Subject: Re: SDL-News: Gates and Block Substructure (fwd)
From: Jose Marķa Parra Sageras (jospar#rest.tel.uva.es)
Date: Tue Nov 15 2005 - 18:13:17 GMT
First of all, I want to thank Andreas Prinz for his quick answer.
---------- Forwarded message ----------
Date: Tue, 15 Nov 2005 08:25:50 +0100
From: Andreas Prinz <andreas.prinz#hia.no>
To: Jose Marķa Parra Sageras <jospar#rest.tel.uva.es>
Cc: sdlnews#sdl-forum.org
Subject: Re: SDL-news: Gates and Block Substructure
Become an SDL Forum Society member
<http://www.sdl-forum.org/Society/members.htm>
The originator of this message is responsible for its content.
-----From Andreas Prinz <andreas.prinz#hia.no> to sdlnews -----
Dear Jose Marķa Parra Sageras,
you wrote:
>> I am developing a program (in SDL-92) that requires a substructure
>>inside the
>> definition of a Block Type. I have a doubt about the use of gates in
>>this
>> case.
>Please note that in general it is not recommended to use substructure any
>more.
>In SDL'2000 this concept is even deleted. It is far better to use the
>object-oriented constructs, e.g. inheritance and redefinition.
I know that with SDL'2000 this matters have been changed but the latest
SDL version supported by the tool I am currently using is SDL-96 so I
have no alternative.
>> In my particular programming context, there is a Block Type definition
>>that
>> only contents a block substructure. Inside the substructure there are
>>several
>> channels connecting some blocks with the substructure environment. I
>>have
>> defined the gates I need in the block type boundary (I cannot do this
>>in the
>> block substructure) and I have specified the name of the appropriate
>>gate in
>> the substructure boundary (as I am supposed to do) in order to connect
>>each
>> channel with the block type environment but it does not work.
>The basic idea behind substructure is that you have *two* definitions of
>the
>same thing. This means, you write the first way to do the things using an
>ordinary block type (This one must not be empty!). Then you add a
>substructure
>definition and put the other alternative in there.
>Please note that the substructure defines a static alternative. This
>means you
>have to decide which version to use when you simulate it.
>I suppose that you start simulation as is, which means your top-level
>version
>is selected (the one with nothing in it). This again means that nothing
>happens.
I agree with you that one idea behind substructures is that you have two
definitions of the same thing. However another option is a block type
specification containing only a block substructure specification.
>> Is that all correct or it is needed any other element? I have look some
>>book
>> but noone tell anything apart from what I am doing.
>So the first thing to do is to make sure that your spec makes sense even
>without the substructure. The substructure is then an alternative to
>this, and
>the spec should make sense when replacing the block type definition by
>the
>substructure.
I see what you mean but taking into account my specs, I need to define a
group of blocks inside a block. I use a block substructure as a way to
achieve this. So I cannot specify anything at the top-level version of the
block type because it is not permmited defining a block in the upper level
block (in SDL-96). I have considered other implementations but there is no
way for them. I need at least three blocks defined inside a block.
Perhaps there are some other posibilities for obtaining my objetives but I
am not considering them.
>
>Please note that it is a matter of the tool how you select the version
>that
>should be used for simulation. This is not defined within the language
>itself.
>
>Best regards,
> Andreas
-- Andreas Prinz --End text from Andreas Prinz <andreas.prinz#hia.no> to sdlnews --- For extra SDL Forum Society benefits join at <http://www.sdl-forum.org/Society/members.htm>--End text from =?ISO-8859-1?Q?Jose_Mar=EDa_Parra_Sageras?= <jospar#rest.tel.uva.es> to sdlnews --- For extra SDL Forum Society benefits join at <http://www.sdl-forum.org/Society/members.htm>Yours faithfully,
Jose M. Parra Sageras
This archive was generated by hypermail 2a23 : Thu May 09 2013 - 16:05:50 GMT