Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!lll-winken!aunro!alberta!ubc-cs!uw-beaver!fluke!ssc-vax!dmg
From: dmg@ssc-vax (David M Geary)
Newsgroups: comp.lang.c++
Subject: Re: How to define a List class
Message-ID: <4173@ssc-bee.ssc-vax.UUCP>
Date: 21 Jun 91 17:31:27 GMT
References: <285E7568.4E93@tct.com> <1991Jun19.151546.15605@intellistor.com> <1991Jun20.220432.955@lia>
Sender: news@ssc-vax.UUCP
Reply-To: dmg@ssc-vax.UUCP (David M Geary)
Organization: Boeing Aerospace & Electronics
Lines: 49

In article <1991Jun20.220432.955@lia> jgro@lia.com (Jeremy Grodberg) writes:
]In article <1991Jun19.151546.15605@intellistor.com> wicklund@intellistor.com (Tom Wicklund) writes:
]>In <285E7568.4E93@tct.com> chip@tct.com (Chip Salzenberg) writes:
]>
]>>According to wicklund@intellistor.com (Tom Wicklund):
]>>>I think what's desired (and should be implementable) is a way to
]>>>define a list class using templates and have the compiler be smart
]>>>enough to create a single implementation for the list methods.
]>
]>>That can already be done with templates.  Create a ListBase class with
]>>all the interesting code.  Then create List<class T> template in which
]>>all member functions are inline.  Presto.
]>
]>But the challenge is to do this without requiring inline functions.  A
]>list class shouldn't need to know anything about the objects which are
]>being placed in the list except for size (depending on implementation)
]>and the compiler can easily pass the size to a method if required.
]>Unfortunately it requires better compiler technology than most current
]>compilers.
]
]This is not entirely true.  The list class needs to know the type of
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
]object that is being added and extracted.  Templates allow for 
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

 Why?

]compile-time checking, whereas other methods require run-time checking.
]Yes, it is a little inelegant to duplicate the code, but it is not
]much code to duplicate.  
]
]We aren't using computers with 4 K of core and which execute a few 
]thousand instructions per second anymore.  I'm getting tired of people
]complaining about implementations which waste a few hundred bytes of 
]code without ackowledging the benefit of increased reliability and
]decreased programming time.  In many cases, templates even produce more
]time-efficient code.  Don't get me wrong; I'd love to have a solution
]which provides all the benefits of templates without the extra code,
]but in the mean time, I think the cost is well worth it.
]
  Oh, how true.  This is deja vu - reminds me of those refusing to
  use functions years back, because the overhead of a function call
  was more than a goto...

-- 
|~~~~~~~~~~       David Geary, Boeing Aerospace, Seattle, WA.       ~~~~~~~~~~|
|-----------------------------------------------------------------------------|
|~~~~~~  Seattle:  America's most attractive city... to the *jetstream* ~~~~~~|
|-----------------------------------------------------------------------------|