Re: [Boost-users] [statechart] history not saved

Subject: Re: [Boost-users] [statechart] history not saved
From: Ilya Orlov (iorlov_at_[hidden])
Date: 2012-11-09 04:25:42

On Fri, Nov 9, 2012 at 2:10 AM, Andreas Huber <
ahd6974-spamboostorgtrap_at_[hidden]> wrote:

> Ilya Orlov <iorlov <at> gmail.com> writes:
>
>
>
> > I missed your advice in previous message. I changed transition code to A1
> state and it worked perfectly. But ideally, I'd like to transit from B not
> to
> A1 but to A state. How I can change this code to get it working with
> saving all
> history for A-substates?
>
>
>
> I guess you are missing the fact that the template argument passed to
> deep_history is actually the *default* *state*. This is the state that is
> entered when no history has ever been saved. If history has been saved (as
> is
> the case in your example code) then transit< sc::deep_history < StateA1 >
> >()
> will transit to whatever inner state of A was active before A was exited.
> This
> is explained in more detail here:
>
>
>
> <
> http://www.boost.org/doc/libs/1_52_0/libs/statechart/doc/tutorial.html#History
> >
>

In fact I did not missed this part of tutorial. But since documentation is
almost non-existent in area of history types description, my first attempt
was to set the same default state (A11) in every declaration of A
substates. My idea here was that if I stated deep_history for A substates
and for A itself, all history will be saved and we need default state only
if there was not action inside A state before transition into it. So A
should have only one default state. For me it makes sense. But it did not
worked for me at all.
Next, I defined different default states for each direct A substate. I
still dont understand why I should do it and how it could be that A have
several default substates. At the same time I added transition from B to
one of exact default states. For A1 default state was A11 and I put it into
return transit < sc::deep_history < StateA11 > >(); My thought here was if
I defined A11 as default state, all transactions to A should use exactly
this state. This did not worked too, as I described in my first message.
And next, using your advice I changed transit from A11 to A1, but I still
dont understand why I should do it this way.
As for me, most logical behavior should be like this - if A uses deep
history (I take it as 'total history for all substates'), A should have
single default state for cases like this C -> B -> A. But for cases A -> B
-> A in B -> A transition user should use outermost state A in transition
destination and do not care about how history is saved inside it or about
its default state.

Thanks,
Ilya

>
>
>
> HTH,
>
>
>
> Andreas Huber
>
>
>
>
>
> P.S. In your example you're also specifying sc::deep_history< StateA1 > as
> InnerInitial of A. So you could even transit directly to A, as follows:
>
>
>
> return transit < StateA >();
>
>
>
> and get exactly the same behavior as when you write:
>
>
>
> return transit< sc::deep_history < StateA1 > >();
>
> _______________________________________________
> Boost-users mailing list
> Boost-users_at_[hidden]
> http://lists.boost.org/mailman/listinfo.cgi/boost-users
>