OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

office message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: RE: [office] Public Comment #217 - Authoritative Version of Specification

I think, as a practical matter, the defects that occur when one uses an ODF
implementation that is not the one used for the authoring are far more
disturbing than the blemishes you remark about concerning disparities in
PDFs derived from the as-authored ODF documents.  There is a problem with
the current state of interoperability for ODF consumers in comparison with
the ubiquity and stability of PDF at this point.  

The proofing-errors problems seem to apply to any renditions of the
document, although I agree it is difficult to catch those kinds of problems
when introduced in the ODF to PDF transformation and they are not stand-out

Signing the (authoritative) ODF version would be helpful, although we don't
know how many consumers would verify the signature, especially since it is
only being introduced with ODF 1.2 (although implemented in OO.o 2.x and
later).  Unfortunately, passing signature verification does not help us deal
with the consumer then presenting that document incorrectly, although it is
some safeguard against documents that have been inadvertently altered.  It
would be great not to get defect reports against any of those (altered or
misrendered) as reported blemishes and defects of the specification.

Note that our published errata are all against the PDF versions of the
documents because of the PDF stability advantage in this regard.  It appears
that ISO only provides PDF versions (and expects PDF submissions).  So the
risk of un-noticed defects is abated, somewhat, by the many eyes that will
eventually find them when making serious use of the material, attempting to
make translations, etc., especially if they are material defects.

Robin, thanks for raising your concerns.  It is important that we look at
the relative risks of all options.  (I have no idea what the conclusion will
be, though apparently HTML is not on the table [;<).

 - Dennis

-----Original Message-----
From: Robin Cover [mailto:robin@oasis-open.org] 
Sent: Monday, March 30, 2009 07:27
To: Dennis E. Hamilton
Cc: Michael.Brauer@Sun.COM; 'OpenDocument Mailing List'; Robin Cover
Subject: RE: [office] Public Comment #217 - Authoritative Version of

On Mon, 30 Mar 2009, Dennis E. Hamilton wrote:
> I don't think that the easily-editable form of the specification should be
> considered authoritative.  Also, because of device and implementation
> differences, font substitutions, page reflow, etc., we have no control
> variations in appearance and the difficulty of associating comments and
> defect reports against a fixed-form.
> I recommend that the PDF be considered authoritative (and if it could be
> digitally signed that would be good too).  This provides more reliable
> page-chapter-verse cross-referencing and discussion in reports, submitted
> defects, etc.

The technical committee is at liberty to select an authoritative format,
so the following comment (not official advice or direction from OASIS) is
mere informal FYI -- personal comment for consideration.

I always recommend that the editable format be considered authoritative.
Here are two of the most important reasons:

1. The OASIS rule about designating an authoritative format (from among the
    required formats (editable source, PDF, (X)HTML) arose from a concern
    that there might be a "difference" between formats, owing to document
    conversion/translation operations.  That is: conversions are sometimes
    (often) imperfect.  The notion was: if there *is* a difference, we would
    a single point of appeal.  From this POV, it makes sense to designate
    authoring format as authoritative, as it evidently represents the
    authoring/editorial intent.  What are the chances that the authoring
    format is "incorrect" from the standpoint of authorial intent, and that
    some derived format (the result of conversion) "correct"?  Near zero.
    If there is a difference, it's because the conversion created the
    and one not representing authorial intent.

    The differences we *have* observed resulting from PDF generation and
    conversion processes are usually small (or very small) insignificant
    changes that the human eye does not catch, because proof-readers see
    what they know must be written, not what is actually written.  That's
    why catching typos and similar mistakes is only asymptotic: several
    people acting as proof-readers will all miss the same mistake.  Now:
    why would one want a PDF (which has the typo/mistake NOT representing
    the author's intent to be authoritative?  It's a process that
    guarantees adoption of inadvertent errors -- in known classes of cases.
    Better to designate the source (the original) and not the derived
    format "authoritative".

2. OASIS allows the creation of derivative works for a range of uses,
    "without restriction of any kind."

    Official text: This document and translations of it may be copied and
    furnished to others, and derivative works that comment
    on or otherwise explain it or assist in its implementation
    may be prepared, copied, published, and distributed, in
    whole or in part, without restriction of any kind.

    Now, suppose someone does want to create an annotated (commentary)
    version, or other help system that uses the specification text
    in an online context.  We would want the creator of that derivative work
    to *start with* the authoritative format, not with some
    format.  If PDF is selected as authoritative, the creator of the
    derivative work would need to convert the PDF -- by some means, and all
    known to pose serious conversion risks, and often generate artifact
    (Ever try to convert PDF into editable format for such a purose?)

That's enough.  I'm quite sure Dennis has his own counter-arguments.  In
this context, the above text does not count, except as argument for the
TC's consideration.  Your game, your choice.

- Robin (speaking simply as a human being, without institutional


> - Dennis
> -----Original Message-----
> From: Michael.Brauer@Sun.COM [mailto:Michael.Brauer@Sun.COM]
> Sent: Monday, March 30, 2009 05:20
> To: OpenDocument Mailing List
> Subject: [office] Public Comments 214 to 256
> [ ... ]
> 217: Question which document (HTML, PDF or ODF) of ODF 1.1 is the
> authoritative version.
> The OASIS TC process nowadays requires that one of the versions is
> declared to be the authoritative document. This was not the case for ODF
> 1.1.
> Resolution: For the future, declare ODF version to be authoritative.
> [ ... ]
[ ... ]

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]