- From: Mike Dierken <mike@dataconcert.com>
- Date: Mon, 11 Feb 2002 09:50:37 -0800
- To: noah_mendelsohn@us.ibm.com, distobj@acm.org
- Cc: mnot@mnot.net, skw@hplb.hpl.hp.com, xml-dist-app@w3.org
> -----Original Message----- > From: noah_mendelsohn@us.ibm.com [mailto:noah_mendelsohn@us.ibm.com] > > The only difference I see per web architecture between > > http://numbers.com/multiply?inputnumber1="3"+inputnumber2="4" > > and > > http://numbers.com/?operation="multiply"+inputnumber1="3"+inpu tnumber2="4" > > is that we can guess that to implement both multiplication and division, > the first style would result in two separate parameterized web resources > (a multiplying resource, and a dividing resource) while the second would > use a single parameterized resource (an arithmetic resource). > I don't see the two as deeply different from a web architecture point of view. There are large number of arithmetic operators and with the first URI, the server can specify where each operator lives and the client fills in the blanks, with the second the client needs to know /that particular servers/ name for the operator and also fill in the blanks. A 'prepackaged' callback (like an HTML FORM) with default arguments (hidden form fields already filled in) can help, but requires more coordination on the format of the 'callback' structure. Mike
Received on Monday, 11 February 2002 12:52:17 UTC