[Haskell-cafe] Disadvantages of de Bruijn indicies?

[Pepe Iborra]

On 13/05/2007, at 12:44, Neil Mitchell wrote:

> Hi
>
> Thanks for all the responses, I'm busy reading through them.
>
> I'm still trying to decide whether I should use them or not. They
> complicate things, are less intuitive than names. But on the other
> hand, the language I'm working in is untyped and has only letrec.
> These things make binding errors easier to occur, and harder to
> detect.
>
> Thanks for the helpful throughts,
>
> Neil

The Calculus of Indexed Names and Named Indices (CINNI) [1] looks  
really neat:

"The Calculus of Indexed Names and Named Indices (CINNI) is a new  
calculus of explicit substitutions that combines names and indices in  
a uniform way. It contains the standard named notation used in logics  
and programming languages as well as de'Bruijn's indexed notation as  
sublanguages. "


Disclaimer: I haven't read the Epigram paper nor Berkling, so I don't  
know how it compares. And I can't really talk from my own experience,  
since I have not played myself with CINNI yet. I wish I had a build-a- 
language assignment, just so that I could put CINNI to work...

[1] - http://formal.cs.uiuc.edu/stehr/cinni_eng.html