-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathnotes.txt
24 lines (14 loc) · 1.38 KB
/
notes.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
* Introduce CDI
* Try to remove ProviderSubscriber which sets the gravityAcceleration on each step. To connect the scStructure to the gravityModel use a simple CDI approach, and have some kind of CDI event/subscriber approach to set the values on each iteration step.
* Remove the SimVisThread from the scStructure
* The numeric models should implement an interface TimeStep (with methods related to the numeric computation, like timeStep). The SeqSimMod would work with a container of TimeSteps
* Create a ModelInfo type which would be a field in each Model.
* The SimVisThread for Celestia accesses the position and velocity of the S/C bythe manipulator, which manages a map of manipulable attributes. The access could be done polling, using jmx.
----
The rocket is the object Rocky. Rocky contains all the pipes, structures, fluids, etc.
* Use JMX instead of the current commands / @Manipulable approaches to manage the simulation. MXBeans seem OK.
The purpose of OpenSimKit is simulation of the commanding aspect. Maybe it is nice to code the management in JMX and provide a client whose commands invoke JMX functions. Of course, client commands can also be coded with the normal OpenSimKit classes
---
Kernel
- inject timehandler
java -classpath "weld-se-1.1.5.Final.jar;*" org.jboss.weld.environment.se.StartMain ../Rocket-stage-inclined2-engine-off.xml out.log