resetting cursor lists programmatically

[Wonger]

Definitely should not happen from updating the cursor list programatically... I will investigate.

[Wonger]

I can't reproduce getting into this state from a fresh VI... there must be something more to reproducing it than just updating the cursor list programatically.

If you have steps to get into this state from a new graph please let me know!

[X.]

Working on that, but I don't want to damage my current VI.

Try the attached one (LV 12), there is anothergoodie to be found.

Run the VI and create a cursor (I use Single Plot but that doesn't matter).

Store the cursor list

Restore the cursor list: watch how the cursor coordinates have disappeared. If you move the cursor around, there is no coordinate display...

[Wonger]

So the bug here is actually that you can ever see the coordinates on the single plot cursor which is not locked to a plot. You are not supposed to be able to see coordinates unless you are snapped or locked to a plot, the fact that you can right after creation is a bug.

I could not reproduce this using free cursors, they should always show the coordinates.

[X.]

Not sure why you would not have coordinates for cursor supposedly attached to a plot but in practice not attached to any. A cursor is a cursor and it has by definition coordinates...

Regarding the Cursor List, I have concluded that it has to be kept at a safe distance and never approached under penalty of random bug strike.

[Wonger]

Well the main reason is that a single plot's cursor gets it's scale from the plot it's attached to. You can't specify it manually like you can with the free cursor. This works fine if there is only one scale but once you start adding multiple scales it would be unclear which scale the coordinates are referring to.

[Wonger]

In lieu of a way to reproduce the state I have implemented a change in 2013 which will fix it when it is detected (when you attempt to access one of the "fake" cursors, for example by popping up on one). This repopulates the legend from the graph's private data, in this case there is no cursor data so it clears all the "fake" cursors from the legend.

Hopefully this will make it more obvious when it happens and make it easier to find the root cause, as well as getting the graph back into a good state.

[X.]

Just in case I can interest you in having a look at this other annoying (lack of) feature of yet another component of graphs (namely, the plot legend)...

[MG70]

Hello!

I am trying to find the proper way by which cursors should be programmatically removed in LV 2012. Despite this thread is pretty old and probably the mentioned bugs should be fixed, it seems to me that the problem reported at the beginning of the thread is still alive.

I programmatically create a list of cursors to highlight spikes on a waveform graph (See cursors_created.jpg).

Then I cleane them all, except for the Threshold cursor. I remove all entries from the cursor list except the first one and rewrite the property.  (see CleanUp_cursors_Vi.png).

The result I get is visible in Cursors_cleaned.png. What I wish to have is only the Threshold cursor visible in the Cursors area, without the coordinates of the deleted ones. 

Can anyone advise about how to fix this issue? The workaround proposed by Paul Davidson doesn't work for me.

Thanks a lot!

MG

[Ben_Co]

I've been trying to recreate this issue, but I haven't had any success. After looking into some bug reports we have, it seems like our R&D team has also had some issue reproducing it as well. It seems to be caused by corrupted cursors, but you could try creating a new graph and programmatically adding the cursors. I'll add the information to the bug reports that we have.