Re: TAG-Organized Unconference in San Francisco on 20 April: Extensible Web Summit

https://lists.w3.org/Archives/Public/public-webpayments-ig/2015Mar/0134.html

Indeed an interesting topic since the Web (browser) of today [1,2,3] is anything but "Extensible".

URIs, Linked Data, etc. is great but they don't give you access to things inside of the client platform making the browser a dumb terminal.
With W3C's SysApps closing without reaching its intended goal, the downward spiral is continuing.

Wait a minute, aren't "Permissions" supposed to enable access to the client platform?
Yes, it is obvious that "Your Location" fits pretty well but it is easy finding things that do not:
http://webpki.org/papers/permissions.pdf

Given these facts, the chance for a browser-vendor-supported "Web Payment API" appears to be very low.

Is that a problem?  No, we rather need a generic and standardized browser extension mechanism capable of dealing with a much wider set of applications, ranging from entirely proprietary one-of-a-kind-systems to community-based schemes, and eventually even including a few universally recognized de-facto or industry standards.

By building on the already established "AppStore" concept holding vetted applications, native extension modules could be distributed in a scalable fashion without the drawbacks of predecessors such as ActiveX and NPAPI.

- Anders Rundgren

1] Native-mode plugins have been deprecated and other extension methods like services running on "localhost" are in the process of being locked down
2] Although sometimes useful, "polyfilling" doesn't give you access to client-side resources
3] Open or closed source; the vendors unilaterally decide what goes into the final product

Received on Saturday, 28 March 2015 16:32:25 UTC