12-06-2006 05:59 AM
12-06-2006 07:49 AM
Regarding your question in your other thread, YES I am very interested. see this thread.
http://forums.ni.com/ni/board/message?board.id=170&message.id=112440&query.id=39020#M112440
and the BLOG entry "Ode to a State Diagram" in the LabVIEW Champions BLOG here
http://forums.lavag.org/blog/champions/index.php?
I did not reply to that thred because I did not have a machine with the SDE installed to confirm your observation. I did request a moderator get an AE to confirm your findings. (Note: If there activity on a thread, it is ignored by the AE's. They only look at threads that have gone stale for 48-72 hours).
I believe there should be a new SDE released SOMETIME to correct a bug I found in it (and possibly yours too!).
I have subimited a long laundry list of changes for the SDE and I hope they get implemented.
I have encouraged the use of the SDE in development of the applications my company delivers. At this time I believe we have delivered more than 1000 SDE's to our customers.
I believe part of the lack of interest in the SDE is due to the design philosophy that NI pushes as shown in the their wizards and Express VI's.
The "bottom-up" approach is appropraite for the market they want to focus on.
In our case, we develop applications that are better if not "Expressed" but are developed using a "Top-down" design. THis is were the SDE lets me step into a cube were a 1000 VI application is misbehaving and quickly determine where the issue lies. But now I am repating myself!
Please see the link I posted above for more of my thoughts.
Ben
12-06-2006 07:54 AM
I forgot to ask,
What do you plan to use the single-step mode for?
I experimented with that years ago and decided it was worthless.
If you know how to use that mode, I'd love to learn more!
Ben
12-06-2006 12:41 PM
Hello Ben,
Thanks for your answer. To start with your last question. I have not yet used the single step mode, but thought of using it if I encountered more complex cases in this way: Assume a "regular" State Machine A calling a "Single Step" State Machine B. B may require certain steps in A to be executed before it can transition to the next State. B's output is then the next State in A to transition to.
This is not very well thought through at this point in time, and probably would be better to use an Hierarchical State Machine (aka LabHSM).
As I think I said, if I unlocked my State Machine from the State Editor, then LV did not crash. However, I am not very comfortable with a Development Environment crashing, at least not in version 8 of it 😞 Makes me concerned about NI's testing an QA.
Still, the Norwegian NI representative said he "think he has seen an early version of SDE somewhere".
Do you know anything more about this? (Did not have time to read all in the threads you pointed me to)
I am also trying to do something like this:
Make up a Front Panel simulating the Process with all movements and I/O (A Visual Model - VM). Then interface the State Machine Model (SDM) to this VM. When the Vm is correct and the SDM work, the SDM should also be correct, right?
I am currently doing some very rudementary VM building, using Sliders to represent movements and their position to trigger "Sensors" I/O.
Do you have any experience and tips about this kind of modelling ?
Geir Ove
12-07-2006 09:53 AM
Hi Geir Ove,
Your probing has revealed my weak point.
I have to confess that I am an engineer/physist with virtually no computer science training.
What I know was discovered the hard way or by listening to others.
But don't let me lack of knowledge stop you! Please post this Q to the LAVA forum found at this URL
http://forums.lavag.org/forums.html
The LAVA-ites are more CS oriented and someone there will probably understand the terms you mentioned.
The guy who wrote LabHSM frequents that forum and may reply to your queries.
Trying to help,
Ben