[Engine-devel] REST-API: Exposing correlation-ID

[Livnat Peer]

On 29/05/12 09:03, Itamar Heim wrote:
> On 05/29/2012 08:56 AM, Livnat Peer wrote:
>> On 28/05/12 21:31, Itamar Heim wrote:
>>> On 05/28/2012 02:35 PM, Ori Liel wrote:
>>>> Quite a few people liked flow-id, and no one objected to it
>>>> explicitly, so I'll just go with that.
>>>>
>>>> If someone feels strongly against, please reply.
>>>
>>> I still like 'label' better.
>>> it doesn't have the context of a unique id, and is much more correct to
>>> what this is - allows the user to label a command (or a set of
>>> commands).
>>> but also doesn't imply it's unique in any way (i.e., it's like a "tag",
>>> just a better, non overloaded term for it).
>>>
>>
>> I think that flow-id is confusing. This id has nothing to do with flow,
>> it can aggregate multiple commands and it is not associated with a
>> specific user flow.
>>
>> Correlation-Id is a common name for such Id, we took it from the
>> terminology used in JMS queues, but Microsoft and Oracle are using CID
>> too.
>>
>> * http://www.microsoft.com/en-us/download/details.aspx?id=23842
>> *
>> http://sharepoint.microsoft.com/blogs/GetThePoint/Lists/Posts/Post.aspx?ID=353
>>
>> *
>> http://docs.oracle.com/cd/B14099_19/integrate.1012/b25709/com/oracle/bpel/client/CorrelationId.html
>>
> 
>  but all of those conform to the concept of an "id" uniquely identifies
> the correlation. in our case, it is not unique, and just a label the
> user sets.

I don't think the uniqueness is an issue, if not abused it will be
unique per flow/flow sequence.
Usually correlation Id enables the user to correlate between multiple
components or between multiple flows, which fits our usage of this ID.