It seems to me that nested divs at the same level don't imply 
sequential content, parallel content, the same content, different 
content or anything else except a comparable level in the hierarchy. 
Three nested divs at the same level with attributes ORDER="1", 
ORDER="2" and ORDER="3" _would_ overtly indicate sequence, but then 
what do three divs with ORDER="1", ORDER="1" and ORDER="1" indicate - 
parallel content, the same content or something else? Probably 
parallel content if anything, I guess, so that doesn't help. Maybe 
you could appropriate the CONTENTIDS attribute, and assign the same 
URI to each? Probably a hack as well, but a little clearer than using 
the TYPE attribute. Then you could do something like:

div
    div CONTENTIDS="[URI-x]"
       fptr
    div CONTENTIDS="[URI-x]"
       fptr
    div CONTENTIDS="[URI-x]"
       div ORDER="1"
          fptr
             area
       div ORDER="2"
          fptr
             par
                area
                area
       div ORDER="3"
          fptr
             area

--Eric

At 05:04 PM 12/10/2006, you wrote:
>Hello all-
>
>Is there a specific rationale behind allowing both <seq> and <area>
>within <par>, but only allowing <area> within <seq> (and not <par>
>within <seq>)? I'm facing a situation now where I've got three files
>representing the same content in a <div>. Two of these are simple
><area>s within files, but the third is more complicated. One audio file
>plays for a while, then two play together, then back to a single file
>(all within this particular content <div>). What I really need to do is:
>(hope the formatting for this works ok)
>
>div
>         fptr
>         fptr
>         fptr
>                 seq
>                         area
>                         par
>                                 area
>                                 area
>                         area
>
>... but that's not valid - <par> within <seq> isn't allowed. If there's
>not a definite reason for not allowing <par> within <seq>, I'd find it
>useful if this change were made in a future version of METS. :-)
>
>In the meantime, I can put a hack in place using nested <div>s with TYPE
>attributes implying the <div>s at a certain level represent the same
>content rather than a sequence, but it's very much a hack, especially
>when elements exist to do what I need just not in the combination I
>need. Or is there another way to do this I'm not seeing?
>
>Jenn
>
>========================
>Jenn Riley
>Metadata Librarian
>Digital Library Program
>Indiana University - Bloomington
>Wells Library E170
>(812) 856-5759
>www.dlib.indiana.edu
>
>Inquiring Librarian blog: www.inquiringlibrarian.blogspot.com


=====================
Eric L Stedfeld
New York University
Digital Library Program
[log in to unmask]