AnsweredAssumed Answered

Bug with global variables?!?!

Question asked by Archival Jones on Jul 15, 2011
Latest reply on Apr 17, 2013 by Asdf Aawer

Like • Show 1 Like1
Comment • 8

I created two global variables X and Y and added them in a design table using $VALUE@X@Equations and $VALUE@Y@Equations.

When I block up external updates to the design table the design table does not update the global variables(so if the design table says x = 3 but it was x = 2 before then it stays x = 2). If I do not block updates then the equations editor ends up updating the table so that whatever the value was gets pushed to the table.

So I never end up with consisent data.

It only seems to do it to some variables. (In this case X)

I've attached a simple demo project that hopefully someone can varify.

Just select Other configuration and the two diameters should be 1.5 and 1.75 but you'll notice that the inner dimension will stay 1.0 which was the original. If you block table from external updates then nothing will change.

Seems like a pretty obvious bug?!? I'm obviously able to get around it by working with the dimensions directly in the design table but this isn't always the case.

Attachments

Test.SLDPRT.zip79.0 KB

No one else has this question

Outcomes

8 Replies

Denny Metcalf Jul 18, 2011 1:08 PM
Mark Correct
Correct Answer
It seems to work as intended.
Using your model I can add more configurations with different sizes.

In all honesty you really shouldn't be trying to do this with both equations and a design table. Your equations first set the global variables... then the next set of equations pushes values in the sketch based on the previous equations (therefore locking access to the dimension values within the sketch)... then you want the global variables that are DEFINED in your equations to be updated by the design table entries and/or vice versa.

This isn't a bug, it's just convolutedly circular.

Just do it with a design table and be done with it. Even if you trick it to work you're going to take a major performance hit... it's a washer.
Like • Show 0 Likes0
Actions
Alin Vargatu Jul 18, 2011 1:22 PM
Mark Correct
Correct Answer
Archival, you are really pushing the envelope for DT functionality to the limit. You might want to take a look at DriveWorks.
Like • Show 0 Likes0
Actions
- Denny Metcalf @ Alin Vargatu on Jul 18, 2011 1:59 PM
  Mark Correct
  Correct Answer
  Alin, his DT has only two rows and two columns in it. Unless that was sarcasm, I fail to see how this is pushing the functionality to the limit unless it is implied that design tables are utterly useless (in which case I wouldn't be completely opposed to such a statement, lol).
  
  Archival needs to decide whether he wants to drive his washers by configurations or a design table. Equations are nothing but mess here. If he wants global variables to be used elsewhere he should not define the values within the equation manager but instead have them pull values from the model.
  
  Please see attached:
  
  Archival, I removed your reference annotations RD1 and RD2. I renamed the sketch dimensions as OD and ID. I got rid of your equations and set up the global variables X and Y to be defined by the values within the sketch. (I'm assuming you have some reason for wanting global parameters X and Y... or else why go through all of this trouble?)
  
  At this point you have two viable options. You can either use a design table or not use a design table, lol.
  
  If you're going to use a design table, you should have it push values onto OD@Profile@Washer and ID@Profile@Washer. Do not try to control your global variables X and Y with the design table as they are being defined by equations. X and Y will automatically update whenever the sketch profile updates.
  
  If you're not going to use a design table, then just use the built in configure dimension table to supply values to OD and ID. I actually perfer this method as it has less overhead.
  Like • Show 1 Like1
  Actions
  - Archival Jones @ Denny Metcalf on Jul 18, 2011 4:07 PM
    Mark Correct
    Correct Answer
    The problem with your method, at least from my perspective, is that in some cases equations are desired for simplicity. SW's has the ability to drive global variables through DT's and you can block updates(supposedly) to the DT from SW.
    
    You have the realize that the washer is a simplified example. In my real cause
    "X" and "Y" are driving several equations so there is no way to "invert" them. Now I could use some dummy dimensions to drive them but they are no longer global variables and it requires an extra step.
    
