[This post
has been originally published on my new technological blog on Toojoor.com - check it out at
www.toojoor.com/blog]
When
Google launched the Opensocial initiative, it was clearly stated that the current status is beta. As a proof, the current version of the API is 0.5.
The opensocial containers have to implement 2 APIs to be conform with this initiative:
As
opensocial initiative specs are still beta and many partners are already and will get involved, I noticed very early on in my first experiences discrepancies between implementations on the various containers that I can already access (or at least their sandbox...) :
igoogle itself, Google's
Orkut,
Hi5,
Ning.
So, I decided to implement this Gadget IGCheck since those discrepancies have to get eliminated over time in order to reach the initial promise of Opensocial "Write Once, Run Anywhere".
The purpose of IGCheck is to test and report which functions of the Google Gadget API (i.e., prefixed with _IG_ in their Javascript name) are available in which containers. IGCheck then produces a report comparing the current test with reference data.
To get current status, check
this page to see the most current result that we can produce concerning igoogle,
orkut, hi5 &
ning.
As the number of opensocial partners that are open for tests remain limited as of now, the following conclusions can be drawn as of now :
- Orkut uses the gmodules infrastructure of Google Gadgets to run the opensocial api (see the line "host" in the result array)
- Hi5 does the same and as such seems to have "outsourced" the task of running its opensocial container to Google (see the line "host" in the result array)
- Ning has followed a more independent approach by implementing a fully independent container named "proxy.ning.com" (see the line "host" in the result array)
- The choice of Ning has currently drawbacks: IGCheck finds only 6 _IG_ functions whereas 55 are present in the others. (see the line "count" in the result array)
- Ning seems to have created its own function(s) with _IG_Prefix: a function "G_Tabs" is present in the Ning container but nowhere else (see the line "_IG_Tabs" in the result array).
The conclusions above are preliminary results only:
- the current and future containers will have to converge after the current pioneering phase in order to reach the promise "Write Once, Run Anywhere" although the fact that IGCheck itself is a proof of partial achievement (IGCheck source code is unique for all containers)
- the opensocial initiative will have to get more precise on Google Gadget API requirements: the debate on this topic is currently hot on the Opensocial discussion group as nothing (at least to my knowledge...) has been clearly and officially stated in this area until now . A poll is under way. The target is to get opinions of developpers and eventually to define to which extent opensocial containers have to implement the _IG_ API.
To use IGCheck in the above container, please use
http://www.toojoor.com/opensocial/js/google/gadget/ig-check.xml as URL for application / gadget in the appropriate box on your social network of choice.
Stay tuned (via
RSS also): more result to come when we can access more containers....
If you want to reproduce my tests and don't know how to access, read my 2 tutorials for
Orkut &
Hi5. The one about Ning will follow shortly.
PS: if you now of other containers that are open for test, please, let us know!
Source: blog
Media & Tech (par didier durand)
avec 250 millions de visiteurs uniques, Yahoo Mail en troisième position cartonne toujours loin devant GMail et Microsoft Hotmail
Je m'étais inscrit sur Orkut pour tester mon gadget IGCheck sur ce réseau social. J'ai des extensions de ce gadget en prévision, mais pour les tester, j'ai besoin "d'amis" au sens Orkut. Je me suis donc permis de contacter certains d'entre vous. merci d'avance de donner suite.
Les prévisions pour les années à venir sont tout aussi encourageantes: la croissance de la publicité sur les réseaux sociaux y sera très supérieure (voir les fins de ligne ci-dessus) à la croissance de la publicité Internet en général: +24% pour la pub Internet globale contre +155% pour les réseaux sociaux
En regardent ce tableau, il y a donc des industries (fournitures de bureau, sous-vêtements) où le trafic payé permet d'atteindre 50% de l'objectif idéal. Par contre, pour d'autres, on ne peut espérer que le tiers voire le quart en ouvrant le porte-monnaie même en grand...
Si vous voulez le rejoindre, ce sera bien pour mes tests.: il n'y a pour l'instant que moi et Unique, le Maine Coon familial (un peu frimeur... voir ci-dessus) . Pour y adhérer: aller à protopensocial.ning.com mais ce sera juste afin de voir, je n'y mettrai pas beaucoup de "vrai" contenus. Merci par avance de votre aide.
