X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["4382" "Tue" "1" "December" "92" "11:54:29" "GMT" "Adrian F Clark" "alien@SX.AC.UK" nil "83" "Re: LaTeX 3 floating strategy: plates" "^Date:" nil nil "12"]) Return-Path: Received: from sc.ZIB-Berlin.DE (serv01) by dagobert.ZIB-Berlin.DE (4.1/SMI-4.0/1.9.92 ) id AA06643; Tue, 1 Dec 92 14:48:01 +0100 Received: from dagobert by sc.ZIB-Berlin.DE (4.0/SMI-4.0-sc/19.6.92) id AB17008; Tue, 1 Dec 92 14:46:43 +0100 Message-Id: <9212011346.AB17008@sc.zib-berlin.dbp.de> Received: from DHDURZ1 by vm.urz.Uni-Heidelberg.de (IBM VM SMTP V2R2) with BSMTP id 7655; Tue, 01 Dec 92 13:36:37 CET Received: from DHDURZ1 by DHDURZ1 (Mailer R2.08 R208004) with BSMTP id 6166; Tue, 01 Dec 92 13:36:34 CET Received: from DHDURZ1 by DHDURZ1 (Mailer R2.08 R208004) with BSMTP id 6164; Tue, 01 Dec 92 13:36:30 CET Reply-To: Mailing list for the LaTeX3 project In-Reply-To: David_Rhead%vme.nott.ac.uk@uk.ac.essex.mailhost's message of Wed, 25 Nov 92 12:46:49 +0100 Date: Tue, 1 Dec 92 11:54:29 GMT From: Adrian F Clark Sender: Mailing list for the LaTeX3 project To: Multiple Recipients of Subject: Re: LaTeX 3 floating strategy: plates Status: R X-Status: X-Keywords: X-UID: 877 I haven't seen any discussion of David Rhead's interesting message of 25th November regarding plates. Since most of my `real' work is concerned with pictures, this is a subject close to my heart; so here are some comments and suggestions concerned with floating inserts in general and pictures (images, not line-drawings) in particular. WHERE PLATES APPEAR. As far as I can tell, having just looked closely at a few books, plates are typically inserted between signatures in older (dare I say more carefully prepared?) books, or in a separate signature at the end. I only have two books in which colour plates appear as floating inserts on `normal' pages -- both these books were very expensive (and produced by Springer-Verlag). WHAT PLATES CONTAIN. The tendency is to produce grey-scale pictures (halftones) as `conventional' floats, with only colour plates pushed into strange places in the document. The use of PostScript for rendering grey-scale pictures means that the tedious messing around with process cameras is no longer needed. (Having said that, the quality from PostScript halftones doesn't match that from a process camera; but that's another matter.) DIFFERENT TYPES OF FLOATS. David's message implies that plates require a separate numbering sequence from figures and tables. I can think of several other cases in which things that one might wish to have float would also require separate numbering: for example, sections of code ("Listing 4.2: The radix-2 Cooley-Tukey fast Fourier transform"), grey-scale images ("Image 4.2: Result of applying the program of Listing 4.2 to out test image"), colour images, and `pictures' of a user's screen ("...after running this command, your terminal screen should look like that in Screen 4.2"). I have seen separate floats of all these kinds in books. A SUGGESTION. Just as there are \newcommand and friends, surely it is possible to produce a \newfloat command? I'm thinking along the lines of \newfloat[\placement]{listing}{chapter} As well as the name of the floating entity, this would require indication of when the associated counter has to be reset (`chapter' in the above example). Of course, \newfloat would also have to define an appropriate \listoflistings command. Since I recall Frank saying that he is moving towards a single auxiliary file to hold all cross-referencing and contents information, this shouldn't result in an explosion of new files! Perhaps, as the example shows, it would also be possible to present a macro to determine the placement of the class of float --- if this is feasible, it would allow a different placement policy to be used for colour plates from other floats...and allow for special requirements, as in David Rogers' memorable (but plain) floating challenge. Of course, LaTeX could pre-define floats called `plate' `colorplate' (yes, with American spelling), and `listing' in addition to the current `figure' and `table' in terms of the more primitive macros in a manner analogous to \section. CHANGE OF BEHAVIOUR FOR [h]. Might I also suggest that a change of behaviour of the [h] option for figure placement is also worthy of consideration? On a number of occasions, I have been writing documents that use explicit placement of a figure or table to come out between paragraphs. When there is insufficient room for the unfloated float to appear on the current page, it is retained until the end of the chapter, dragging all other floats with it. Could I suggest that, if feasible, a new page is started and the [h] float rendered at the top of it? This is similar in concept to the `.ne' directive in olde worlde text formatters. Oh, and one last thing regarding tables, or more precisely tabulars. When preparing overheads with \raggedright in effect, shouldn't p{} cause paragraphs to be produced ragged-right too? I hope this is of interest/help. ..Adrian Dr Adrian F. Clark JANET: alien@uk.ac.essex INTERNET: alien%uk.ac.essex@nsfnet-relay.ac.uk FAX: (+44) 206-872900 BITNET: alien%uk.ac.essex@ac.uk PHONE: (+44) 206-872432 (direct) Dept ESE, University of Essex, Wivenhoe Park, Colchester, Essex, C04 3SQ, UK.