[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Streams and Reports and StreamBase
I had a critical realization about Streams when talking with Steve Ray about the use of Streams in FSGIM. A StreamBase consists of dtStart duration related-to (interval identifier) Sequence (a set of intervals) And a Payload In effect, it is a Gluon and Sequence packaged as one. A StreamBase can have two uses. It can specify a stream. Or it can describe the computation of the Sequence that will later be conveyed as part of a stream. If we take the simple case for a Stream, Measure the Temperature at 5 minute intervals starting at 3:00 PM and going through 10:00 PM Our focus has been on conveying 3:00 PM (dtstart) 5 Minutes (duration) [01] (related to interval 1) Room temperature (payload) 01:84 (first interval UID and measurement) (note interval UID can be anything that enables ordering. The actual time is a natural choice) 02:84 03:85 04:85 … 1) It is perhaps far more useful, to our readers and others, the be able to request the stream reporting, i.e. 3:00 PM (dtstart) 5 Minutes (duration) [01] (related to interval 1) ß-not needed because implicit Room temperature (payload) 10:00 PM (dtend) Is this a stream specifier? If so, it is akin to the ReportSpecifier in Energy Interoperation. I think the Stream is easier to understand if we *start* with the Specifier and then describe the StreamBase 2) Partitions and Sequences. As written, a Stream can be any sort of Sequence. Is it useful to include any type other than the Partition? If it is not restricted to the Partition, what does it mean? I can imagine a complex Sequence, perhaps measuring a cluster of different items at a scattershot of times, but is it useful? 3) Temporal Relations. This is a special case of the Sequence discussion. If the measurement sums (for example) some activity over a three minute interval, then would it make sense to convey that “wait 10 minutes and start another sum?” I know that notions of waiting minutes is the source of a lot of confusion in the FSGIM. They think of a measurement as an instantaneous act separated by durations, while I have always implicitly considered a meter reading as summing the delta over the interval. tc “The single biggest problem in communication is the illusion that it has taken place.” – George Bernard Shaw.
|
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]