[OpenAjaxIDE] [OpenAjaxGadgets] Widget constructors - property bag *not* passed

Rich Thompson richt2 at us.ibm.com
Thu Aug 14 06:13:38 PDT 2008

The last time I was involved in a discussion on this topic, the decision 
was to pass the WidgetWrapper instead of the property bag since it 
provided a superset and passing both would be redundant. Did this not make 
it all the way to the spec write-up?


"Howard Weingram" <weingram at tibco.com>
Jon Ferraiolo/Menlo Park/IBM at IBMUS, <ide at openajax.org>, 
<gadgets at openajax.org>
08/13/2008 09:48 PM
Re: [OpenAjaxGadgets] Widget constructors - property bag *not*  passed

I am not sure about this, Jon.
The property bag being passed in could include properties
(such as local security callbacks) that are not intended for 
external advertisement.
Could we please leave this in here for now? Thanks.

From: gadgets-bounces at openajax.org [mailto:gadgets-bounces at openajax.org] 
On Behalf Of Jon Ferraiolo
Sent: Wednesday, August 13, 2008 1:07 PM
To: ide at openajax.org; gadgets at openajax.org
Subject: [OpenAjaxGadgets] Widget constructors - property bag *not* passed

As we worked on the reference implementation, we discovered that there was 
no need to pass a property bag to the widget constructor, as we have 
talked about in the past. The widget can simply call 
widgetwrapper.getPropertyValue() for any property values that it needs, 
and can make the call to getPropertyValue() at the time when the widget 
needs the information. Another approach to getting property values is to 
take advantage of the @@foo@@ macro substitution feature by including 
something like <javascript>var backgroundcolor='@@bgcolor@@';</javascript>

In our reference implementation, we maintain copies of property values on 
both sides (i.e., one copy in the mashup editor and one copy in the widget 
itself), so when a widget calls getPropertyValue(), there is negligible 
performance impact. It seems likely that other implementations would use a 
similar approach.

Does anyone see any problems with removing the preliminary section in the 
spec (i.e., still has red-colored comments) that talks about passing the 
property bag to the widget constructor? To me, this is a vestigial section 
that no longer makes sense. Here is the section of the spec that I propose 
to remove:


If no one objects by the end of the week, I'll remove this section.


gadgets mailing list
gadgets at openajax.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://openajax.org/pipermail/ide/attachments/20080814/12e5d228/attachment.html 

More information about the IDE mailing list