LV 8.0.1 'NOT RESPONDING' and/or 'split' and unusable front panel while scrolling

[FLT CTRL 6]

Hello,

Is anyone aware of any display or front-panel glitches and bugs with LV 8.0.1?  I have been using LV for years, but I have never seen the front panel tear itself (literally) in sections while I use the standard scoll bars at the edges of teh screen, then become completely unusable and go into [NOT RESPONDING] mode (???). In other words, my front panel displays have always moved as one continuous piece when scrolling.  During these recent episodes, I would use the scroll bars and parts of my front panel displays would not move while the other section(s) tore away and scrolled as I commanded.  When this happens, my entire PC would slow down, the CPU usage would rise and a approach 100%, and the VI was unusable.  I have been writing a particular VI for several months, starting in LV 6.1, then LV 8.0 PRO-DEV, and recently upgraded to the LV 8.0.1 'maintenance' release.  The problem that  I describe here began to occur within two weeks of the maintenance upgrade.  The problem seemed to be propagating to my back-up version after I opened each VI in unsuccessful attempts to find a VI that did not appear corrupted.  It seemed as if any VI that I opened was going to be plagued by the bug that I described above and I thought that I was going to lose months of expensive work and my job!!!!.  In a last ditch effort ( and a stroke of pure genius) I started saving some of these coorupt version as 'PREVIOUS' in LV7.1 format.  This seems to have turned the situation around, for now, and saved my VIs from becoming corrupted and completely unusable. 

Please let me know if anyone else has seen this.  Also, please let me know how to avoid this, in the future.  Does this indicate corruption of my LV 8.0.1 PRO-DEV environment or the VI files (or both)?  Does NI have tools to correct these corruptions?

Thank you.

[sarahk]

Hi, FLT CTRL 6.

I haven't found any reference to such a bug, but this definitely sounds like something worthy of investigation. Does the problem occur during development, or is it only when the VI is running? Would you mind attaching a version of the code that causes this crash so I can try to reproduce it? If it is in fact a bug, I'll want to report it to R&D for a fix in a future version.

[FLT CTRL 6]

Thank you, but I cannot attach the code since it is proprietary to my employer.  The glitch occurs before execution of the VI.  When this glitch was noticed, the computer would not execute the VI, most of the time, if any of the time.

[Wendy L]

Hello FLT CTRL 6,

As mentioned by sarahk, I haven't found any reference to this behavior as of yet.  So you mentioned that you cannot post the code that you are currently seeing the behavior in, do you have a test VI or ideas that could help reproduce the problem?

Thanks.

[FLT CTRL 6]

Thank you. I can only point to the following events/ techniques that were somewhat different from my previous experiences/ techniques:

1) Performed 'SAVE AS... ' new file name to these files over several months and several times a day using similar filenames with only one number character different between them

2) Logging long data files to the root C: drive and closing in on less < 5GB HDD remaining (could this be overwriting soem temp files that LV uses to manag graphics???)

3) After saving to LV 7.1 and re-opening with LV 8.0, I get a message that is new to me that explained something about wiring a Boolean constant to a case structure and how LV 7.1 or previous used this to keep a sub-vi in memory (???).  Unfortunately for me, I did not capture the exact syntax of this message and I do not wish to intentionally repeat this graphical feature.

4) I copied a small block of code that contained a 'cloned' sub-vi from one block diagram to the a similar VI that , shortly afterward, exhibited the front-panle 'tearing' feature.  This is something that I usually do not do, and, again, is one of the only new actions from anything that I have done in LV for the past several years.

5) After the message described in (3), above, the rescued LV 7.1 version had moved a handful of the most recently added indicators to a very remote location, about 30 seconds of edge-scrolling time, to the left of my front panel graphics.  I perfromed a 'drag' operation by selecting these and bumping them to the right edge of the screen for about 30 seconds to get them back to the vicinity of my front-panel graphics.  I have never experienced a VI that shifted inidcators so far off the edge of the screen.

These are the only ' unique' events that I can recall before the 'tear; effect.

Thank you.

[JeffBoettcher]

Hello,

Well you've managed to capture (at least) three NI employees' interest / concern with this issue, but I'm afraid that if we can't reproduce it, it will be difficult to diagnose.  Some other things that I'll reiterate or add from the previous posts:

• Do you have any other computers with LabVIEW 8 on them?  If so, can you reproduce the problem there, or does it appear to be specific to a particular machine?
• I understand that your work is confidential, but is there anything that you can do to re-create the problem in a new VI such that the new VI doesn't contain any confidential information, so that you could then post that VI?
• Particularly if the previous suggestion is not an option, is it possible to recreate this problem in a VI without any confidential information on the front panel, and take a screenshot of this?

I also know that some facilities are very strict about what can be posted, but I had to ask again, as without some or all of this it might be hard to diagnose this problem which I'm pretty sure we've never seen or heard of before.

Thank you for your patience and understanding.

[FLT CTRL 6]

Here is the screen shot of the phenomenon that appears in a handful of my LV 8.0.1 VIs.  I do nto have a second LV 8 Pro Dev license to try these VIs.

Thank you

[JonMendez]

Hi there.

I was shown this thread to try to investigate, and I found that I have a VI from a bug report that exhibits similar behavior. Perhaps if we can find a common situation between this VI and yours it will help in solving the problem.

My first guess: Is your front panel set to scale all the objects with the window?

[RWiersma]

I found this thread because I'm seeing nearly the same issue.

I'm using LV 8.01, and the problem exists on two seperate PC's.

Two differences I have between the original post:
-My tear line runs horizontal, and it's not always at the same height (the poster shows a screen shot with a vertical tear).

-My app does NOT lock up. All objects in the "scrolling" section of the tear are usable, all objects in the "dead" zone are unclickable. I can still save, close, etc.. just like normal.



It looks like an unintentional split-screen, but one half of the screen becomes unusable.

Any pop-over windows (say, open Windows Explorer, then close it) will remain visible in the "dead" zone. Almost like screen isn't refreshing in that region.

Block diagram works fine.

I'm not scaling object to fit screen size.

Problem only exists in top-level VI, any/all sub vi's work properly.

I have the Vision module loaded and in use in this VI.

Problem does not always exist. I can close LV, reopen...work for a period of time, and at some point it starts 'tearing'.

The tear exists regardless of whether the VI is running or not.

My email address is rwiersma@tpprobes.com

Please advise.

I'd be able to provide a neutered version of the VI to an appropriate NI support person. (There are -tons- of sub VI's, I need help with the logistics of sending such an involved program).

[RWiersma]

Update:

Loaded 8.2, loaded the same VI, problem STILL exists.

Please advise.

-Ryan
rwiersma@tpprobes.com