Re: COMMENTS on Tentative Design Summary v0.93

[Radu Grosu <grosu@informatik.tu-muenchen.de>]

Please find below my comments to V0.93.

Best regards,  radu

Name: 
     What about `CoFI Light'? Light should be understood as - light 
     algebraic Language. 

Subsorting:
     In explainig equivalence of terms, spend a few more words about what
     it means to `adjust cast injections appropriately'.

Compound Identifiers: 
     I think, they simulate quite nice polymorphic sorts as they appear in
     functional languages.

Cyclic sections: 
     I liked the way the semantics of cyclic sections was given.

Parameterized specifications:
     It seems that in the current version, one cannot use the same 
     specification twice as parameter. For example, given a partial order
     specification PO with principal sort Elem, one might want to write 
     something similar to the following OBJ code:

     obj example[X::PO, Y::PO] is
      
       f: Elem.X -> Elem.Y
       ...

     Moreover, it is nothing said about the semantical relationship between
     the formal and the actual parameters and how does this relate to 
     subsorting.

Architectural specifications:
     Both the notion and the terminology are quite satisfactory. By the way, 
     I also prefer the term `architectural specifications' to
     `decompositional specifications'.

Let declarations:
     They seem to improve quite often the readability of functional programs. 
     (Surely, there are also other reasons to use them in functional
     languages like let-polymorphism etc).


Typos:

     pp16: GEN-FUN-SYMB  ::= gen-fun-symb FUN-SYMB* FUN-TYPE?
        Is it not intended FUN-SYMB+ instead of FUN-SYMB*?

     pp19: Replace cast_s' : s -> s'  and cast_s : s -> s' by
                   cast_s1 : s2 -> s1 and cast_s2 : s1 -> s2
    
     pp24: Replace LIBRARY-SECTION ::= LINEAR-LIBRARY | CYCLIC-LIBRARY by
                   LIBRARY-SECTION ::= LINEAR-SECTION | CYCLIC-SECTION


------- End of forwarded message -------