainly makes sense if we have isViewer and isOwner.

~Arne

--
OpenSocial IRC - irc://irc.freenode.net/opensocial

[Evan Gilbert]

New APIs don't really have these "function" things any more on the person object (and we need a way to represent in REST/RPC). So fine having isAnon() in the old JS, but think id==-1 is probably the right thing for JSON/REST/RPC.

[Scott Seely]

Please write up a patch.

[Lane LiaBraaten]

Kevin and I would also like to propose moving the "Versioning Applications" [1] feature to v.NEXT.  There is some working code in Shindig that we plan to shop around to app developers to get more feedback and make sure this concept is fully baked.

Please reply if you feel strongly that this should remain in v0.9.

Thanks,
Lane

[1] http://wiki.opensocial.org/index.php?title=Versioning_Applications

On Mon, Mar 30, 2009 at 6:16 PM, Scott Seely <sSe...@myspace.com> wrote:

[Scott Seely]

Is there anything in the spec that is limiting? I’ve shopped this around to some of our developers as well as the application approval team and the details seem about right.

 

 

 

From: opensocial-an...@googlegroups.com [mailto:opensocial-an...@googlegroups.com] On Behalf Of Lane LiaBraaten
Sent: Thursday, April 02, 2009 1:42 PM
To: opensocial-an...@googlegroups.com
Subject: [opensocial-and-gadgets-spec] Re: Proposals to push to v.Next

 

Kevin and I would also like to propose moving the "Versioning Applications" [1] feature to v.NEXT.  There is some working code in Shindig that we plan to shop around to app developers to get more feedback and make sure this concept is fully baked.

[Kevin Brown]

The biggest concern that I have with it is that it's already a week into April and we haven't really had a chance to see how well the manifests address the issues developers have raised for versioning. We will be deploying shindig implementation as is to google containers at some point in the next few weeks, but I'm reluctant to set it in stone until we get some useful feedback.

[Lane LiaBraaten]

On Tue, Mar 31, 2009 at 12:54 PM, Evan Gilbert <uid...@google.com> wrote:

New APIs don't really have these "function" things any more on the person object (and we need a way to represent in REST/RPC). So fine having isAnon() in the old JS, but think id==-1 is probably the right thing for JSON/REST/RPC.

This is a good point...can we just punt on isAnon() since we're looking to deprecate the old API anyway?

-Lane

[Scott Seely]

I’m fine with this reasoning. Given the roughness of the osapi, it seems that we should do the same there as well.