Re: implicit self

[David.Ungar@Eng.Sun.COM (David Ungar)]

Wow, nice message!



At 10:41 AM 9/13/95, Randy Smith wrote:
>>
>> I wanted to better understand the meaning of sending implicit
>> self messages, so I tried the following:
>>
>>     ( | x <- 4. silly = ( | x <- 2. k: y = (x+y) |
>>                                  k: 0 ) | )
>>
>> I wondered if sending "silly" to this object should return 2 or
>> 4 ( my system returns 2 )
>
>Great question Jecel.
>
>Allow me to paraphrase slightly :
>
>(| x     = 4.
>   silly = (| x = 2.
>              y = (x). |
>              y  ).
>|) silly
>
>Should this be 2 or 4?  I think you could make a coherent story for 2,
>but I think you get a slightly nicer story for returning 4.
>
>So this should be 4 in my book.
>
>Perhaps it seems goofy at first glance, but there must be a distinction
>between these 3 concepts:
>
>        The current receiver.                    (self)
>        Where message lookup starts.             (thisContext)
>        The object holding the looked-up method  (method holder)
>
>(we ran through merging all of these possible pairwise combinations
>during the original Self design, but gave up for lack of energy and
>insight.)
>
>More explicilty why it should be 4.
>
>   (| y     = (x).
>      x     = 4.
>      silly = (| x = 2 |
>                 y  ).
>   |) silly
>
>I hope we all agreee THIS should be 4. And this too...
>
>   (| p*    = (| y = (x)|).
>      x     = 4.
>      silly = (| x = 2 |
>                 y  ).
>   |) silly
>
>..and this...
>
>  (| p*    = (| p* = (| y = (x) |)|).
>      x     = 4.
>      silly = (| x = 2 |
>                 y  ).
>   |) silly
>
>So in the above examples, we have the method holder moving up and down
>the lookup path. But while "silly" is being looked up, and while "y" is
>being looked up, the virtual machine has its self marker on the same
>object. It would be odd to suddenly slip and let the self marker "drop
>down" to the currentContext just becasue the methodHolder is at
>currentContext.
>
>Make sense?
>
>        --Randy
>
>(We used to have local methods, does anyone recall how they evaluated?
>I hope I am not making a liar out of us. Note, since we allow
>"local parents" we can try in today's system
>
> (| x    = 4.
>   silly = (| p* = (| y = (x) |).
>               x = 2 |
>           y ).
> |) silly
>
>
>which does in fact return 4.)
>
>
>

-- Dave