X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["2250" "Tue" "7" "October" "1997" "17:49:35" "PST" "Donald Arseneau" "asnd@TRIUMF.CA" nil "49" "Re: Extended include" "^Date:" nil nil "10" nil nil nil nil nil] nil) Received: from listserv.gmd.de (listserv.gmd.de [192.88.97.1]) by mail.Uni-Mainz.DE (8.8.5/8.8.5) with ESMTP id CAA06145; Wed, 8 Oct 1997 02:55:16 +0200 (MET DST) Received: from lsv1.listserv.gmd.de by listserv.gmd.de (LSMTP for OpenVMS v1.1a) with SMTP id <3.C95D6A6C@listserv.gmd.de>; Wed, 8 Oct 1997 2:53:15 +0200 Received: from RELAY.URZ.UNI-HEIDELBERG.DE by RELAY.URZ.UNI-HEIDELBERG.DE (LISTSERV-TCP/IP release 1.8b) with spool id 210166 for LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE; Wed, 8 Oct 1997 02:53:06 +0200 Received: from REG.TRIUMF.CA (reg.Triumf.CA [142.90.100.2]) by relay.urz.uni-heidelberg.de (8.8.7/8.8.7) with SMTP id CAA00039 for ; Wed, 8 Oct 1997 02:53:03 +0200 (MET DST) Received: by triumf.ca (MX V4.0-1 VAX) id 247; Tue, 07 Oct 1997 17:49:36 PST Message-ID: <009BB6D1.D6B5A400.247@triumf.ca> Reply-To: Mailing list for the LaTeX3 project Date: Tue, 7 Oct 1997 17:49:35 PST From: Donald Arseneau Sender: Mailing list for the LaTeX3 project To: Multiple recipients of list LATEX-L Subject: Re: Extended include Status: R X-Status: X-Keywords: X-UID: 2390 David Carlisle wrote: % % > add reason 2A) for using \include: To produce a partial document % > that is identical to part of the whole document. % % Yes that as well, but to be reliable for that use the pages generated % have to be *exactly* correct and it is always difficult to be sure of % that with an include system. This is the primary use of \include. Yes, the most *common* (correct) use is for quicker latex runs at the draft stage, but in that case the cross-references are unlikely to be correct and nobody really cares. The primary function of the \include mechanism is to make minor changes to a completed final copy. For this use, the \clearpage is practically essential! Without it, and with unincluded parts sharing pages with included parts, just about any change at all would require reprocessing the entire document, as adding just one line will change all the ensuing page breaks. When \clearpage separates sections, it is rare (but possible) for a small change to affect pagination after the included section. % So you \includeonly{chapter3} but the % first page of chapter3 `should' start with a float from chapter2, % (which is now allowed because we assume an improved system with no % \clearpage) I am always amazed by the difference between dialects of English. In North America, "improve" means "make better", not "ruin"! Floats should not be allowed to wander very far for reasons of good typesetting, not just for technical convenience. So an extended include should begin (and end) with a \FloatBarrier or equivalent (placeins.sty). Still, I don't see any case where a non-page-breaking include would be beneficial; when sections run one into the other, you must process the entire document to get the locations right after any change. I will have to look at Matt's work. I believe a non-breaking \include is possible (but not useful imho). It requires: abolish \immediate writes (no problem) Either: write to only the main aux file (goodbye chapterbib) with tags to gobble sections of it when interested in just a part. Or: abolish \immediate \openout. Easy! \FloatBarrier checkpoint \pagetotal etc. Donald Arseneau asnd@triumf.ca