- From: Mark Baker <distobj@acm.org>
- Date: Wed, 26 Jun 2002 14:08:10 -0400
- To: "Cutler, Roger (RogerCutler)" <RogerCutler@ChevronTexaco.com>
- Cc: w3arch <www-ws-arch@w3.org>
That's a fine summary, Roger. I agree with both "definitions", in the context of Web services. But I'd just like to point out where the Web's version of late binding differs. On Wed, Jun 26, 2002 at 08:31:18AM -0700, Cutler, Roger (RogerCutler) wrote: > The late binding scenario would be to write code which discovers at runtime, > by examining the WSDL of the web service, what the interface is and then > uses that interface. That's exactly right. The problem with this approach in a distributed setting, is that it requires a priori knowledge of the interface. So if I came across some WSDL that said that http://www.cherontexaco.com/foo/bar implemented the "baz", "goo", and "gurgle" methods, I don't know what those mean, if it's safe for me to invoke them, what might happen when I do, etc.. What the Web does, that goes beyond OLE, is that it defines methods that actually allow you to use the service. COM/OLE just allowed you to pass object references around, instantiate new objects, etc.., but not use it, in the sense of being able to take advantage of the service(s) that it offers. [1] http://www.mvps.org/word/FAQs/InterDev/EarlyvsLateBinding.htm MB -- Mark Baker, CTO, Idokorro Mobile (formerly Planetfred) Ottawa, Ontario, CANADA. distobj@acm.org http://www.markbaker.ca http://www.idokorro.com
Received on Wednesday, 26 June 2002 13:57:40 UTC