Bug with radio buttons?

but if it fails to produce the goods then do the ganging directly

As already noted it is usually not obvious that the bug is visible since it changes a variable after the usual interaction (window with radio buttons is open) happens.

But I take that there's no interested in fixing this bug.

I have already had a look at this and I found that there would be a great deal of work needed to fix this problem. Maybe even a redesign of %ga.

Given the ease of doing things differently I have decided that further work in this are would be too expensive.

I hope that we can now draw a line under this issue.

Thanks

I have already had a look at this and I found that there would be a great deal of work needed to fix this problem. Maybe even a redesign of %ga.

The code snippet I've posted above does NOT use %ga at all but still shows the bug. When using callbacks to simulate the %ga effect, the same buggy behaviour when closing the window appears, the callback function of the first radio button is (erroneously) called.

I hope that we can now draw a line under this issue.

I won't comment on this sentence.

My apologies. I thought that you were asking for a problem with %ga to be fixed.

I can see the problem and can identify the point of failure in ClearWin+. A fix may take a little while.

Thank you very much, your help is appreciated!

I have had a look at this and I cannot find a way to block the message that is causing the problem without causing other problems elsewhere.

The best that I can do at the moment is to provide an option [isolate] on %ww that will fix the problem locally. For the next release the following code will work correctly with this option.

winapp 
Program Test_Radio_Button 
integer :: iw,winio@,ipos,jpos,lw_handle,irb 
irb = 0
iw = winio@ ('%lw&',lw_handle) 
iw = winio@ ('%`ca[Test]&') 
iw = winio@ ('%rb[RB1]%nl&',irb) 
iw = winio@ ('%gp',ipos,jpos) 

iw = winio@ ('%ww[no_sysmenu,isolate]&') 
iw = winio@ ('%`ca[Window]&') 
iw = winio@ ('%sp&',ipos+20,jpos+20) 
iw = winio@ ('%4bt[Close]') 

end 

It does not solve that the problem might not be immediately noticed, but thank you anyways for this workaround!