Re: Question about history nodes

That is a good counterpoint.

We decided to dis-allow initial states that aren't direct children for our
purposes to make creating an editing environment simpler.  The initial
transition, if there is one, could be quite involved but we do not allow
most possibilities here from our editing environment (although the
implementation of the runtime environment does allow all these things).


Chris



On Tue, Jun 25, 2013 at 2:46 PM, Jim Barnett <Jim.Barnett@genesyslab.com>wrote:

>  Interesting question.  We discussed this in the group and came across
> the following problem:  what if the parent state’s initial state is not an
> immediate child?  That would be fine as a default for a deep history state,
> but not for a shallow one.  I’m   sure that there would be a way around
> this (maybe we could say that the default history state is the first child
> state), but we decided it was simpler to just make the transition
> mandatory.  ****
>
> ** **
>
> I’d be happy to take this up in a future version of the spec, particularly
> since the change would be backwards compatible (1.0 documents that
> specified a transition would  still be valid once the transition was
> optional.)****
>
> ** **
>
> **-          **Jim****
>
> ** **
>
> *From:* chris nuernberger [mailto:cnuernber@gmail.com]
> *Sent:* Tuesday, June 25, 2013 4:14 PM
> *To:* VBWG Public (www-voice@w3.org)
> *Subject:* Question about history nodes****
>
> ** **
>
> The SCXML specification states that the history node *must* have a valid
> transition for its initial state.  It seems a logical default value for the
> initial transition would be the parent's initial state.  ****
>
> ** **
>
> Given this, is it *really* absolutely necessary to have initial
> transitions on history nodes?****
>
> ** **
>
> ** **
>
> Chris****
>
> ** **
>
> ** **
>
> --
> A foolish consistency is the hobgoblin of little minds - Emerson ****
>



-- 
A foolish consistency is the hobgoblin of little minds - Emerson