Y'ello Pierre!
16 Apr 97, Jerry Coffin had to endure Pierre Phaneuf's ramblings (shown 
below).
 PP>>> There is *many* questionable things about C++... Like for
 PP>>> example, (regarding pure virtual functions) what happens if
 PP>>> I put something else than 0 after the =??? Like in "virtual
 PP>>> void fn() = 42"?
 JC>> Assuming your compiler works correctly, you get an error
 JC>> message.
 PP> Why did they choose to make such a simple thing, obviously crying
 PP> for a simple "flag keyword", such a easy to mess up thing as this
 PP> "= 0" crap?
Come on, Pierre, only a turkey could mess it up.  It's not as though it
changes, it's always = 0.  Stop being a big baby and take it like a man. ;)
 PP> This is the arch-example of multiple
 PP> constructor in Borland Pascal, as this is used throughout the
 PP> streaming mecanism of Turbo Vision.
Don't come the TurboVision stuff with me.  We were forced to use it for a
group project, thinking it would be quicker than developing our own 
nterface.
What a nightmare.  There is no way to escape from the UI.  Everything you do
relies on direct contact with TV, and as soon as you want to make a change, 
you
have to hunt through all your source code.  Nightmare.  Maybe that was just 
us.
 JC>> they're talking about.  I think Booch said it well: MI is a
 JC>> bit like a parachute.  You don't need it very often, but
 JC>> when you do need it, you REALLY need it.
 PP> I have yet to see a case where I'd *REALLY* need it.
But why, Pierre?  It seems perfectly logical that a class could have the
characteristics of two seperate super-classes.  In C++, I think the I/O 
tream
heirarchy looks like:
                     stream
                       |
                +------+------+
                |             |
              istream       ostream
                |             |
             ifstream      ofstream
                |             |
                +------+------+
                       |
                    iostream
Is that right?  I can't remeber if that's it exactly, but, in any case it 
makes
sense that iostream should inherit from both the input and the output 
classes. I
must admit, getting your head around it by the time you come to code is 
awkward.
Java does a sort of multiple inheritance which (forgive me) is a bit easier 
o
use.  A class can EXTEND a base class, but also IMPLEMENT a number of other
abstract base classes.  Clever, because the object IS-A 
but at the same time, must provide implementations for the methods within the
abstract classes which it IMPLEMENTS.  You probably know all that, but I 
hink
it's interesting and perfectly clear to see when it's used.
 PP> I think of it as the GOTO of
 PP> object-oriented programming and makes me puke just like the
 PP> original GOTO... ;-)
Come, come.  :)  Like you, I loathe GOTO, and would always use variable to
terminate a loop, or use an IF the execute relevant code, but MI is an unfair
comparision.  GOTO slaps structured programming in the face whereas MI is a
natural compliment to OOD, consistent with the methodology.
 PP> I remember seeing a framework for a particular version of Object
 PP> Pascal that supported multiple inheritance... It actually
 PP> *relied* on it and positively was the most illogic and barfulous
 PP> piece of design I ever saw in a distributed library.
Yep, that's Pascal for ya.  ;)  Couldn't have phrased it better myself. Ooh.
MLIBTYL ;b
                            Fido = 2:259/33
  --Craigzilla++        Internet = McMoose@sol.co.uk
                             WWW = http://www.taynet.co.uk/users/mcmoose/
--- GEcho 1.00
---------------