Semantic Model Mail Archive: SM> Integrate "-actual&quo

SM> Integrate "-actual" attributes into the document object spec

From: Dennis Carney (dcarney@us.ibm.com)
Date: Tue Apr 15 2003 - 19:16:28 EDT

  • Next message: HALL,DAVID (HP-Vancouver,ex1): "SM> Where to publish xsd / wsdl"

    In "Appendix A: Change Log" of the IPP "-actual" attributes extension
    (ftp://ftp.pwg.org/pub/pwg/ipp/new_ACT/pwg-ipp-actual-attrs-latest.pdf), is
    the following text:

        Version 0.3, 16 December 2002, as a result of the PWG Semantic Model
    telecon, December 12, 2002:
            1. Removed all references to the document object. Extending this
    concept to the document object will be done
                in the document object specification only. In this way, moving
    this specification forward on the standards
                track will not be held up.

    As far as I can see, the concept of the "-actual" attributes has not been
    extended to the Document object in the latest version of the Document
    Object spec. I assume we still want to do this--is there any feeling
    otherwise among the group?

    If we *do* want to do this, how to do it? In fact, I'm not sure in reading
    the spec (maybe I missed it?) that it is ever made clear that Document
    Template attributes have corresponding "xxx-default" and "xxx-supported"
    attributes--is that because all Document Template attributes are also Job
    Template attributes and as such have "xxx-default" and "xxx-supported"
    attributes by definition (from 2911)?

    However, the "-actual" attributes are a bit different. For every *Document
    Template* attribute, there would be a corresponding *Document Description*
    "-actual" attribute. This fact cannot, I don't think, be considered to be
    implied by 2911 or by the "-actual" spec (which as shown above does not
    even mention the document object).

    So:
    ISSUE 1: In chapter 7, OK to add a mention of the extension of the
    "-actual" concept to the Document object?
    ISSUE 2: In addition, does the way that "xxx-default", "xxx-supported", and
    "xxx-ready" apply to the Document object need to be made more clear?

    In addition, in Table 7, which lists all the Operation attributes, there
    are columns for "xxx-default" and "xxx-supported". I guess the concept is
    to extend the concept of "xxx-default" and "xxx-supported" to certain
    Operation attributes, instead of simply having those concepts apply to
    Template attributes.

    ISSUE 3: In this case, would we want to do the same for "-actual"s?

    This is a complicated situation, I believe, since we already have two
    cases:
    1) For job-impressions, job-k-octets, and job-media-sheets, we already have
    the concept that these are copied to Job Description attributes and that
    these three (and *only* these three, I believe) can be updated by the
    printer to contain a "more accurate" (see 2911, section 4.3.17) value than
    was provided. So for these, they are sort-of their own "actual" attribute.
    2) For document-format and document-format-version, we have created new
    attributes to essentially be "actual" values: document-format-detected and
    document-format-version-detected. We created the "detected" name to make
    sure it was not confused with the "-actual"s, since the "-actual" concept
    (as written) doesn't extend to Operation attributes.
    My personal vote is to *not* extend the concept of "-actual"s to Operation
    attributes, but I wanted to bring it up to see what others think. My
    reason for not extending them is to keep the concept "clean"--it only
    applies to Template attributes, not to all Template attributes, plus those
    Operation attributes that have currently been identified as having them.
    Also, extending the concept to Operation attributes conflicts somewhat with
    the situation in #1 above--2911 says that those three *are* some sort of
    "actual"s. If we extend the concept, do we deprecate that behavior? But
    then clients that don't implement the document object will no longer get
    the same "results" from servers that do implement it.

    Comments anyone?

    Dennis Carney
    IBM Printing Systems



    This archive was generated by hypermail 2b29 : Tue Apr 15 2003 - 19:16:40 EDT