Re: [scxml] sixth WD of SCXML is published

On 1/8/2010 10:15 AM, Jim Barnett wrote:
> Greg,
>   You're right about the intended effect of namelist.  It is shorthand
> for something that could also be done with<param>.  (Note that the

All right, and I see the no-@expr form of <param> too. I'm assuming 
these are places where _data is implicit, and if you want anything from 
outside of _data (like _event, _x, etc.) then you must use <param> with 
@expr?

If this is the case, is there anywhere else where _data is implicit 
besides send/@namelist, invoke/@namelist, and param/@name?

Also, I saw the first bullet of section 4.3.4 which explains what gets 
replaced in invoked machines when <param> is used. Is this basically the 
same logic that determines what part of the data model is read in the 
invoking machine (or anywhere else <param> is used)? Would that mean 
that we're not really dealing with location expressions at all here, but 
rather names that are only good for top-level properties of _data? (I 
guess for ECMAScript, this would be like the difference between 
_data.a.b.c and _data["a.b.c"].)

> syntax of the actual message that is sent will depend on the event I/O
> processor, though.  The SCXML Event I/O processor and the BasicHTTP
> processor will produce very different messages from the same 'namelist'.
> Appendices G and H give examples.)

Right, of course.

> The statement that 'namelist' cannot occur with<param>  is an error and
> will be fixed in the next draft.

Got it. Thanks.

-- Greg

Received on Friday, 8 January 2010 21:21:36 UTC