Well there is still the issue of whether or not amendments or amendlist is valid as part of booklist in the frontmatter regardless of the name we choose. Because whatever we have in booklist can be valid in frontmatter and backmatter.
I assume there was a good reason for only including amendments in the backmatter. Essentially the same thing would occur if we wanted a revision history list. Is it only valid in the frontmatter or backmatter.
If revisions are frontmatter material and amendments are backmatter, then does it make sense to have one element and let the processing deal with them as revisions or amendments?
Ãric Sirois
DITA Toolsmith
IXIASOFT
825 Querbes, Suite 200, MontrÃal, QuÃbec, Canada, H2V 3X1
tel + 1 514 279-4942 / toll free + 1 877 279-4942
mobile + 1 647 462-3620
eric.sirois@ixiasoft.com /
www.ixiasoft.com From: Scott Hudson <
scott.hudson@jeppesen.com>
Sent: Monday, August 13, 2018 5:22 PM
To: Ãric Sirois <
eric.sirois@ixiasoft.com>; DITA TC <
dita@lists.oasis-open.org>
Subject: Re: [dita] Bookmaps and booklists
What if we deprecated the <amendments> element and created a new <amendlist> that followed the similar content model of the other <booklist> elements?
Then you could still move the functionality, without the content model issues?
--Scott
From:
<
dita@lists.oasis-open.org > on behalf of Ãric Sirois <
eric.sirois@ixiasoft.com >
Date: Monday, August 13, 2018 at 3:12 PM
To: DITA TC <
dita@lists.oasis-open.org >
Subject: [dita] Bookmaps and booklists
Hi,
I have been having a discussion with Eliot and Kris about some f the changes to bookmap for DITA 2.0, specifically the amendments update. The issue at hand is that while the update is doable with RNG and XSD, we
will have issues when converting them to DTDs. So if we wanted to support this in 2.0, we would need two different elements to differentiate the booklists from the backmatter and the frontmatter. As well there is Scott s note about revision history in booklists
as well that we need to discuss.
Here is a snippet of the discussion:
<snip author= Eric >
I m working on the changes to bookmap for DITA 2.0 and one change we wanted to do was to move <amendments> into <booklists>. That s easily doable with RNG and XSD, but no so much with DTDs and likely the reason
why it's designed that way in the current DTDs. The concern that I have at the moment is that when we do an RNG to DTD conversion, we are going to run into the same issue again because it s not technically feasible in the DTD world unless I m missing something.
The technical issue is that the content model for booklists front matter would be different than the content model for booklists in the backmatter. If we can t get the design we want to work in the DTD world, then
I think we should leave things as is for DITA 2.0.
</smip>
<snip author= Eliot >
The RNG must conform to DTD constraints, unfortunately, so a given element must have the same content model everywhere it is used. That means booklists has to be a union of everything that would be wanted at the
front or back of a manual or you would need two different booklist element types in order to have different content models. That seems like a reasonable solution, e.g. booklists-frontmatter , booklists-backmatter . Ideally, we would have a domain that defines
the abstract base type from which specific booklist specializations would be derived (i.e., move current booklists to a domain and then use that domain as the basis for whatever specialized booklists you need). But I realize that goes beyond the current remit
for BookMap 2.0.
</snip>
Kind regards,
Ãric Sirois
DITA Toolsmith
IXIASOFT
825 Querbes, Suite 200, MontrÃal, QuÃbec, Canada, H2V 3X1
tel + 1 514 279-4942 / toll free + 1 877 279-4942
mobile + 1 647 462-3620
eric.sirois@ixiasoft.com /
www.ixiasoft.com