[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: FW: book


I most certainly agree that the first edition had nothing on types, safety,
and gc, which is what I teach in addition to the principles of the book.

The authors have recognized these shortcomings and address some of them in
the second edition. It still lacks a discussion of GC, however, and I tend
to refer my students to Paul Wilson's survey article for a reasonably short
introduction. In a semester long course, I also make them build a simple GC. 

As for couching these results in terms of different syntaxes, I suggest we
agree to disagree that this is a topic of importance. 

Interfacing to legacy systems is a topic that has only come up some five
years ago. Indeed, I would claim that if practical, un/anti-academic LLs
have a distinct advantage over impractical, academic LLs, then it is the
appearances of their ability to integrate legacy code safely into a modern

The reason I used the word "appearances" is that personally I find nothing
less satisfying than integrating an existing program into one of my scripts
via strings or char streams (and yes, I do that sometimes). And if you mean
FFIs, then please do look at PLT Scheme's FFI and let me know how to
improve on it without giving up safety. 

But you're correct. This topic is missing from the discussions and the
books. Perhaps it should be a topic for LL2 or LL(n+2). 

-- Matthias