RE: [dita] Request for clarification of DITA Technical Committee MeetingMinutes: 17 April 2007

[Michael Priestley <mpriestl@ca.ibm.com>]

I think the spec is actually clear on
this:

>The collection-type attribute indicates how a particular
set of sibling topicrefs 
>relate to each other. The collection-type attribute
is set on the container 
>element for the sibling topicrefs. The collection-type
value can indicate whether 
>to generate links among the siblings, and what kind
of links to generate (for 
>example, next and previous links for a sequence, or
sibling links for a family). 
>The collection-type attribute can also indicate how
the parent topic should link 
>to its children (for example, showing the child links
as a numbered list when 
>the collection-type is sequence).

When collection-type is set on a relcell
that contains topicrefs, it should behave according to the spec - ie, in
the same way it would if the topicrefs were in a collection defined by
any other container element. There is no parent topic, but there is no
parent topic in a topicgroup either.

The reason relcolspec and reltable elements
get problematic is just that they cannot contain topicrefs (at least not
directly). So according to the spec, collection-type on those elements
should do nothing, which is what they do - but given that the attribute
is there, we probably want it to do something at some point, so we want
to reserve it for future use.

Michael Priestley
IBM DITA Architect and Classification Schema PDT Lead
mpriestl@ca.ibm.com
http://dita.xml.org/blog/25

"Grosso, Paul" <pgrosso@ptc.com>

04/30/2007 06:17 PM

To	"DITA TC list" <dita@lists.oasis-open.org>
cc
Subject	RE: [dita] Request for clarification of DITA Technical Committee Meeting Minutes: 17 April 2007

I don't have a record of a response to this.

Don, please put this in the agenda for our upcoming meeting.

paul

> -----Original Message-----
> From: Grosso, Paul [mailto:pgrosso@ptc.com] 
> Sent: Tuesday, 2007 April 24 17:37
> To: DITA TC list
> Subject: RE: [dita] Request for clarification of DITA 
> Technical Committee Meeting Minutes: 17 April 2007
> 
>  
> 
> > -----Original Message-----
> > From: Grosso, Paul [mailto:pgrosso@ptc.com] 
> > Sent: Monday, 2007 April 23 9:05
> > To: DITA TC list
> > Subject: [dita] Request for clarification of DITA Technical 
> > Committee Meeting Minutes: 17 April 2007
> 
> > Re:
> > 
> >  * Question about collection-type on reltable or relcolspec
elements
> >  * http://lists.oasis-open.org/archives/dita/200704/msg00008.html
> > 
> >   Discussion on the merits of making this change.
> > 
> >   Michael: We should clarify this in the 1.1 spec. Question
> is should 
> >   we remove them in 1.2?
> > 
> >   Don is concerned about how this would look if we then
> > rearchitect this in the future.
> > 
> >   DECISION: This will be treated as a 1.1 spec comment.
> 
> Unfortunately, I don't think I got as much clarity as I had
> hoped for during today's telcon.  Maybe I'm just missing
> something obvious on this subject, but I think I need something
> in writing.
> 
> My notes from today's telcon say this about what I heard:
> 
>  The official word is that DITA 1.1 will not define what
>  it means to use the collection-type attribute on the
>  reltable element.  It will be reserved for future use--and
>  presumably defined in DITA 1.2.
> 
> When I transmitted that back to the rest of Arbortext, I
> was asked:
> 
>  So in DITA 1.1 the use of collection-type is undefined on
>  both reltable and on relcolspec, but its use is defined
>  elsewhere including on relcell?
> 
> I have no idea what to answer.
> 
> It doesn't seem right that it's so hard for a TC member--much
> less someone else--to figure out what our decision is in this 
> case especially after we've ostensibly discussed it at two
> consecutive TC meetings.
> 
> Can we please get a definitively worded decision in writing?
> 
> Michael, I think you might be the best one to take a crack
> at this.  I don't mean spec-ese, I just mean what we decided
> in layman's terms as the answers to the questions at
> http://lists.oasis-open.org/archives/dita/200704/msg00008.html
> 
> paul
>