On Thu, Mar 20, 2003 at 06:19:57PM -0800, bill@daze.net wrote:
[ on the OpenSRS<->client interface and the OpenSRS<->Registry interface ]
> The problem with this is that #1 is dependent on #2. For most
> commands (i.e. registry commands -- domain availability, register
> domain, renew domain, transfer domain), OpenSRS passes your request
> to the registry and returns to you the registry return code. If
> they shut off #2, #1 will never complete or always fail for registry
> commands. The only pieces they could break out are internal OpenSRS
> stuff like profile management.
This is only going to be a problem if OpenSRS relies on other
registries to error check client submissions, which I'm sure they
don't.
The point about eliminating #2, the other registries, is that we
*want* OpenSRS to fake up a response, having error-checked the
submission to their satisfaction. This is the only way that they can
provide repeatable test cases for their API.
Chris.
This archive was generated by hypermail 2.1.3 : Tue Oct 19 2004 - 23:37:39 EDT