- From: Ugo Corda <UCorda@SeeBeyond.com>
- Date: Thu, 7 Aug 2003 15:20:10 -0700
- To: "Christopher B Ferris" <chrisfer@us.ibm.com>, "Cutler, Roger (RogerCutler)" <RogerCutler@chevrontexaco.com>
- Cc: <www-ws-arch@w3.org>
> Note also that your characterization of "someone willing to wait an hour, > or for someone with an attention span like mine" is suggestive of a human making a request. While it is true > that one might imagine that Web services might be invoked by a "client" that also exposes a human > interface (with a human at the controls), I thought that we had established that Web services were primarily > addressed at machine-to-machine interaction. I don't think we should push the machine-to-machine point too far though, at least when it comes to response times. Given the positive responses received so far by WSRP, and given the fact that portals are very important in the enterprise, there will probably be at least an important and popular case where the WS consumer (specifically a WSRP Consumer) is regularly driven by a human interface (the browser interacting with the WSRP Consumer). In that case, response time expectations are directly transferred from the human interface to the machine-to-machine interaction (WSRP Consumer and WSRP Producer). Ugo
Received on Thursday, 7 August 2003 18:20:16 UTC