On Wed, Oct 22, 2008 at 7:54 PM, Dan Becker <[EMAIL PROTECTED]> wrote:
> I've not read the OSOA HTTP binding spec, but I do have some Tuscany > experience with the HTTP and Atom bindings. I am especially interested in > their support for caching and conditional commands. > > I gladly would like to help out on this one. > > > Simon Laws wrote: > >> On Wed, Oct 22, 2008 at 6:01 PM, Raymond Feng <[EMAIL PROTECTED]> >> wrote: >> >> I briefly read the draft. My impression is that it tries to come up some >>> sort of poor-man's Web Service support over HTTP using RPC style >>> (tunneling >>> the invocations over HTTP). What are the advantages over SOAP/HTTP or >>> JSONRPC/HTTP? I'm wondering if it would be better to focus on the REST >>> style >>> by mapping HTTP methods into a set of business operations that deal with >>> resources. >>> >>> Thanks, >>> Raymond >>> >>> From: ant elder >>> Sent: Wednesday, October 22, 2008 3:50 AM >>> To: dev@tuscany.apache.org >>> Subject: Draft OASIS spec for binding.http >>> >>> >>> >>> A draft spec for an HTTP binding has been posted to the OASIS bindings >>> mailing list - >>> http://lists.oasis-open.org/archives/sca-bindings/200810/msg00078.html >>> >>> It has some interesting things such as using the new wireFormat element >>> and >>> the suggestion that you could extend that for atom and json support. I'm >>> interested in doing an implementation of this spec, anyone interested in >>> helping? >>> >>> ...ant >>> >>> >> I think this will fall into the different strokes for different folks >> category. If OASIS pick this up then people will want to use this binding >> so >> we should look at it. We can then probably give OASIS some good feedback >> on >> how to improve the spec. In the PHP SCA implementation we had a similar >> "REST" binding and people quite liked it. Don't have an opinion on basing >> the Web2.0 bindings on it. >> >> Simon >> >> > > -- > Thanks, Dan Becker > Sounds good Dan. I should point out that, in my previous post, I wasn't suggesting that we should wait until someone asks us for this. I was trying to say it's often an advantage to have several different approaches. Gives us a wide view on what works and what doesn't Simon