    If your saying that I should use either a DT or CT because solidworks is buggy as hell trying to mix and match then you might be right. But if you are saying it because it's wrong then I have to disagree. SW's has built in DT parameters to access equations so they did allow for mixing.
    
    The reason question I'm asking is not for alternatives but if it is a bug. In your first post you say it works as intended.
    
    Does that mean that when you changed the configurations the ID and OD changed? (One didn't stay the same)? I would like to know if my SW installation is buggy(as I've had some other problems), if it's a bug in SW, or if it's intended behavior(I don't think it is since X and Y are setup exactly the same but one behaves differently than the other).
    
    Can you confirm to me that the part I attached(not a modified one) is changing both diameters when you change the config? Specifically if the ID *should* change.
    
    BTW, it shouldn't be circular. The fact that the OD works proves this. The OD and ID are independent but identical in setup. They should behave exactly the same.
    
    You have to realize I'm not driving the sketch diameters in the DT BUT I'm driving the global variables.
    
    The chain goes like
    
    1. DT sets global variables
    2. global variables set equations
    3. equations set sketch dimensions.
    
    There is no circular logic(unless I'm missing something) and SW designed DT's to be able to drive global variables. Again,if you take a look at the setup there should be no difference between and X and Y are setup. They are no dependent on each other in any way. Yet X, IIRC, works as I expect and Y doesn't. If there were circular stuff both should be invalid.
    
    All the test file has is
    
    "X"=1.5
    "Y"=5
    "X@Sketch1" = "X"
    "Y@Sketch1" = "Y"
    
    This is fine!! no circular definitions. No interdependence between X and Y. NOW the design table looks like:
    
    $VALUE@X@Equations $VALUE@Y@Equations $Parent
    Default 1 2
    Other 1.5 1.75
    Another 3 5 Other
    
    (You can ignore Another/$Parent as it was added after and doesn't not affect the problem)
    
    http://help.solidworks.com/2011/English/WhatsNew/AllContent/SolidWorks/Core/Parts/wn_2011/wn2011_Global_Variables.html
    
    So, the DT drives X and Y and X and Y drive OD and ID(or what I have as X@Sketch1 and Y@Sketch1 for some innane reason).
    
    There should be no circular logic.
    Like • Show 0 Likes0
    Actions
    - Denny Metcalf @ Archival Jones on Jul 18, 2011 4:51 PM
      Mark Correct
      Correct Answer
      While it seems like there's no "circular logic" I probably shouldn't have used such a term...
      It works as it's intended to.
      
      I have to ask... what leads you to believe that SolidWorks will evaluate the DT before the equations? It might, it might not. Your logic chain is dependant on it though.
      
      Archival, have you played with the API yet?
      You might want to take a peek at my post here:
      https://forum.solidworks.com/thread/40681
      
      What you're trying to do, even if you trick it to work, is going to have an incredible amount of overhead. Even if you're not familiar with C#/VB, you might want to have a read at the equation manager and see for yourself it's limitations and possibilities.
      
      IEquationMgr Interface
      http://help.solidworks.com/2011/English/api/sldworksapi/SolidWorks.Interop.sldworks~SolidWorks.Interop.sldworks.IEquationMgr.html?id=2743ec8061b64fc7900ee5686445a936#Pg0
      
      IEquationMgr Interface Members
      http://help.solidworks.com/2011/English/api/sldworksapi/SolidWorks.Interop.sldworks~SolidWorks.Interop.sldworks.IEquationMgr_members.html
      
      Anyway, I'll get straight to the point...
      
      Based on the model that you have posted in your opening post, without altering it at all, your equations are overriding your table based on one single configuration.
      
      Can you confirm to me that the part I attached(not a modified one) is changing both diameters when you change the config? Specifically if the ID *should* change. BTW, it shouldn't be circular. The fact that the OD works proves this. The OD and ID are independent but identical in setup. They should behave exactly the same.
      
      No offense, but this is wrong. I'll prove it to you. Go and download the file you posted and then edit the equations to set X to 10 and Y to 20 for equations 0 and 1. Then rebuild and switch configurations. Then open your design table. Only one parameter stays at 10 and that is equation 0 because it's the first one to get evaulated before failing. Switch configurations again and rebuild and look at your design table again. The design table will still have the values it originally had and that is the descrepancy that you are experiancing.
      
