Re: Ichibah flames, and flames out over, Ada 9X

[jgg@evb.com (John Goodsen)]

wellerd@ajpo.sei.cmu.edu (David Weller) writes:

>Tucker would be quick to point out that the word "record" is also
>an implementation detail, but that it's become generally accepted
>in the programming domain.  

This is irrelevant to the class vs. tagged type discussion.
Or is it?  "class" is generally accepted in the OO programming
domain.  Tagged Type ?  really not even known in the OO programming
domain.


>Frankly, I don't care for the word
>"tagged" either, but "class" doesn't fit in its place.  
>
By what measurement?  C++ syntax? Again, if you want to make
an impact against markets like the C++ market, you need to at
least speak the language.  Change the semantics for
technical superiority, but you'll never get the OO message out
using terminology that the OO community doesn't (for the most
part) understand, namely "tagged types".  



>One other thing that's bothered me -- why are we raising such a fuss
>about the word "tagged" to mean "this record may have it's state
>description extended"?  Perhaps we're being pedantic?  

The whole argument for class vs. tagged type is based solely upon
the need for Ada 9X to gain support as an OO language.  This is
not pedantic.  It is realistic awareness of the programming community
and how to grow a business centered on Ada technology.


-- 
John Goodsen
Software Process & Environments
EVB Software Engineering
jgg@evb.com