marc@dumbcat.sf.ca.us (Marco S Hyman) (04/28/91)
In article <jls.672698352@rutabaga> jls@rutabaga.Rational.COM (Jim Showalter) writes: > Right. Now reread my example: the existing class hierarchy is delivered > in BINARY. This is the real world: people want to protect their source. Jim, as long as all of us are stacking decks to make our viewpoint come out best let me do another shuffle. Selecting the proper off-the-shelf package is part of the _design_ of the program. Either I buy a binary package that does do what I want it to do or I make a design decision to wrap the package with enough code to give the needed function. Repeat this is a _design_ decision, not a coding decision. Now, to your example. Let's assume that a design decision has been make to use the package provided in binary form for whatever reasons. A wrapper needs to be written. Simple enough in some languages. For C++ it might be something like: class supershape : public shape { private: // whatever public: supershape() {} virtual int getRadius() {return 0;} }; Since using the wrapper and the wrapper interface becomes part of the design the rest of the code will either use specific leaf classes of shape or the base class supershape. Nothing else to modify. Design issue solved once in one place. How would you do it in a language that doesn't support inheritance? > Now, I know from > experience that this is a load of crap...but until the snake oil salesmen > stop selling snake oil, I think it's fair to hold them to their claims, > don't you? No. Holding meaningless people to meaningless claims is a waste of time. Better is to educate others on how meaningless the claims are so they wont waste their time either. > >Is this some kind of contest, where solutions are required not to > >make use of any language features also found in Ada? > > Frankly, yes. since I was responding to Bjarne Stroustrup's claim > that inheritance was a paradigm not well-supported by Ada, in contrast > to C++. So it seems QUITE fair to me to come up with what I consider > simple examples of places where C++ support for inheritance degenerates > into the exact same solution you'd use in a language that doesn't even > claim to HAVE inheritance. Does this seem unfair to you? If so, WHY? Fair. Unfair. Don't care. The world is not fair so I certainly don't expect people to be... However, I see this particular slight of hand. Bjarne points out a problem that can be solved easily in using inheritance and you counter with one that can't. Ok. I'll agree that inheritance is not a panacea and that there are classes of problems that C++ and Ada solve equally well -- or equally bad. Now you tell me how you would solve the original problem in a language that doesn't support inheritance. Subject changes not allowed. Why are you so down on inheritance? It's just another programming tool that can be used for both good an bad. I don't know how it will scale to very large projects, do you? > -- > * "Beyond 100,000 lines of code, you should probably be coding in Ada." * > * - P.G. Plauger, Convener and Secretary of the ANSI C Committee * I think he's off by an order of magnitude. 100 Kloc is easily (and successfully) done in C these days. -- // marc // home: marc@dumbcat.sf.ca.us pacbell!dumbcat!marc // work: marc@ascend.com uunet!aria!marc
jls@rutabaga.Rational.COM (Jim Showalter) (04/30/91)
>Jim, as long as all of us are stacking decks to make our viewpoint come out >best I stand fairly accused, but then so does Bjarne Stroustrup. I started out arguing that Ada was inherently more readable than C. He countered that C++ more naturally expresses inheritance than Ada. I countered that at least one common form of inheritance is poorly-enough supported in C++ that the Ada solution is just as good. Which deck is stacked sort of depends on who's ox is gored, yes? ;-) >Holding meaningless people to meaningless claims is a waste of time. >Better is to educate others on how meaningless the claims are so they wont >waste their time either. Precisely what I'm trying to do. I see this Next Great Wave of momentum toward the wholesale adoption of OO-everything, without, in many cases, any decent discussion of what any of it means and what, if anything, it actually adds. I somehow doubt that it is the advent of, say, polymorphism that will make us all rich and famous and yet, to hear the OO evangelists talk, that's all we need. What we REALLY need are decent software architects, industrial-strength tools, and workable process models. Whether or not a language supports virtual functions is so trivial in comparison it boggles my mind when I hear the claims. >Now you tell me how you would solve the original problem in a >language that doesn't support inheritance. Subject changes not allowed. Same as it was always done in the past: variant record indexed by enumeration type, case statement on the type. > > -- > > * "Beyond 100,000 lines of code, you should probably be coding in Ada." * > > * - P.G. Plauger, Convener and Secretary of the ANSI C Committee * >I think he's off by an order of magnitude. 100 Kloc is easily (and >successfully) done in C these days. 2MSLOC is easily and successfully done in Ada these days. Know any C projects of similar complexity and scope that aren't total clusterfucks? (I know of several that are so screwed up they introduce, a la Brooks, greater than one bug for every bug they "fix".) -- * "Beyond 100,000 lines of code, you should probably be coding in Ada." * * - P.J. Plauger, Convener and Secretary of the ANSI C Committee * * * * The opinions expressed herein are my own. *
dlw@odi.com (Dan Weinreb) (05/01/91)
In article <jls.672984893@rutabaga> jls@rutabaga.Rational.COM (Jim Showalter) writes:
I stand fairly accused, but then so does Bjarne Stroustrup. I started
out arguing that Ada was inherently more readable than C. He countered
that C++ more naturally expresses inheritance than Ada. I countered
that at least one common form of inheritance is poorly-enough supported
in C++ that the Ada solution is just as good. Which deck is stacked
sort of depends on who's ox is gored, yes? ;-)
Not at all. Bjarne's comments were direct and to the point. Your
reply is a non-sequitur. C++ does express inheritance more naturally
than Ada, even if there are places where C++ inheritance doesn't do as
much as it might.
Suppose I say that suspension bridges are a better way to get across
major rivers than are beanbags. You reply that suspension bridges
aren't useful to cross the center of Lake Ontario, showing that they
fail to do some water-crossing jobs that people might like to do.
That in no way negates my statement about suspension bridges.