Can a child object link back to the parent?

idbruce

lllllllllllllllll

Heater.

Short answer. No.

If you try to put an OBJ clause in a child that uses the parent object you will get a "circular reference" error.

Long answer, I bet someone here has figured out a way to do it.

My answer. Even if there is a way to do it you probably don't want to, it is likely to be an unfathomable hack:) There must be a better way to structure you code.

Heater.

idbruce:

...or set values within the parent?

Well of course you can pass pointers to data from the parent to the child and in that way the child has a means to access data within the parent.

Dave Hein

You could pass a pointer to a callback routine from the parent to the child.

idbruce

Dave

I would be interested in learning more if you have the time to provide the details.

Bruce

Mike G

You could add the wanted logic to the child application. Then you would have two copies of the code, not good. How about abstracting the logic to make a new object that can be shared between the parent and child.

idbruce

Hello Everyone

I do not have a real definition for my problem yet, but I would like input on this subject.

Bruce

Dave Hein

idbruce wrote: »

I would be interested in learning more if you have the time to provide the details.

The MethodPointer object is located in the OBEX at http://obex.parallax.com/objects/697/ . It was discussed in the thread at http://forums.parallax.com/showthread.php?128397-Callbacks-and-Method-Pointers-in-Spin&highlight=pointer . It uses a 4-word buffer to store the calling information for a method. This includes the starting address of the object, the starting address of the VAR section, the method's starting address and the stack space needed for local variables.

idbruce

Dave

Thanks for that information. I will take a peek at it in the very near future.

Bruce

idbruce

Dave

The link to the thread for this information is a dead link. Can you repair this link please?

Bruce

Dave Hein

I fixed the link in my previous post. It works now.

mynet43

I have a simple method for calling a number of spin routines from an assembly language routine.

I define a single variable in the spin routine that the assembly language can write to.

Then in the spin routine, I have a routine that is periodically called. In that routine, there is a case statement looking for that variable. Depending on the value of the variable, it calls a particular routine and stores the results where needed. When it finishes, it clears the variable.

The assembly routine just waits until the variable is cleared, then resumes operation, just as if it had called the routine.

Please let me know if you would like code examples.

Jim

Dave Hein

Jim,

That technique is used in several objects posted in the OBEX that use multiple cogs, and it works very well. However, Bruce is asking about calling a method in a parent object from a child object in the same cog, which is a different problem. I've encountered this problem in the past, and resolved it by moving the method from the parent object to another child object.

Dave

mynet43

Hi Dave,

Thanks for clarifying that for me.

Has anyone answered your suggestion, about passing a pointer for the parent routine to the child routine and calling it that way?

Or have you tried it? Shouldn't this also work for variables?

It seems like this is almost equivalent to linking back to the parent.

Jim

idbruce

@Jim and @Dave

Here is the situation.

In another post ->http://forums.parallax.com/showthread.php?129118-Looking-for-suggestions-pertaining-to-new-object<- , I have an object that I am working on. If you like you can take a peak at the source code. Anyhow the problem is described as such within the primary post:

As was pointed out to me, although I was aware of the potential problem, if this object were used for highly repititious transmissions, there would be a race condition for getting the available data out and overwriting the available data. In it's current state, the parent object would have to get the data out before a new transmission is received. Even though this will work for me, I believe it would be a much better object for all people if a solution were provided.

The parent creates a child item, the child item creates a cog, the cog updates two variables, the parent needs these updated variable values but must get the information before the cog updates the variables again. That is the problem in a nutshell. However, the nut inside has taken roots, and gone a little deeper. Still thinking about this one.

Bruce

$WMc%

Have you looked at Jonnymac's Spin Zone Article in the Jan 2011 Nuts&Volts?

idbruce

$WMc%

I just got a message pertaining to this article yesterday. Reading it is definitely on my to do list. I am currently studying microsoft serial communication protocol, but I will get to that issue shortly. However, thanks for pointing me in the right direction.

Bruce