RE: [cgmo-webcgm] request for new style propertity

["Bezaire, Benoit" <bbezaire@ptc.com>]

Hi,

 

I have a hard time understanding why a new style property is needed. Maybe I'm missing something, but a 'state' property sounds like metadata to me.

 

Why can't the state of the objects be kept in JS variables; or alternatively into namespace attributes set/getAttributeNS (on WebCGMNode interface)?

 

Benoit

---

From: Weidenbrueck, Dieter [mailto:dweidenbrueck@ptc.com]
Sent: Wednesday, October 08, 2008 1:17 PM
To: Lofton Henderson; Forrest Carpenter; WebCGM
Subject: RE: [cgmo-webcgm] request for new style propertity

All,

 

quite honestly, I don't understand Forrest's request.

 

"I would like to propose an additional style property for WebCGM 2.1. This would be a generic text string that could be used to define the state of an application structure. Values might be 5 volts , 30 psi , switch position 3 , etc. In my example I have used state as the name of the style property, but some other name like comment would also work."

This sounds like a generic APSATTR of type string. However, Forrest apparently suggests to use this attribute to describe a "state" listing several examples. These examples suggest an implied semantics that can not be understood without ambiguity.

 

Example:

an APSATTR with the name "Voltage" could carry a string or numeric value of "5". The semantic of this attribute would have to be clearly defined so that an interpreter would know what to do with it. However, as there is no built-in logic for such a semantic, e.g. "hide an object and show another one depending on this value",  it would be difficult to standardize the viewer behavior.

But in this case, it is even worse, there is not even a clearly to be understood value, it may be anything.

So there is no way

- for the interpreter to understand the semantics

- for the interpreter to act accordingly

 

Now one could say, why not treat it as an arbitrary text attribute?

The reason is that then we're back to the old times, where Forrest would probably carry a voltage around for an interactive wiring diagram, Don might put some seismic value into it, and we might do something completely different with this attribute.

This would put us right into a situation, where only one application would understand it and do the right thing, whereas a standards-compliant viewer would have no chance to handle the intended functionality accordingly.

 

Therefore, I object to this, unless I have completely misunderstood the intention.

 

Regards,

Dieter

---

From: Lofton Henderson [mailto:lofton@rockynet.com]
Sent: Mittwoch, 8. Oktober 2008 18:20
To: Forrest Carpenter; 'WebCGM'
Subject: Re: [cgmo-webcgm] request for new style propertity

Stuart -- please put this on the agenda for discussion.

All -- 

(Email discussion is also welcome, of course.)

Per the principles 
that I expressed in a just-sent message [1], I think a wise approach for 
comments like this is:  TC reach consensus about Forrest's comment;  
then, Forrest send his comment (possibly revised) to the W3C public list for the 
Last Call review in the WebCGM WG.

Of course Forrest is free to process 
his comment in any way he pleases, since the spec is now under WG control, but 
some up-front coordination might help expedite the completion of the OASIS-W3C 
collaboration on 2.1.  (See [1] for further explanation.)

[1] http://lists.oasis-open.org/archives/cgmo-webcgm/200810/msg00007.html

Regards,
-Lofton.

At 
12:06 PM 10/7/2008 -0500, Forrest Carpenter wrote:

> All,
>
>  
>
> I would like to propose an additional style property for 
>   WebCGM 2.1. This would be a generic text string that could be used to define 
>   the state of an application structure. Values might be 5 volts , 30 psi , 
>   switch position 3 , etc. In my example I have used state as the name of the 
>   style property, but some other name like comment would also 
>   work.
>
>  
>
> Regards,
>
> Forrest
>
> ---------------------------------------------------------------------
> To 
>   unsubscribe from this mail list, you must leave the OASIS TC that
> generates 
>   this mail.  Follow this link to all your TCs in OASIS at:
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php