Enhancement #498

Integrate causal vector for events

Added by J. Wienke over 12 years ago. Updated over 12 years ago.

Status:ClosedStart date:09/08/2011
Priority:UrgentDue date:
Assignee:J. Moringen% Done:

100%

Category:ProtocolEstimated time:5.00 hours
Target version:0.5

Description

Events are caused by other events and this information is useful very often. We need a way to tell which event caused which other event, also in the case where we extend the uri schema such uris need to be stored in events somewhere.

There is actually need from the HUMAVIPS project:

3) it is obvious that the result of (any) computation should reference the
source image in some way (it's also mentioned in D1.3), is there a
recommended way of doing this? (timestamps, unique IDs, pointers, ...)


Subtasks

Tasks #560: Add cause checking to integration testResolvedJ. Moringen

Tasks #561: Implement cause vector in pythonResolvedJ. Wienke

Tasks #562: Implement cause vector in javaResolvedJ. Wienke

Tasks #563: Implement cause vector in clResolvedJ. Moringen

Enhancement #656: Logger should display cause vectors in detailed formattin...ResolvedJ. Moringen


Related issues

Blocked by Robotics Service Bus - Tasks #553: Introduce EventId comparable to C++ Resolved 09/07/2011
Blocks Robotics Service Bus - Tasks #565: Let RPC mechanism use cause vector Resolved 09/09/2011

Associated revisions

Revision 9f722ecb
Added by J. Wienke over 12 years ago

add a basic cause vector to each event. Protocol and C++ implementation adapted.

refs #498

Revision ea5e82da
Added by J. Wienke over 12 years ago

add a basic cause vector to each event. Protocol and C++ implementation adapted.

refs #498

Revision 382beb79
Added by J. Wienke over 12 years ago

increment and require higher wire format version

refs #498

Revision 8ffbc8c2
Added by J. Wienke over 12 years ago

increment and require higher wire format version

refs #498

Revision 79da8648
Added by J. Wienke over 12 years ago

adaptions for new protocol version

refs #498

Revision fcff9600
Added by J. Wienke over 12 years ago

adapted to new protocol version

refs #498

History

#1 Updated by J. Wienke over 12 years ago

I thought a bit about this and to me it seems we should completely get rid of the uuids for events. instead we should have one class that encapsulates sequence number amd sender id amd optionally converts to uuid. otherwise i have no efficient way of handling ids for causes without requesting an expensive uuid creation. what do you think about this? Something like eventid for java as it formerly existed would be the solution.

#2 Updated by S. Wrede over 12 years ago

EventId in Java is actually existing since version 0.4 in more or less the way you suggest above. I changed this when working on the participant / sequence number changes for 0.4. Cf.: https://code.cor-lab.org/projects/rsb/repository/entry/trunk/java/core/src/rsb/EventId.java

Internally in the Java implementation I am mostly working with this class as identifier for notifications and not with UUIDs. BTW: I moved most of the UUID-related stuff to UUIDTools.java… (For instance the V5 generation code...)

#3 Updated by J. Wienke over 12 years ago

  • % Done changed from 0 to 50

cpp + protocol adapted. other languages missing right now

#4 Updated by J. Wienke over 12 years ago

  • Assignee changed from J. Wienke to J. Moringen

#5 Updated by J. Wienke over 12 years ago

  • Target version changed from rsb-0.10 to 0.5

#6 Updated by J. Moringen over 12 years ago

  • Status changed from New to In Progress

#7 Updated by J. Moringen over 12 years ago

  • Status changed from In Progress to Closed

Also available in: Atom PDF