Re: An object-oriented network DBMS from relational DBMS point of view

Oops! paul c <toledobythesea_at_oohay.ac> was seen spray-painting on a wall:
> Bob Badour wrote:
>> Alfredo Novoa wrote:
>>> ...
>>> "To instantiate" means to give an example, and it is not an apt for
>>> mathematics term.
>> Actually, "instance" and "instantiate" come from logic, and in
>> logic, instances are values. However, OO abuses these words as badly
>> as it abuses so many others.
>> ...
>
> Just a side comment - I remember reading one of the earlier Java
> specs, around 1996 or so and couldn't counter the impression, as much
> as I wanted to, that the spec was written after the first
> implementation. I wonder what the spec for "Oak" looked like.
>
> Not to disparage Jave, there's much behind it, if not in it, that I
> like but I'm trying to limit myself to three or four sentences these
> days.

Well, CLOS, which is probably still the most complete object implementation around, was built based on specifying the union of all of the major flavours of object systems at the time, notable being LOOPS and MIT Flavors.

If des Rivieres or Kiczales were to make comment on what *they* think "objects" are, I daresay I'd take their comments *very* seriously.

Of course, they were involved back before it had become fashionable to pretend everything's better if it's object oriented...

-- 
output = ("cbbrowne" "_at_" "linuxfinances.info")
http://linuxdatabases.info/info/slony.html
"I think fish is nice, but then I think that rain  is wet, so who am I
to judge?"  -- Ruler of the Universe, HHGTTG