QTVCP Bugs?

Good evening, I am working with Phill on QTPlasmac (I mainly break it and tell him so he can fix it ). At any rate, I have noticed a few things that I believe are QTVCP issues at root:

1. Holding shift + other jog axes does not seem to give full rapid as it normally does in other GUIs. It seems to be affected by the "JOG" slider, i.e. higher JOG slider = higher rapid while holding the shift key
2. The preview window shows the yellow plot of jogging the axes during a homing sequence. This results in seeing yellow lines outside of the red bounding box. For reference of "normal" behavior, Axis and Gmoccapy both do not show any yellow lines until after all axes are homed. I would think QTVCP should be similar?

Thanks for all the great work thus far! And please let me know if I can provide any additional information.

cmorley wrote:

1) that's a qtplasac issue, keyboard commands are handler in the screen code.
for instance qtaxis jogs faster (double) if you press shift first. the Code just needs to be added to QtPlasmac

I actually think that's exactly what I am seeing. Shift is probably doubling the jog speed. I am just used to Shift giving full axis velocity as it does in Axis. Good to know

2) I just checked master and don't see this behavior in AXIS or GMoccapy. All three use mostly common code.
Mind you I tested python2 on an older mint system - I'll check on a buster version later.

I do use mint 19.2. I think python2. it's weird. I can get you a video of it if it might help. After the push for the cam stuff goes through that is

Thanks for testing and giving feedback

Be careful what you wish for haha... I'm pretty good at breaking software

It occurred to me that I could back out the changes manually until the push goes through.

Here is what I am talking about with the preview, and homing (yellow lines outside of the red bounding box, and the cone moves during homing). All I've done here is click "HOME ALL", but it shows axis movement during the homing sequence when it really shouldn't.

Of course, on a real machine parked away from the switches, the yellow lines are much longer.

cmorley wrote:

Sorry, what I meant was I see those line on all screens.

Interesting.

So what I've seen is that Axis doesn't move the cone at all during homing as can be seen here:



And QTPlasmaC (QTVCP) does move the cone during homing as can be seen here:


Of course on a real machine, the yellow lines and movements are much longer. On a SIM the homing is obviously really short.

If you need me to get a recording from the actual table I can.

PlasmaC sim, it does show the yellow lines when homing, but pretty sure the machines i have do not as they have tandem axis on Y:

tommylight wrote:

PlasmaC sim, it does show the yellow lines when homing, but pretty sure the machines i have do not as they have tandem axis on Y:

Interesting, that's the difference! Well at least between the Axis SIMs vs what I'm seeing on my Axis with PlasmaC installs (sim and real machine).

Maybe this is something that got "fixed" for QTVCP?

At any rate, would it be possible to "Clear live plot" at the end of the homing sequence? It's a bit odd seeing the yellow outside of the bounding box. Just a suggestion!

Thanks again!

To follow up on this, Phill pointed out to me that this may have started with a recent .ini file change I made to remove the kinstype line from the [KINS] section. He was right.

Removing:
Causes the homing stuff I've been seeing now. Putting it back causes the cone to stay still and no yellow plot lines to appear.

Personally I prefer the latter, as axis movements on the live plot are erroneous (at least on open loop machines) until they're homed. But if nothing else, I wanted to follow up as to what caused the change I noticed.