On Jan 5, 2011, at 3:02 AM, Nigel Garvey wrote:

> John Delacour wrote on Wed, 5 Jan 2011 02:21:03 +0000:
> 
>> I've recently switched from 10.4.11 to 10.6.5 and
>> am having various problems, among them this one.
> […]
>> property eudora : "Eudora"
>> 
> […]
>> I can't get or set its properties.  In other words this fails with an error:
>> 
>> tell application "eunotice" to get its eudora
> 
> I can confirm getting the same results (or lack thereof) on my 10.6.5
> machine. However, if eunotice contains a handler to return the property
> value, it can be told to execute it. Similarly with setting.

I have a question that arises from some experimentation I did. We have Macs where I work with three different OSes. Here are the results of attempts to directly and indirectly get a property from a stay-open script applet:

OS Version	Property direct query
--------------------------------------------------------
OS 10.6.6	Error -1728: Can't get [property name]
OS 10.5.8	Error -1750: Scripting component error
OS 10.4.11	Return [property value]

OS Version	Property query via handler call
---------------------------------------------------------
OS 10.6.6	Return [property value as handler result]
OS 10.5.8	Return [property value as handler result]

In all cases, whether an error or the property is returned, the stay-open applet seems to respond only when idle and is unresponsive while executing code[1]. So, is this consistent applet behavior or am I jumping too quickly to a conclusion? If one is using this technique, knowing when to expect a response would be helpful.

Stan C.
_____
[1] The test property is set to 'true' as the applet begins executing and is set to 'false' just before going idle. Repeated queries (hundreds) to the applet through many idle cycles have returned only 'false' and never 'true' values. In the case of errors, the lengths of time for executing and idling were known, so I could approximate when in the cycle an error occurred.