atom feed38 messages in org.oasis-open.lists.ditaRE: [dita] Complexity of bookmap cont...
FromSent OnAttachments
Paul PrescodJun 12, 2006 2:28 pm 
Robert D AndersonJun 12, 2006 3:11 pm 
Erik HennumJun 12, 2006 4:07 pm.gif, .gif, .gif, 7 more
Paul PrescodJun 12, 2006 6:09 pm 
Erik HennumJun 13, 2006 8:10 am 
Paul PrescodJun 14, 2006 12:13 pm 
Robert D AndersonJun 14, 2006 12:53 pm 
Paul PrescodJun 14, 2006 1:10 pm 
Robert D AndersonJun 15, 2006 7:16 am 
JoAnn HackosJun 15, 2006 7:34 am 
Paul PrescodJun 15, 2006 7:48 am 
JoAnn HackosJun 15, 2006 7:52 am 
Esrig, Bruce (Bruce)Jun 15, 2006 7:57 am 
Robert D AndersonJun 15, 2006 9:27 am 
Esrig, Bruce (Bruce)Jun 15, 2006 9:39 am 
Robert D AndersonJun 15, 2006 9:43 am 
JoAnn HackosJun 15, 2006 10:08 am 
Paul PrescodJun 15, 2006 11:19 am 
Paul PrescodJun 15, 2006 11:20 am 
JoAnn HackosJun 15, 2006 11:22 am 
Robert D AndersonJun 16, 2006 7:09 am 
ma...@mekon.comJun 16, 2006 8:28 am 
JoAnn HackosJun 16, 2006 10:51 am 
Robert D AndersonJun 16, 2006 6:13 pm 
Paul PrescodJun 16, 2006 6:24 pm 
Paul PrescodJun 19, 2006 9:56 am 
Robert D AndersonJun 19, 2006 10:12 am 
Robert D AndersonJun 19, 2006 11:26 am 
Esrig, Bruce (Bruce)Jun 19, 2006 12:11 pm 
Robert D AndersonJun 19, 2006 12:20 pm 
Gershon L JosephJun 26, 2006 7:48 am 
Grosso, PaulJun 26, 2006 8:51 am 
Robert D AndersonJun 26, 2006 3:04 pm 
Grosso, PaulJun 26, 2006 3:20 pm 
Paul PrescodJun 26, 2006 3:38 pm 
Robert D AndersonJun 26, 2006 5:03 pm 
Paul PrescodJun 26, 2006 5:59 pm 
Grosso, PaulJun 27, 2006 7:06 am 
Subject:RE: [dita] Complexity of bookmap content model
From:Paul Prescod (paul@xmetal.com)
Date:Jun 15, 2006 11:20:29 am
List:org.oasis-open.lists.dita

Luckily it seems we are agreeing much more than disagreeing and converging quickly!

-----Original Message----- From: Christian Kravogel [mailto:di@seicodyne.ch] Sent: Thursday, June 15, 2006 11:19 AM To: Esrig, Bruce (Bruce); Robert D Anderson; di@lists.oasis-open.org; 'JoAnn Hackos'; Paul Prescod Subject: Re: [dita] Complexity of bookmap content model

We are going to use the new DITA 1.1 bookmap for two customers (Kone and Novartis), and we indeed need a high flexibility of the base bookmap contentmodel when we are going to specialize it.

After checking the latest proposal from Robert and comparing it with the needs for those two projects, I can definitely support Roberts proposal. It works fine to us.

Having booklists in frontmatter meets our requirements perfectly, but I can imagine cases where some users might prefere to have booklists in the backmatter as well. E.g. someone requires toc, figurelist and tablelist before the chapters but glossarylist and indexlist after the chapters. And if, for any reason, this user can not leave the control for that order to the stylesheet.

If nothing speaks against booklists in front and backmatter, I would suggest to have it in both.

Best regards

Christian Kravogel SeicoDyne GmbH Eichenstrasse 36 CH-6015 Reussbühl +41 41 534 66 97 www.seicodyne.com

----- Original Message ----- From: "Robert D Anderson" <roba@us.ibm.com> To: "Esrig, Bruce (Bruce)" <esr@lucent.com> Cc: <di@lists.oasis-open.org>; "'JoAnn Hackos'" <joan@comtech-serv.com>; "Paul Prescod" <paul@xmetal.com> Sent: Thursday, June 15, 2006 6:27 PM Subject: RE: [dita] Complexity of bookmap content model

To take a little step forward - the revised model, without getting in to shared front or back matter items, would be this: <!ELEMENT bookmap (title?, bookmeta?, frontmatter?, chapter*, part*, backmatter?, reltable* )>

<!ELEMENT frontmatter (booklists | draftintro | abstract | dedication | preface | topicref)*>

<!ELEMENT backmatter (appendix | notices | specialnotices | amendments | colophon | topicref)*>

<!ELEMENT part (topicmeta?, (chapter | topicref)*)>

I'll just note here that the title is optional, just because it is in the current implementation. Earlier emails implied it would be required.

About placing items in both the front and the back - I'm not voicing a strong opinion on that either way (though some pretty clearly make sense in only one spot, like preface in the front). The original idea was to have some sort of style setting that controlled where the items went. For example, the index is in <booklists>, which is only in front matter; default processing would usually stick it at the end. A style setting could move it before the appendix, after it, or whatever. You know, that old XML principal - I just have to say I want an index, and my build process controls where to put it.

I'll note that allowing <booklists> in both spots was not possible before the <backmatter> and <frontmatter> containers, which may be why it never came up. There's no technical limitation on it once those are added.

"Esrig, Bruce (Bruce)" <esr@lucent.com> wrote on 06/15/2006 09:57:48 AM:

What about the extra topicref for specialization?

To second that, Lucent allows an overview topic in a part before thechapters.

Regarding front matter and back matter, a flexible content model is better.

Which items do we know are either front matter or back matter but not both? For example, would the order of booklists in the book be completely up to the processing, or is it under author control?

-----Original Message----- From: JoAnn Hackos [mailto:joan@comtech-serv.com] Sent: Thursday, June 15, 2006 10:53 AM To: Paul Prescod; Robert D Anderson Cc: di@lists.oasis-open.org Subject: RE: [dita] Complexity of bookmap content model

I would support having colophon in both front and backmatter as options. Standards among publishers are variable around this. I would not enforce order in the backmatter either. JTH

-----Original Message----- From: Paul Prescod [mailto:paul@xmetal.com] Sent: Thursday, June 15, 2006 8:49 AM To: Robert D Anderson Cc: di@lists.oasis-open.org Subject: RE: [dita] Complexity of bookmap content model

Now that we're getting down to detail...

<!ELEMENT bookmap (title, bookmeta?, frontmatter?, chapter*, part*, backmatter?, colophon?, reltable* )>

The current model has two ways of representing the title. I don't totally understand that design but hope to within a few days.

Can we move colophon into the backmatter?

<!ELEMENT frontmatter (booklists | draftintro | abstract | dedication | preface | topicref)*>

Looks good.

<!ELEMENT backmatter (appendix*, notices? specialnotices*, amendments?, topicref* )>

Is there a reason to enforce order and cardinality on the backmatter but not the frontmatter?

<!ELEMENT part (topicmeta?, chapter*)>

What about the extra topicref for specialization?