|
It depends (of course). Any change to a service needs to be communicated to the service consumers. Ultimately, it will depend on whether or not the changes are acceptable to the consumers. Clearly, changing the results without notifying the customers is a recipe for disaster. While this may occur in the real world more often than we like, it is something that should be avoided. If this happens, it is likely that the interface was not explicit enough. I ran into this situation where we had a data service that had a query by last name. The original consumers only needed a search by preferred last name. A subsequent consumer wanted to add in checks against legal name, maiden name, etc. This could have been done without changing the interface, but could result in a change in behavior for existing consumers. It all stems back to an assumption on what fields constitute "last name". Assumptions are bad. In this scenario, it would still have been okay to continue with a getByLastName operation with the new behavior, as long as all consumers agreed on the behavior and the documentation was updated so that the behavior is explicit. -tb On Sep 27, 2006, at 8:33 AM, jeffrschneider wrote:
SPONSORED LINKS
Your email settings: Individual Email|Traditional
Change settings via the Web (Yahoo! ID required) Change settings via email: Switch delivery to Daily Digest | Switch to Fully Featured Visit Your Group | Yahoo! Groups Terms of Use | Unsubscribe __,_._,___ |
- Re: [service-orientated-architecture] if f(x) changes, d... Todd Biske
- Re: [service-orientated-architecture] if f(x) chang... Chuck D'Antonio
- [service-orientated-architecture] Re: if f(x) c... Gervas Douglas
- Re: [service-orientated-architecture] Re: i... Michael Poulin
- RE: [service-orientated-architecture] R... David Orchard
- RE: [service-orientated-architectu... Michael Poulin
- Re: [service-orientated-architecture] if f(x) chang... Anne Thomas Manes
- Re: [service-orientated-architecture] if f(x) chang... Mark Baker
- Re: [service-orientated-architecture] if f(x) c... Michael Poulin
- Re: [service-orientated-architecture] if f(... Radovan Janecek
- Re: [service-orientated-architecture] i... Mark Baker
