Re: [www-vrml] Re: [Vrspace-dev] Cortona createVrmlFromURL bug test code

[Rob Meyers <rob@vrspace.org>]

> Second, it's a bad (automatic?) translation from Russian.  That's what
> I've wrote: "I can not see how it can break program logic, taking into
> account that createVrmlFromUrl() completion time and callbacks calling
> order are non-deterministic."

      The fact that createVrmlFromUrl completion time and callbacks are 
non-deterministic
is precisely the reason why you want a callback for each call.  How else 
are you going to
know if, how, and when the operation completed?


> The problem arises when several calls to createVrmlFromUrl() where
> completed in the same frame (see. ISO/IEC 14772-1, 3.33 frame).
> Cortona then sends multiple addChildren events to the receiving node.
> Changes in the children field can fire one or more children_changed
> events, but only one is sent out (see the last paragraph of `4.10.3
> Execution model' and `4.12.9.3 Sending eventOuts').
>
> > > Hah, hah, hah,  well you see, since the standard is vague we decided to
> As you can see, the standard is absolutely clear. You just have to
> become accustomed to read it.

     I am not so familiar with the standard, so I have read through the 
sections that you
have highlighted.  I'm not quite sure I see how some of them apply, but I 
did see this statement:

4.10.3:
Nodes that contain eventOuts or exposedFields shall produce at most one 
event per timestamp

This is what you are arguing?  If you complete two different 
createVrmlFromURL calls and the two
associated MFNode events are received in the same timestamp/frame, then 
only one children_changed
event is produced as a result of this?  Where does it say in the standard 
that you cannot coordinate your
completions of the createVrmlFromUrl calls so that you send IN only one 
event per timestamp?  i.e.
onCreateVrmlFromUrl completes:
     if ( already sent a createVrmlFromUrl node to target event in this 
frame ) {
         wait frame
     } else {
          send nodes
     }
Contact apparently does this, I think its a much better handling of the spec...


> > > choose a retarded method of implementing it, so therefore, its not a bug,
> > > its a feature!!!   No competent software designer would consider random
> > > callbacks to an asynchronous api call acceptable.   Come on now...
>
> Well, explain me, as one competent software designer to other
> competent software designer what for do you use callbacks if their
> count is so important?

       Again, its an asynchronous call in the first place, so you want to 
know when it
completes.  But here are some examples of why you would want to know:
a)  Oops, it didn't work, maybe I should try to get the object again
b)  Uh oh, I have tried n times and still no luck, I'll give up now.
c)  Great, I got the object, now I can commence with initializing it.  i.e. 
this
is the vrml code associated to a shared object.  Now that it has been 
created, I want
to
i) Start listening to its events out so I can propagate them across the net.
ii) Send in its initial events and tell it to start functioning.

> And at last a simple example of events handling in different browsers.
> Load the file in a browser, open a console, and click a sphere. You've
> got one message in Cortona and Cosmo. Guess, how much do you get in
> Contact?
>
> #VRML V2.0 utf8
>
> DEF G Group {
>  children [
>   Shape {
>    geometry Sphere      {
>    }
>   }
>  ]
> }
>
> DEF TS TouchSensor {
> }
>
> DEF S Script {
>  eventIn SFTime touchTime1
>  eventIn SFTime touchTime2
>
>  eventOut MFNode addChildren1
>  eventOut MFNode addChildren2
>
>  eventIn MFNode children_changed
>
>  url "javascript:
>  function touchTime1()
>  {
>   addChildren1 = new MFNode(new SFNode('Shape { geometry Cone {}}'));
>  }
>  function touchTime2()
>  {
>   addChildren2 = new MFNode(new SFNode('Shape { geometry Box { size 1 5 
> 1}}'));
>  }
>  function children_changed(v, t)
>  {
>   trace('G.children_changed: ' + v.length);
>   trace('time: ' + t);
>  }
>  "
> }
>
> ROUTE TS.touchTime TO S.touchTime1
> ROUTE TS.touchTime TO S.touchTime2
>
> ROUTE S.addChildren1 TO G.addChildren
> ROUTE S.addChildren2 TO G.addChildren
>
> ROUTE G.children_changed TO S.children_changed
>
> #END OF VRML
>
> Dmitry Egorenkov
> Cortona Team Leader
> ParallelGraphics
>
>
> -------------------------------------------------------
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from real users.
> Discover which products truly live up to the hype. Start reading now.
> http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
> _______________________________________________
> Vrspace-dev mailing list
> Vrspace-dev@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/vrspace-dev

-------------------------------------------------------------------------
for list subscription/unsubscription,
go to http://www.web3d.org/cgi-bin/public_list_signup/lwgate/listsavail.html