From adrian.lawrence@oucs.ox.ac.uk Sun Oct 31 15:49:05 2004
From: A E Lawrence <adrian.lawrence@oucs.ox.ac.uk>
To: occam-com@ukc.ac.uk
Date: Thu, 12 Oct 2000 15:26:10 +0100
Subject: [Fwd: Priority revisited: a new primitive]
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=us-ascii
MIME-Version: 1.0
X-Mailer: Mozilla 4.73 [en] (X11; U; Linux 2.2.17 i586)
Message-ID: <39E5CA02.79F5F6A4@oucs.ox.ac.uk>

> >
> > On the subject of MALT, I think that
> >
> > MALT
> >
> >   c ? x & d ?y
> >     blah
> >
> > is just
> >
> > ALT
> >   c ? x & d ? y
> >     blah
> >
> > if you intrepret "&" as merge. In other words, just omit the M (!).

Gerald replied:-
> 
> Yes, the MALT is the same as the ALT with the event merge symbol (diamond
> operator).
> 
> > I wrote \merge as \diamond in my LaTex syntax for no very good reason,
> > and then used <> to match in ACSII. Either we overload <> or &. I don't
> > think either case leads to any ambiguity. & has the advantage of a
> > single keystroke.
> 
> Doesn't symbol & confuse with conditional guards and with merging events, as
> in,
> 
> ALT
>   cond & event1 & event2
> 
> or is
> 
> ALT
>   cond & event1 <> event2
> 
> different?
> 
> >
> > The point about "merged" events is that they can occur anywhere, not
> > just in the guards of (M)ALTs. And we need them for other reasons,
> > especially in hardware compilation. Which is why I think MALT should be
> > suppressed before it gets a hold. KISS !
> >
> 
> I agree! The MALT was only introduced as another implementation of the ALT
> for multiple guards.
> 
> > I haven't sent this to the list, although I guess the last part of of
> > genearl relevance.
> 
> If you like you can send our exchange of thoughts to the list.
> 
> Gerald.

-- 
Dr A E Lawrence