IFX Mail Archive: IFX> FW: [Comments on UIF spec]

IFX> FW: [Comments on UIF spec]

From: McDonald, Ira (imcdonald@sharplabs.com)
Date: Wed Jun 27 2001 - 11:54:46 EDT

  • Next message: mjoel@netreon.com: "IFX> IPP Fax Spec Corrections"

    Hi folks,

    Here are comments from my colleague John Thomas at Sharp on UIF spec.

    We should consider them at this afternoon's telecon, if we have time.

    Cheers,
    - Ira McDonald, consulting architect at Sharp and Xerox
      High North Inc

    -----Original Message-----
    From: Thomas, John
    Sent: Tuesday, June 26, 2001 5:36 PM
    To: McDonald, Ira
    Cc: Thomas, John; Olbricht, Eric; Whittle, Craig; Koss, Scott; Murdock,
    Joe; Hurtz, Robert
    Subject: RE: IFX> IPP FAX telecon agenda, Wed, June 27, 10-12 PDT (1-3
    EDT)

    Ira -

    Most of my questions and observations are about Universal Image Format (UIF)
    specification.

    1) The text fields for "ImageDescription", "DocumentName", "Software" and
    "DateTime" all use ASCII encodings. As we have discussed in the past, this
    limits these fields effectively to US English. (Is there an encoding for the
    British pound symbol in 7-bit ASCII?). Is this an oversight, or is this
    because of UIF's TIFF-FX/ITU legacy?
    NOTE: IPP-Fax identifies the "native language", at least for notification.

    2) What is the intended use for the "DateTime" field? Creation date? If
    so, it would be nice if the specification stated this. If an image has a
    creation date, shouldn't it also have an author/copyright holder field?
    Yes, I know this would be easy to edit out, but that would be a felony.
    Right? :-)

    3) The UIF spec expects the name and revision of the authoring software in
    the "Software" text field. I assume this text is format-free? Human
    readable (that is, if you are an English speaking human). And while we are
    on the subject of "Revision", shouldn't the specification identify its own
    revision? This is a common technique to provide "future backward
    compatibility".

    4) Why do profiles C, F and J (optionally) allow centimeter resolution
    units, but profiles L and M only allow "inch" resolution units? TIFF-FX
    again? Again I ask, how would this specification change if the United
    States FINALLY went metric? Is there a reason to prohibit cross-unit
    compatibility like, for example, the computational cost of a good
    sub-sampling algorithm?

    5) Do all (international) bitmap format standards order scan lines
    left-to-right and top-to-bottom? If not, is there a need for UIF to specify
    this order? This scan-line order inherited from the base formats (e.g.
    TIFF-FX)?

    6) I prefer the presentation of compliance requirements in the IPP-Fax
    specification to that in the UIF specification. IPP-Fax makes it clear what
    is "receiver" responsibilities and what is "sender" responsibilities. UIF
    doe not.

    7) Line 152 of the IPP-Fax specification says: "Universal Interchange Format
    (UIF)". I think it should say "Universal Image Format (UIF)".

    Please pass on to the appropriate authorities any of these thoughts which
    have any merit.

    Thanks.

    John
    jct@sharplabs.com



    This archive was generated by hypermail 2b29 : Wed Jun 27 2001 - 11:59:43 EDT