Path: utzoo!attcan!uunet!samsung!zaphod.mps.ohio-state.edu!rpi!uupsi!sunic!liuida!mikpe
From: mikpe@majestix.ida.liu.se (Mikael Pettersson)
Newsgroups: comp.lang.scheme
Subject: (input-port? (close-input-port (open-input-file "/dev/null"))) ??
Summary: is section 6.10.1 of [R4RS] inconsistent?
Message-ID: <1892@majestix.ida.liu.se>
Date: 26 May 90 14:52:34 GMT
Organization: CIS Dept, Univ of Linkoping, Sweden
Lines: 35


From reading the R4RS (actually, the R3.99RS and the draft IEEE P1178/D4),
section 6.10.1. Ports, one can obtain the following definitions:

(A)	[at the top]
	"..an input port is a Scheme object that can deliver characters
	 upon command.."

(B)	[just below (close-input-port port)]
	"Closes the file associated with port, rendering the port
	 incapable of delivering .. characters."

(C)	[next sentence below (close-input-port port)]
	"These routines have no effect if the file has already been closed."


Now, from (A) and (B) it would seem that once a port has been closed,
it ceases to be a port. On the other hand, (C) seems to imply the
opposite.

Maybe I'm nit-picking (probably :-)), but we don't want accidental
undefinedness in the language definition, do we? [the report is
otherwise pretty careful in describing what is and what is not
undefined behavior]

On a related issue: it's explicitly stated that "It is an error to
read from a closed port.", but no such statement is made about trying
to _write_ to a closed port..  what's the deal here?



/Mike
-- 
Mikael Pettersson, Dept of Comp & Info Sci, University of Linkoping, Sweden
email: mpe@ida.liu.se or ...!{mcsun,munnari,uunet,unido,...}!sunic!liuida!mpe