LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Icon bounds are no longer honored! (2015 bug?)


@Bob_Schor wrote:

How did you do that, Christian?

 


 

It is possible that these icons go way back to before the new style icon editor was introduced.

 

EDIT: Actually, I take that back. the ln(sinh) subVI was built 6-18-2014, so it is relatively new. Probably written in labVIEW 2013 originally)

 

(All pre-2015 LabVIEW versions handled them just fine, the problem occured only in 2015.)

0 Kudos
Message 11 of 26
(1,784 Views)

To the best of my knowledge, there were no icon editor changes in 2015, and I know there were none in the backend C++ code that computes the icon mask.

0 Kudos
Message 12 of 26
(1,718 Views)

Old LabVIEW versions (before the new icon editor) explicitedly required a B/W icon with the desired shape as a closed polygon. If the polygon was open somewhere or the B/W icon was missing, the "mask" for what would be used to determine what was icon area and what was background would end up to be the entire 32*32 pixel area. At some point LabVIEW tried to automatically create a mask based on the lowest available icon bit depth but it still had to have a closed border (of pixels that would coerce to black if used as source for the B/W icon) somewhere. With the new icon editor the icon resource was completely changed and instead of a B/W, 16 color and 256 color icon now consists of multiple layers of 256 color each, with still some masking that is infered from the outline of the combination of all layers.

Rolf Kalbermatter  My Blog
DEMO, Electronic and Mechanical Support department, room 36.LB00.390
Message 13 of 26
(1,707 Views)

Ah, yes, I remember when creating Icons in LabVIEW 7.0, I learned (from sad experience!) to always make a B/W "copy" of my Icon even if I never intended to use it.  Didn't know about the Closed Polygon rule as I always used the Rectangle Tool to outline my Icon ...  When the three "versions" of icons went away, I forgot about this practice (thank you, NI, for improving this fieature!).

 

BS

0 Kudos
Message 14 of 26
(1,685 Views)

Yeah not sure about you but I always just did a select all, copy, then paste into the other two icons.  The BW icon border not being closed is a pretty good guess at what might be causing the issue.

0 Kudos
Message 15 of 26
(1,679 Views)

@AristosQueue (NI) wrote:

To the best of my knowledge, there were no icon editor changes in 2015, and I know there were none in the backend C++ code that computes the icon mask.


Well, "something" changed, because it happens to all my reduced size icons (just found a few more, all are in LabVIEW 2013). Maybe it is not the icon editor or the mask calculation, but they way it is drawn.

 

All the icons show the full whilte background in the upper right corner of the VI, and as soon as I open the icon editor and close it with OK, the mask shrinks. to the valid region. 100% reproducible.

 

Everything I tested so far came from LabVIEW 2013. I just looked at some icon versions that were still 2012, and they don't seem to have the problem. I will investigate what the exact history pattern is that leads to the described issue.

 

It's not a very big problem, because we have a workaround, but having to rattle the icon and re-save means the VI has now new changes, even though the code has not changed. I think it it still worth to investigate.

 

Can anyone else independently confirm this new 2015 problem with their own old code?

0 Kudos
Message 16 of 26
(1,650 Views)

@Bob_Schor wrote:

Ah, yes, I remember when creating Icons in LabVIEW 7.0, I learned (from sad experience!) to always make a B/W "copy" of my Icon even if I never intended to use it.


Yes, I was aware if this. The uppermost icon (B&W) needed to have the desired bounds for it to work, and it always worked the way I used it.

 

But as I said, the ln(sinh) function shown earlier is (1) only a few years old (2) was definitely created in the new icon editor, (3) has a black bounding box, and (4) still shows the problem.

 

0 Kudos
Message 17 of 26
(1,641 Views)

@altenbach wrote:
Can anyone else independently confirm this new 2015 problem with their own old code?

I tried, but can't get the problem you are seeing to occur.  My Laptop (I'm "on the road") doesn't have LabVIEW 2013, but does have 2012 and 2014 (and 2015), and I'm pretty sure that I tried this on my earlier responses to this post.  if noone else can do a 2013/2015 test, I suppose I can Remote Desktop into work and try it there ...

 

BS

0 Kudos
Message 18 of 26
(1,633 Views)

So, after some testing, I think I found the problem!

 

Test 1: I was opening the same VI from old source distributions built in LabVIEW 2011, 2013, 2013, 2014, and 2015.

 

Result 1: Opening them in LabVIEW 2015 ONLY shows the problem with the 2015 source distributions. The older ones look fine in 2015.

 

Test 2: In LabVIEW 2015, I corrected the icons of these subVIs (open icon editor, OK, save VI) and things look good in the development system. Now I built a source distribution. Close LabVIEW completely and open the source distribution.

 

Result 2: The VIs from the source distribution again show the problem!!! Aha!!!!!

Control 2: Closing the source distribution and reopening the saved project shows the icons without the problem. Only the files from the source distribution have the problem!)

 

Preliminary concusion: Building a source distribution in LV2015 corrupts the icon of VIs, adding a full size white background to the icon if the icon is smaller than full sized!

 

 

Message 19 of 26
(1,591 Views)

@altenbach wrote:

 

 

Preliminary concusion: Building a source distribution in LV2015 corrupts the icon of VIs, adding a full size white background to the icon if the icon is smaller than full sized!

 

 


I was unable to see this problem in 2015 building a source distibution.  I may be doing something incorrectly so I attached the project.  I opened View Icon.vi after building a source distribution and the Block Diagram looked the same to me.

Matt J | National Instruments | CLA
0 Kudos
Message 20 of 26
(1,571 Views)