Re: [clisp-list] Case sensitive reader. A call for action.

[Bruno H. <ha...@il...>]

Marco Antoniotti writes:

> I do not know how many people here have been following the current
> debate on C.L.L. regarding the issue of the case sensitive reader.
> 
> I believe this is not going away and that it is something we need to
> deal with.

Yes. Thanks for your initiative.

> Franz has posed a problem

IMO, Franz has given an answer to a problem that has been haunting us
for years.

> 1 - a straw poll about whether a "case sensitive" reader would be a
>     good thing.

YES. The debate whether case sensitiveness in languages is over. The
model presented by all other mainstream languages (except Pascal) has
proven to be good.

Three things are involved:

  - How to turn it on for selective packages?
    (There are packages out there, like the Kenzo package, which
    are partially written in upper case and therefore rely on the
    traditional ANSI CL behaviour.) CLISP has been implementing the
    following for three years:

    MAKE-PACKAGE and IN-PACKAGE accept a keyword argument :CASE-SENSITIVE.
    Similarly, DEFPACKAGE accepts an option :CASE-SENSITIVE. When its value is
    non-NIL, the package will be case sensitive, i.e. the reader will not
    case-convert symbol names before looking them up or creating them in this
    package. The package names are still subject to (READTABLE-CASE *READTABLE*),
    though.

  - How to turn it on for the standardized packages (CL, CL-USER etc.)

  - How to turn it on for the package names?

Which *features* should be standardized to indicate one or the other?

> 2 - what would be the "cost to implementors" (of - at least CMUCL and
>     CLisp) -- e.g. what would it take to change all internal symbols
>     to lower case.

It depends on whether two distinct images are acceptable, or whether
the switching should be doable at runtime.

> 4 - if the ANSI standard needed to be changed, then how should a
>     consensus be reached, so that it somehow binds the implementors to
>     the set of agreed upon choices.

Ask Rainer Joswig if there is currently a group working on
standardization. If not, it is (IMO) best done by communication among
implementor groups.

Bruno