This page is part of the web mail archives of SRFI 56 from before July 7th, 2015. The new archives for SRFI 56 contain all messages, not just those from before July 7th, 2015.
From: "Hans Oesterholt Dijkema" <hans@xxxxxxxxxxxxxxxxx>
Subject: Re: binary vs non-binary ports
Date: Fri, 17 Sep 2004 23:23:13 +0200
> Let the software engineer decide how to handle the "markup" of
> his/hers protocols. If he/she wants to mix clarity of text with
> performance of binary protocols. Let him/her do so.
That's fine. One can implement binary/character mixing port
on top of basic binary ports. The issue here I understood is that
R5RS states open-input-file/open-output-file create character ports,
and some implementations don't want to have binary ports on top of
character ports, not vice versa.
My personal feeling is for leaving this issue to another srfi;
this srfi is already pretty useful, even with a condition that
"some port are not capable of mixing binary/character i/o".
The ideas I proposed (predicates to check port capability, and
open-{input|output}-file-port) are to compromize for some issues
raised recently, but it seems it's going to lead to a loose end.
--shiro