Odd corner case bug with iterations

mikeabuilder

I'm writing a macro to test some new code I'm putting into start.g. The macro is launched from the Jobs tab in DWC. The complete macro is this:

while iterations <10
  m291 S0 P{"Waiting for 10 seconds per iteration. On iteration: "^iterations}
  g4 s10

If I start this "print" job, I see the messages start to print.

The bug happens when I use DWC to pause the job, then cancel the job, then start the job again. When I do this, the variable iterations is not reset when the job starts the second time. Seems like it should happen as part of the cancel process.

o_lampe

@mikeabuilder Can you please add more info about your RRF version? It looks like a bug, especially because AFAIK we can't set 'iterations' to zero ourselfs.

Does it also happen when you use a global variable instead of iterations?

chrishamm

@mikeabuilder I believe I already fixed the underlying bug in the upcoming RRF 3.5-b4. I cannot reproduce this problem on my test setup in standalone or SBC mode with the latest RRF dev build.

mikeabuilder

Sorry for not including my RRF version. I'm on 3.5 rc4. My hw is MB6HC

dc42

@mikeabuilder are you running in standalone mode or with SBC ?

mikeabuilder

Stand-alone mode.

mikeabuilder

Some additional information. I wrote the following in response to @o_lampe's suggestion, though I used a local variable and not a global.

var loops = 0
while var.loops < 10
  m291 S0 P{"Waiting for 10 seconds per iteration. On iteration: "^var.loops}
  g4 s10
  set var.loops = var.loops +1

In this case, the symptom is a little different. The repro steps are

The file is run as a print job, then paused, then cancelled. (Note it runs properly to the end if not cancelled)
The file is run a second time (either Print Again button in DWC Status page, or started from the Jobs page of DWC)

During the second run, the following error is produced:

M32 "0:/gcodes/loop test3.gcode"
File 0:/gcodes/loop test3.gcode selected for printing
Error: in GCode file line 3 column 17: meta command: unknown variable 'loops'
Cancelled printing file 0:/gcodes/loop test3.gcode, print time was 0h 0m

If the file is run a third time, it runs properly. I assume that the RRF error handler properly cleared out whatever was amiss after the Pause/Cancel process.

chrishamm

@mikeabuilder Thanks, I could just identify another issue. We'll see if my upcoming change fixes your problem, too.