[e-lang] Levine the Genius Language Designer

[Peter Van Roy]

Norman Hardy wrote:
> Very good indeed. I strive for concise expressions of an algorithm  
> and especially the notations in which those can be expressed.
> In my mathematical mode I assume that we should all learn to make  
> such notations familiar to us.
> That is a fallacy; for some, and perhaps most of us the cost/benefit  
> of such learning is negative for most of these many notations.
> I believe in specialization; at least for language exploration and  
> probably production we need specialists that learn new language  
> paradigms deeply.
> We also need literature that informs others of the benefits of these  
> notions so as to guide us in what to learn.
> We need to improve the art of describing such notions; tutorials are  
> good but not enough.
>   
I think that many people have the wrong idea that semantics has to be 
complicated.
This starts with their studies: they either see no semantics (in OO 
courses) or they
see a highly complicated semantics with a lot of theory, but that has no 
practical
relevance.  I've been on a crusade for many years now trying to change 
this.  In my
second-year course, students see a full formal semantics of all the 
concepts in the
course (like functions, objects, classes, exceptions, concurrency with 
streams).  It
takes two lectures to teach it, defining the basic concepts and going 
through
examples.

Peter