I am sorry to keep banging on about this, but we cannot brush
asynchronicity under the carpet as a "2.0" issue when we still
actually have it present in the 1.0 interface... I think that we need
to be clear in the specification exactly how we expect this calls to
behave - i.e. whether the web service call is expected to complete
before the data transfer has taken place - if we say that there is no
asynchronicity in 1.0 implementors might take it to mean that
pushDataFromVoSpace should not return until the data transfer has
taken place - which is probably not the behaviour we want - right?
Paul.
On 14.06.2006, at 16:40, Dave Morris wrote:
> Paul Harrison wrote:
>
>> On 14.06.2006, at 13:52, Dave Morris wrote:
>>
>>> This implies asynchronous transfers, which we have not defined
>>> in enough detail yet.
>>
>> We have asynchronous transfers - are not the push bulk data
>> methods asynchronous, web service call finishes before data has
>> arrived? Certainly pushDataToVoSpace is....
>
> Yep, you are right, PushdataToVoSpace is asynchronous.
> However, the client is controlling the push, so it already knows
> when it has finished sending the data.
Received on 2006-06-16Z12:57:51