LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

automating LVDiff or LVCompare

Is there any way of automating the LVDIFF process?

 

We work a lot with classes and other VIs over multiple targets (runninc concurrently) and we're forever having to check files marked as changed by SVN to see if there really WAS a change or if the VI was just re-saved in a new context (in order to deploy to RT for example).

 

I'd love to be able to automate the process of running a check on all of the VIs in my hierarchy versus the current SVN version and revert if no changes reported (no REAL changes that is) and leave marked as changed if changes are detected.

 

Is there such a thing available?  Are tools for such a task available?

 

Shane.

Message 1 of 15
(3,718 Views)

Does separating source from compiled alleviate this issue? If not, I assume you are looking for a way to compare the block diagrams without comparing the underlying binary that may change when the target is switched and VI is saved? Even so, this wouldn't totally fix the issue; someone could always move a VI on the block diagram.

 

0 Kudos
Message 2 of 15
(3,692 Views)

Well moving a VI on the BD IS a change (cosmetic, but a change).

 

I only want to exclude the "VI Recompiled" but nothing else issues.  And seperating source code funnily enough doesn't solvet he ptoblem is conditional disable structures are present.

0 Kudos
Message 3 of 15
(3,682 Views)

Don't even need to work on different targets. Same thing happens with changed typedefs (enum). They require the calling VIs to be resaved, even though there's no change (just recompiled). It would be nice to find those as well.

 

0 Kudos
Message 4 of 15
(3,664 Views)

@Intaris wrote:

Well moving a VI on the BD IS a change (cosmetic, but a change).

 

I only want to exclude the "VI Recompiled" but nothing else issues.  And seperating source code funnily enough doesn't solvet he ptoblem is conditional disable structures are present.


Is this problem causing you to drink and post? Smiley Tongue

0 Kudos
Message 5 of 15
(3,649 Views)

No, that's just what happens when I don't post-edit my posts.  My typing accuracy has gone to hell in the last couple of years.

0 Kudos
Message 6 of 15
(3,635 Views)

There is a private method of the application class to compare VIs using two references.  It might help you out.

0 Kudos
Message 7 of 15
(3,621 Views)

@Darin.K wrote:

There is a private method of the application class to compare VIs using two references.  It might help you out.


 

I was going to dig for something like this to see if it existed, but looks like Darin, the source of all knowledge, beat me to it.

0 Kudos
Message 8 of 15
(3,613 Views)

Also,  just inspecting the Mods.vi property bitwise might be fairly helpful in masking out "Who cares" changes and helping you concentrate on the changes you are concerned with.

 

Getting information on "What the bitfield means?" might involve some under-the-hood information that is pretty difficult to track down although SOMETHING in vi.llb must be doing it somewhere to populate the explain changes strings.


"Should be" isn't "Is" -Jay
0 Kudos
Message 9 of 15
(3,611 Views)

@Darin.K wrote:

There is a private method of the application class to compare VIs using two references.  It might help you out.


That sounds interesting indeed.  I'll try it out in the office tomorrow.

0 Kudos
Message 10 of 15
(3,605 Views)