What factors can cause a Proj file to open Dirty (LV 2009) ?

[Neil.Pate]

---

tst wrote: 

I do have a vague recollection of something about the project changing the order of items in the list,

---

NI has a race condition in their LabVIEW IDE code due to overuse of local variables.

This is primarily the reason they will never open-source it, it would be too embarassing

[George_P_Burdell]

Ben-

Thank you for posting the SR number.  I have forwarded all the responses from this forum to the owner of that SR.  He is in the process of looking into it. 

[Ben]

---

GeorMike S wrote:
Ben-

Thank you for posting the SR number.  I have forwarded all the responses from this forum to the owner of that SR.  He is in the process of looking into it. 

---

Thank you Mike S!

I was getting  the standard line "... If you narrow it down, let me know..." and that made me think I was doing something wrong. Judging by the number of others that have noticed this behaviour, there must be a common cause.

In my case

LVOOP yes

DAQ - yes

TDMS -yes

CW 3D graph -yes

RT -no

Auto-poulate -no

Ben

[Stephen_S.]

Hello,

I was the one talking with Ben about this yesterday. I have been looking into it, and this behavior has been seen to occur in different situations, and it is something that R&D is aware of and investigating. Since it looked like the behavior Ben was seeing is different from the issues that have been recorded already (which have included FPGA VIs, Real-Time VIs, etc.), I filed a corrective action request on the issue. This was reported to R&D (#207809) for further investigation.

Regards,

Stephen S.

[Ben]

---

Stephen_S. wrote:

...

This was reported to R&D (#207809) for further investigation.

 ...

---

Thanks Stephen.

I can track it now.

Ben

[LV_Pro]

Ben,

   It seems to be pretty common, I've seen on my laptop running mostly v2009, on my customer's machine also v2009 and my other customer's systems running v8.6.1. Nothing is being changed, no Xcontrols, etc. I thought that someone asked this last fall and got "sort of" answer, but since, with my PFH (described in excruciating detail in other threads) I have gone from dreaming about LabVIEW (yes, that much of a nerd) to having nightmares, with LabVIEW as a component. So, my reality is becoming blurred, at least with LabVIEW   I've just clicked save, if I haven't done anything and know that no one else has touched the code. It is a bit disconcerting, but that is my constant state these days!

Well it was longer ago than last fall, and the other poster used the phrase "dirty" 

Message Edited by LV_Pro on 02-12-2010 02:42 PM

[Ben]

---

LV_Pro wrote:

Ben,

   It seems to be pretty common, I've seen on my laptop running mostly v2009, on my customer's machine also v2009 and my other customer's systems running v8.6.1. Nothing is being changed, no Xcontrols, etc. I thought that someone asked this last fall and got "sort of" answer, but since, with my PFH (described in excruciating detail in other threads) I have gone from dreaming about LabVIEW (yes, that much of a nerd) to having nightmares, with LabVIEW as a component. So, my reality is becoming blurred, at least with LabVIEW   I've just clicked save, if I haven't done anything and know that no one else has touched the code. It is a bit disconcerting, but that is my constant state these days!

---

Well lets see if we can get this one fixed so you have one less stress factor to deal with.

Ben

[RavensFan]

For my situation:

LVOOP no

DAQ - no

TDMS -no

CW 3D graph -no

RT -no

Auto-populate -no

VISA to interface with serial device - YES

 1 dynamically called subVI, about 56 of my own subVI's in the dependency list, along with assorted VI.lib and OpenG dependencies.

I was trying to keep a closer eye on this dot when working in my project today.  One thing I saw was that if I modified a VI such as adding or removing a subVI, the dependency list changed and that caused the dirty dot to appear.  If I do a change in a VI from outside the project, save it, then later open up the project, the dependency list automatically changed and the project becomes "dirty" right away.

This makes sense, but I don't know if this is expected behavior or not.  I can understand that the project has changed because I have essentially changed the VI hierarchy by modifying some subVI's that added or subtracted some dependencies.  But I would also argue that I didn't actively change anything in the project explorer itself (add libraries, change properties, explicitly add subVI's into the tree) so I don't want the .lvproj file to change.

This leads me to the question, if the dependencies are automatically updated when the hierarchy is loaded when you open the project file, then why is the list of dependencies even stored in the .lvproj file?

(PS.  This is just one thing that I've found that seems to explain my situation at this moment.  There may be some other explanations that could be equally valid as well.)

Message Edited by Ravens Fan on 02-12-2010 07:34 PM

[Ben]

---

Ravens Fan wrote:

For my situation:

LVOOP no

DAQ - no

TDMS -no

CW 3D graph -no

RT -no

Auto-populate -no

VISA to interface with serial device - YES

 1 dynamically called subVI, about 56 of my own subVI's in the dependency list, along with assorted VI.lib and OpenG dependencies.

...Message Edited by Ravens Fan on 02-12-2010 07:34 PM

---

We have an over-lap on VISA

Most of the Classes I am using for I/O (Serial DAQ-AI, DAQ-Counters) use templates to realize servers for each sub-system.

I keep my dependancies pretty clean but I can't claim it clean at the moment.

I am only seeing this on proj files that came from an earlier version (so far).

Makes me think there is some mutaion code in LV 2009 that is not finishing the job when we save.

Ben

[MichaelAivaliotis]

I see this on a daily basis across numerous projects. Very repeatable when it happens. I'm currently seeing this in a project that has no hardware and no classes. I tend to agree with Ben that it may be due to an upgrade of older code. I would love to see this annoying bug fixed.

Message Edited by Michael Aivaliotis on 02-15-2010 09:41 PM