Runs but get 'No Code' when deployed

[sciencewhiz]

While I have not run NetConsole since the beta, it does work with LabVIEW. Back then you had to specifically ftp the NetConsole.out to the cRIO, but now your install log shows it being installed automatically.

[nathand]

I wonder if this is related to your previous post, Problems with autonomous in deployed vs run mode.  There you diagnosed the problem as a missing link to a type definition, but that's not actually correct.  The real problem in that case is that when you compile your code to the cRIO, the front panel is removed.  Many property nodes (such as the one you were using to get the string value of an enumeration) do not work when there is no front panel.  Instead, they generate an error.  If you're using a property node like that now, and you wire the error out from the property node to other functions, that error will propagate, which could cause the problem you're seeing.

[Jon236]

Finally got the code to deploy.  Involved a nonsensical sequence of disabling vi's, building and deploying, then enabling them and repeating.  But once 'Robot Code' came up, nothing worked and the Driver Station reported 'Watchdog Not Fed".  This with code that worked perfectly when run from Windows.  I have run out of solutions and would really appreciate your assistance.

[nathand]

It would help if you provide your code in a more usable form.  The ZIP from the yousendit link is missing a lot of files (where's your Robot Main?  which autonomous mode?).  It would help if you use LabVIEW to build a ZIP file, which you can do by right-clicking on Build Specifications and choosing New -> Zip File.  Then, upload that zip file directly to this forum, with more specific directions as to which files you're actually using.

Having poked through your code only very briefly, my initial thought is that maybe you have a dequeue element somewhere that's never getting an element enqueued.  Since you don't seem to have wired timeout values anywhere, the program will wait forever, which will eventually cause the watchdog to time out.

[Jon236]

Strange...I just downloaded the link from yousendit to my work computer and all the vi's were there

Anyway,  we enqueue elements in teleop without a timeout since we are monitoring joystick button pushes and dequeue them in Kicker &Ball Magnet with a 0 timeout, to avoid delay.  We put a while loop around the case structure in teleop to facilitate shift-registers....the watchdog is fed outside that loop, but since it is inside the Robot Main while loop, isn't that enough?

[nathand]

Odd, I downloaded it again and all the files showed up.  No idea what happened the first time.

There's too much code there for me to make any guess as to what's happening.  I would recommend, on general principle, slowing down some of your loops (I doubt that anything on the robot needs to run 100 times per second), and using Wait instead of Wait on Millisecond Multiple.  Also, there's no need for a wait inside any while loop that only executes once (loops in which you've wired a constant to the stop condition).

Can you try remotely debugging a real-time application?

[Jon236]

I tried to connect in debug mode, but it would not connect  (although I could ping the cRio).