From owner-freebsd-current  Sun Oct 20 14:30:58 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id OAA15754
          for current-outgoing; Sun, 20 Oct 1996 14:30:58 -0700 (PDT)
Received: from critter.tfs.com (disn20.cybercity.dk [194.16.57.20])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id OAA15745;
          Sun, 20 Oct 1996 14:30:44 -0700 (PDT)
Received: from critter.tfs.com (localhost.tfs.com [127.0.0.1]) by critter.tfs.com (8.7.5/8.7.3) with ESMTP id XAA24991; Sun, 20 Oct 1996 23:29:51 +0200 (MET DST)
To: John Polstra <jdp@polstra.com>
cc: bde@zeta.org.au, current@FreeBSD.org, wollman@lcs.mit.edu
Subject: Re: <sys/queue.h> 
In-reply-to: Your message of "Sun, 20 Oct 1996 14:12:38 PDT."
             <199610202112.OAA04687@austin.polstra.com> 
Date: Sun, 20 Oct 1996 23:29:51 +0200
Message-ID: <24989.845846991@critter.tfs.com>
From: Poul-Henning Kamp <phk@critter.tfs.com>
Sender: owner-current@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

In message <199610202112.OAA04687@austin.polstra.com>, John Polstra writes:
>> >Don't do that, then.  It's horrid style anyway (IMAO), and you
>> >/certainly/ don't see any native Berkeley code doing that.  The style
>> >guide should discourage the practice if it doesn't already.

I disagree.

>> 
>> I agree.  Typedef should only be used for scalar types and function
>> types.
>
>Why do you say that?  There's already precedent for using typedefs
>for structs in, for example, the "DIR" type of <dirent.h>.  And it
>is in line with C++ practice, where the struct, class, or union
>keyword is almost never used outside of the declaration.

This is actually why I do it.  The code I'm crunching right now is
pretty much C++ written in C, ie. object oriented.  Not because I
want to rewrite the kernel in C++ at a later date, but because it
is the design model that fits this particular piece.

My approach to fixing <sys/queue.h> would probably be to add the
new macros for each kind of queue, rather than have to change
all the code that uses it now:

	#define SLIST_HEAD_TYPE(name, type)                     \
	struct name {                                           \
		type *slh_first; /* first element */            \
	}

(_TYPE suffix is open for discussion)

I would probably add the _EMPTY, _NEXT, _PREV, _FIRST and _LAST macros
to all the queues as apropriate while I'm there.  I consider it a major 
flaw in <sys/queue.h> that you have to know anyting about the internal
representation.

--
Poul-Henning Kamp           | phk@FreeBSD.ORG       FreeBSD Core-team.
http://www.freebsd.org/~phk | phk@login.dknet.dk    Private mailbox.
whois: [PHK]                | phk@ref.tfs.com       TRW Financial Systems, Inc.
Future will arrive by its own means, progress not so.