From owner-freebsd-doc Mon Oct 25 13:51:23 1999 Delivered-To: freebsd-doc@freebsd.org Received: from nothing-going-on.demon.co.uk (nothing-going-on.demon.co.uk [193.237.89.66]) by hub.freebsd.org (Postfix) with ESMTP id 7E3E614CFC; Mon, 25 Oct 1999 13:51:00 -0700 (PDT) (envelope-from nik@nothing-going-on.demon.co.uk) Received: from kilt.nothing-going-on.org (kilt.nothing-going-on.org [192.168.1.18]) by nothing-going-on.demon.co.uk (8.9.3/8.9.3) with ESMTP id VAA61015; Mon, 25 Oct 1999 21:01:20 +0100 (BST) (envelope-from nik@catkin.nothing-going-on.org) Received: (from nik@localhost) by kilt.nothing-going-on.org (8.9.3/8.9.3) id IAA03987; Mon, 25 Oct 1999 08:11:59 +0100 (BST) (envelope-from nik@catkin.nothing-going-on.org) Date: Mon, 25 Oct 1999 08:11:59 +0100 From: Nik Clayton To: jkoshy@FreeBSD.org Cc: Satoshi - Ports Wraith - Asami , freebsd-doc@FreeBSD.org Subject: Re: #anchors in .html handbook Message-ID: <19991025081159.C2102@kilt.nothing-going-on.org> References: <199910151150.EAA30422@freefall.freebsd.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95.4i In-Reply-To: <199910151150.EAA30422@freefall.freebsd.org>; from jkoshy@FreeBSD.org on Fri, Oct 15, 1999 at 04:50:13AM -0700 Organization: FreeBSD Project Sender: owner-freebsd-doc@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Fri, Oct 15, 1999 at 04:50:13AM -0700, jkoshy@FreeBSD.org wrote: > > Eek. So it was the previous behavior that was wrong? (And may I > > A small correction: I had assumed that the 'name' attribute for > was parsed as an ID. A perusal of the HTML declaration however > shows that it has been declared as CDATA and is thus is not under > the scope of the NAMING case conversion rules. The HTML bit is a red herring -- the case conversion is carried out by /usr/local/share/sgml/docbook/dsssl/modular/lib/dblib.dsl:normalize() according to the SGML declaration in effect at the time. At the time, the SGML declaration being used is the one for DocBook, where IDs are mapped to upper case. At least, that's my reading of it at the moment. We could write our own version of normalize() which didn't do this. I suspect we'd then be wrong according to the spec, which would be a bad thing. Depending on how much work it is on the part of the ports team (and if it's work we can help with then let us know) I'd rather we bit the bullet and moved over. An alternative would be another post-processing of the HTML to make the attributes lower case. This is probably still within the (HTML) spec, so would be allowed. Thoughts? N To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-doc" in the body of the message