Re: Forward: D tags and ICADD thread.

At 11:41 1997/05/20 -0400, Al Gilman wrote:
>...The place where I fall off the track laid down
>by T.V. [Raman] has to do with making references to print versions in 
>the accepted indexing medium of the print realm.
>
>[picking up the quote from T.V.]
>
>  Daniel--  so you have some background about IPP and BPP:
>  
>  The ICADD 22 was designed to help in the production of etexts and
>  braille texts --these were necessarily looking at documents that had
>  already been published.
>  This is why tags representing post-processing information such as IPP
>  and BPP made their way into that DTD.

[BPP] in that draft was just PP. For symmetry Murray's notes made it BPP
to make clear that both Braille and Ink page numbers may be needed.
In general, there are several BPP per IPP.

>  
>  As you point out, these do not have a place except in the printing
>  spec, and I'd redirect this to that working group.
>  
>[Al, here...]
>
>This is where my agreement breaks down.  Specifically, "these do
>not have a place except in the printing spec" goes too far.  The
>Braille user with an ICADD 22 document can get an answer to the
>question "Where am I in the print-version coordinates?"  That
>question should be answerable by some mechanism we support.  It
>is not just a matter of instructing the printing device.  This is
>required for coordination [over the phone] by people using print
>and Braille variants of the same document.  The same goes for
>print and Web versions of the same document.  Braille users are
>not the only users needing cross-index or coordinate
>tranformation services.

The transformation services could use IPP or BPP, though both are
coarse finding aids. 
>
>Compare this with the issue of the West Publishing copyright on
>the pagination that is required as the means of indexing in court
>proceedings in the U.S.
>
>--
>Al Gilman
>
>
Thanks for this insight. It is quite pertinent. 
A significant reference for that is:

    http://www.abanet.org/citation/

On August 6, 1996, the American Bar Association's House of
      Delegates favorably passed a motion for a universal citation
      system recommendation to the courts. The Resolution from
      Report No. 107 recommends that courts adopt a universal
      citation system using sequential decision numbers for each year
      and internal paragraph numbers within the decision. The
      numbers should be assigned by the court and included in the
      decision at the time it is made publicly available by the court. 
      It also recommends that parallel citations to commonly used print
      sources be strongly encouraged. The Citation system is equally
      adaptable to printed and electronic case reports and is thus
      medium neutral. 

      The standard form of citation, shown for a decision in a federal
      court of appeals, should be:

            Smith v. Jones, 1996 5Cir 15, ¶18, 22 F.3d 955. 

      1996 is the year of the decision; 5Cir refers to the United States
      Court of Appeals for the 5th Circuit; 15 indicates that this
      citation is to the 15th decision released by the court in the year;
      18 is the paragraph number where the material referred to is
      located, and the remainder is the parallel citation to the volume
      and page in the printed case report where the decision may also
      be found. 

Another reference is:

    http://law.wuacc.edu/aallnet/citeform.html

March 1, 1995 Report, AALL Task Force on Citation Formats
American Association of Law Libraries

I paraphrase its lengthy content:
 
Objective: medium-neutral, vendor-neutral, and in the public domain.

Recommendation: Document references are sequential by year by court by
decision. Sequentially numbered paragraphs are the primary internal
"pinpoint" targets.

West Law is again' it; they have a vested interest and have copyrighted
their page numbers from their printed versions. Lawyers have been doing
it their way for over a century. Lawyers and judges know how to use
their system. Paper remains the medium for the origin of documents 
with legal effect.


Issues for Generalization:

Above applies to persistent, static legal information. It gets that way
after initial versions are final edited and eventually signed by a judge. 
Adding numbers thereafter could mess up copyfitting and page breaks.
Paper publishing was the durable medium of choice for centuries. 
Page number referencing was OK as that was all there was. Old law on 
paper is difficult to reliably scan/OCR to become electronic. When
done, folio numbers are often at the end of the page, not before them.

Paragraph numbering is finer-grain than page, but subject to ambiguity 
when done algorithmically. Should these be numbered: headings, bulleted
items in a list, block quotes, display equations within a logical
paragraph?

Many electronic documents are dynamic. No sequential paragraph 
numbering scheme would survive revisions, without gaps and sub-numbering
injections. Reorderings would get original numbering out of order. Different
renditions of electronic documents may have different 
content or ordering [e.g., some may be text-only, some may have alt 
text in place of images, some may have D pointers to descriptive text. Some
may have other footnotes by reference vs. by inclusion, some may have
footnotes "nearby", others as endnotes gathered either by division
or by document.]

An SGML way could assign a unique identifier to each tag. A means to have
fractional numbers would allow injections. Missing numbers would indicate
omitted materials. [The military hierarchic numbering of divisions and
subdivisions is a long-standing model for this, possibly to include
a comment about the range of deleted material.] A hierarchic naming
scheme should be able to grow outward, as when one numbered division
becomes part of a new parent, or possibly part of two different parents.
These indicate that the id "uniqueness" is a hard problem. 

So, the IPP BPP issues, when viewed in a larger context, are the URN issues.

The original purpose for IPP and BPP was locators between dissimilar
media for rendition. The finer grain pinpoint location to paragraph
of the legal citations would be useful here as well. Presence of such
pinpoint location as a general practice would be useful for universal
design, not just accessibility.

Regards/Harvey Bingham

Received on Tuesday, 20 May 1997 18:52:19 UTC