Re: N. Wirth

"David Fetter" <david_at_fetter.org> wrote in message news:zr-dnYoU3uSBp8XZnZ2dnUVZ_vGdnZ2d_at_speakeasy.net...
> Jon Heggland <jon.heggland_at_idi.ntnu.no> wrote:
> > David Cressey wrote:
> >> I'll repeat a comment I made a little while ago: Someone should do
> >> for OOP what Pascal did for structured programming. Come up with a
> >> suitable language for use as a teaching tool.
> >>
> >> Someone should do the same for the interface language to a
> >> relational database.
> >
> > Something like Tutorial D, you mean?
>
> I sure hope not. Tutorial D is meant to implement the Date's
> Relational Model, which is equivalent to first-order logic, which in
> turn is provably less powerful than SQL. Why would anybody want to
> bother with a language that's both cumbersome and weaker than what
> people will be working with?

For pedagogical purposes, one might start with a language that's weaker than the one a professional is working with.

An important consideration in the learning experience is to avoid language features that the beginner has to be alerted to, but that can't be reasonably explained to the beginner, yet. Nicklaus Wirth, in the Pascal Report made reference to the omission, usually deliberate, of many of the favorite features of programmers like myself.

The result was a language that can be used to teach programming, as opposed to teaching the idiosyncrasies of that language.

Certain things like preventing missing data or sidestepping change management, which might be fatal in a professional environment. might actually be favorable in a learning environment.

I've said elsewhere that I don't know Tutorial D, so I don't know if the above comments apply to it. My remarks are directed to learning languages in general. Received on Wed May 03 2006 - 13:15:57 CEST