      Now I know that this isn't convincing yet so bare with me. Look very carefully at your feature tree and look at how the equations are displayed. If the equation was accepted properly you will see:
      
      "X"=10
      "Y"=1.75
      
      What you actually end up getting is:
      X=10
      "Y"=1.75
      
      Equations are nothing more than a string of text. That is why I suggested that you go and look up how equations are handled. When you program for equations with the API you actually have to be very careful how you insert your double quotation marks or it will fail when SolidWorks trys to parse. That's the reason I've written my own equation manager library to manage the equations in SolidWorks... well... without SolidWorks, lol.
      
      That should answer your question about why one variable updates and one doesn't. It fails to resolve both.
      
      Like • Show 1 Like1
      Actions
      - Archival Jones @ Denny Metcalf on Jul 18, 2011 11:22 PM
        Mark Correct
        Correct Answer
        Nope. Sorry, it's gotta be a bug. I was able to get it to work intermittedly but what it seems like it's doing is not updating the value from the DT every time. I was able to get work as expected for ID and OD(I changed X and Y) for a while by adding a dummy variable.
        
        It is almost surely not an issue with the order of evaluation(although you might be right that a failure could cause a problem Y was updating properly which was after X).
        
        What happen is that whatever value was used as the global variables default value for it was added to the DT gets "stuck" and always ends up being used. You can "unstick" it at times and it works as expected but I can't reporduce it on demand. It required opening the DT and possibly making a change or adding a new column or row. Once you unstick it it works until you do something to stick it.
        
        This is enough to say it is a bug since it works as intended at times, is most likely an "update" problem, and shouldn't really be a "hard problem".
        
        I've attached the new test file. If your interested try getting it to work by playing with the DT and/or global variables. Don't make any structural changes to ID and OD but just screw around. Once you make a change try to change configs and see if you can get them to update properly(rebuild each time you hit change a config).
        
        I can understand your thought process on why it won't update but I don't think that is going on here(at first I thought it might until I was able to get it to "work"). Also realize that for some configs it works right off the bat and only some "stick". SW simply not updating the global variables properly explains it all IMO.
        
        If you don't believe me that it works I can always make a video of it .
        
        As far as the API is concerned: https://forum.solidworks.com/message/235615#235615
        Attachments
        Test.SLDPRT.zip130.4 KB
        Like • Show 0 Likes0
        Actions
        Denny Metcalf @ Archival Jones on Jul 19, 2011 9:41 AM
        Mark Correct
        Correct Answer
        Archival Jones wrote:
        
        What happen is that whatever value was used as the global variables default value for it was added to the DT gets "stuck" and always ends up being used. You can "unstick" it at times and it works as expected but I can't reporduce it on demand. It required opening the DT and possibly making a change or adding a new column or row. Once you unstick it it works until you do something to stick it.
        
        I can reproduce it every single time using your model. I'm not really sure what to tell you at this point except that if you are a capable programmer then you will have the best results just writing your own libraries and programs to get the job done. If you are so certain that you have discovered a bug and it would be best to work with your VAR and supply them the information necessary. I think you will find many more of these types of bugs as you continue to use this program though.
        
        I'll reply to your other thread.
        Like • Show 0 Likes0
        Actions
        Asdf Aawer @ Archival Jones on Apr 17, 2013 11:06 AM
        Mark Correct
        Correct Answer
        I have the same experience and am wondering if you ever had this fixed / found a solution.
        
        I have some global variables whose values I'd like to set with the design table for different configurations. Other equations are dependent on these global variables. It seems the globals don't get updated every time when switching between configurations. The only time they update properly is when I haven't yet closed the equations manager. What the heck.
        Like • Show 0 Likes0
        Actions

Incoming Links

Cannot edit/delete/suppress Global Variables

Bug with global variables?!?!

Attachments

Outcomes

Attachments

Related Content

Recommended Content

Incoming Links