>From mira@ccs.neu.edu Mon Nov  3 10:03:51 1997
>Received: from taboor.ccs.neu.edu (mira@taboor.ccs.neu.edu [129.10.112.115])
>	by amber.ccs.neu.edu (8.8.6/8.7.3) with ESMTP id KAA10319
>	for <lieber@ccs.neu.edu>; Mon, 3 Nov 1997 10:03:50 -0500 (EST)
>From: Mira Mezini <mira@ccs.neu.edu>
>Received: (mira@localhost)
>	by taboor.ccs.neu.edu (8.8.6/8.6.4) id KAA11979
>	for lieber@ccs.neu.edu; Mon, 3 Nov 1997 10:03:50 -0500 (EST)
>Message-Id: <199711031503.KAA11979@taboor.ccs.neu.edu>
>Subject: Re: examples in Rondo style
>To: lieber@ccs.neu.edu (Karl Lieberherr)
>Date: Mon, 3 Nov 1997 10:03:49 -0500 (EST)
>In-Reply-To: <199710312238.RAA10194@stockberg.ccs.neu.edu> from "Karl Lieberherr" at Oct 31, 97 05:38:16 pm
>X-Mailer: ELM [version 2.4 PL23beta2]
>MIME-Version: 1.0
>Content-Type: text/plain; charset=US-ASCII
>Content-Transfer-Encoding: 7bit
>Status: R
>
>> Hi Mira:
>> 
>> does this look good?
>> 
>> strategies
>>  S1 = from Foo to Y
>> 
>> Foo {
>>   // a traversal method called average
>>   void average() do S1;
>> }
>> adjustment Count {
>>   (@ int total_count; @)
>>   init (@ total_count = 0; @)
>>   change (@ total_count++; call_super(); @)
>>   return (@ total @)
>> } modifies Y
>> 
>> adjustment Sum {
>>   (@ int total_sum; @)
>>   init (@ total_sum = 0; @)
>>   (@ total_sum+=value; call_super(); @)
>>   return (@ total @)
>> } modifies Y
>> 
>> adjustment Average {
>>   (@ int avg; @)
>>   change (@ System.out.println( total_sum/total_count );
>> } modifies Y
>> 
>> Foo f = new Foo() modify average with {Count,Sum,Average};
>> f.average();
>> 
>> Alternatively we could say earlier:
>> 
>> void average() do S1 with {Count,Sum,Average};
>> 
>> =================
>
>
>Hi Karl:
>
>yes, this looks very good. It looks pretty much similar to the
>sample code I wrote down as you left on Friday except for the 
>following few modifications I have done below.
>Please compare.
>
>// my modified code 
>
> strategies
>  S1 = from Foo to Y
>
> Foo {
>   // a traversal method called average
>   void average() do S1;
> }
> adjustment Count {
>   (@ int total_count; @)
>   init (@ total_count = 0; @)
>   change (@ total_count++; call_super(); @)
>   return (@ total @)
> } 								(*)
>
> adjustment Sum {
>   (@ int total_sum; @)
>   init (@ total_sum = 0; @)
>   change (@ total_sum+=value; call_super(); @)			(**)
>   return (@ total @)						
> } 								(*)
>
> adjustment Average {
>   (@ int avg; @)
>   change (@ System.out.println( total_sum/total_count );
> } 								(*)
>
>
> Sum modifies Foo;						(*)
> Count modifies Foo;						(*)
> Average connects Sum Count;					(*)
>
> Foo f = new Foo() modify average with Average;			(*)
> f.average();
>
> Alternatively we could say earlier:
>
> void average() do S1 with {Average};				(*)
>  
>//end of modified code
>
>
>What has chanched is the way "modifies" statements are done (*).
>They are separated from the adjustment implementations, assuming 
>in this case that all three adjustments may be reused to mdify
>several different base classes. 
>In addition I have prefered to declare Average as a connection
>of Sum and Count. This simplifies the "modify average with ..."
>statements. When Average gets activated the adjustments it
>connects (Sum and Count) will be activated as weel.
>
>The second modification is (**)
>-- the addition of "change" in Sum, which I guess you simply forgot.
>I liked using this term ("change") for the methods to be attached to a
>traversal.
>
>
>> Here the container example: I got lost ...
>> 
>> How do we refer to variables in other adjustments?
>> 
>> We need to do this together.
>> 
>> 
>> -- Karl
>
>I got lost too. I tried to understand your implementation in Dem/Java,
>but I got kind of confused. I think, it's better to look at it
>together. 
>
>
>-- Mira
>
>
>PS.: Did you get my previous mail? Could you please tell me what they
>sold me instead of Gregor's talk? That's really sad. Associating
>viewgraphs with his voice would have been perfect. 
>
>