> Will an implementation having getenv but not environ be SRFI-98
> compliant?

I don't think so.

> Although many useful Scheme programs have been written
> using getenv; I have yet to see one which requires "environ".

On CGI application, I want to check all the environ variables for debug usage.
Some Web servers don't pass all the environment variables to CGI application.

--
Taro Minowa(Higepon)

http://www.monaos.org/
http://code.google.com/p/mosh-scheme/

On Mon, Jul 7, 2008 at 5:43 AM, Aubrey Jaffer <xxxxxx@alum.mit.edu> wrote:
>  | Date: Sun, 6 Jul 2008 19:06:16 +0900
>  | From: higepon <xxxxxx@gmail.com>
>  |
>  | Hi.
>  |
>  | > SRFI-96, SLIB Prerequisites, (requested finalization a month ago)
>  | > specifies single-argument GETENV compatibly with SRFI-98.  Should
>  | > it be mentioned?
>  |
>  | When SRFI-98 becomes final status, will you revise SRFI-96?
>
> I can submit a revision (no guess as to how long it will take to
> appear).  SRFI-96 (and SLIB) will not be requiring the "environ"
> behavior.  Although many useful Scheme programs have been written
> using getenv; I have yet to see one which requires "environ".
>
> Will an implementation having getenv but not environ be SRFI-98
> compliant?  SRFI-98 would be a prerequisite for SRFI-96 only if the
> answer to that question is yes.
>
>  | > Also, I have added support for the no-argument getenv to the
>  | > development version of SCM
>  |
>  | Cool!
>  | This is the first implementation of this srfi!
>
> According to the table at the end of srfi-98, STklos preceeded SCM.
>
>  | I read your code SCM scm_getenv(nam).
>  | Thanks.
>