IPP Mail Archive: IPP> RE: Revised ABNF per Monday's Phone C

IPP> RE: Revised ABNF per Monday's Phone Conference

From: Hastings, Tom N (hastings@cp10.es.xerox.com)
Date: Tue May 15 2001 - 20:57:24 EDT

  • Next message: Hastings, Tom N: "RE: IPP> Revised ABNF per Monday's Phone Conference"

    Ron,

    I sure didn't hear any agreement to add all the class names to the ABNF.

    We didn't want the class names in the ABNF because we wanted to be able to
    register additional class names as needed, without revising the standard.
    Registration means just adding the names to a list on the PWG server.
    During the following year, we could republish with any registered names
    added.

    On the other hand, the advantage of having the class names in the ABNF was
    to tie each class name to either the "mm" or the "in" units, so that the
    class names wouldn't be mis-used with the other units.

    So I thought the compromise win-win that we agreed to was when we recognized
    that the "in" class names was a very small set and very unlikely to grow,
    compared to the "mm" class names. Therefore, we only needed to add the "in"
    names to the ABNF. The result is that any other class name MUST go with
    "mm" units.

    We also discussed whether or not to allow "-" and we agreed not to, but did
    agree to allow "." (at least I didn't hear any objections, when Ira
    suggested it) since it is used in registry names to indicate hierarchy, in
    case we ever need it.

    What do other think?

    Thanks,
    Tom

    -----Original Message-----
    From: Bergman, Ron [mailto:Ron.Bergman@Hitachi-hkis.com]
    Sent: Tuesday, May 15, 2001 15:54
    To: 'Hastings, Tom N'; Bergman, Ron; Ira McDonald (E-mail); Don Wright
    (E-mail)
    Cc: 'ipp@pwg.org'
    Subject: RE: Revised ABNF per Monday's Phone Conference

    Tom,

    I known we discussed both of your issues, but I do not recall such an
    agreement.

    It was my understanding that all class names need to be registered as either
    "inch class" or "mm class". This was the only way to guarantee their
    uniqueness.

    And to keep the class names simple, they were to be a single word.

    What is the advantage of not adding the class to the ABNF? We must always
    update the document to add any new names, and the ABNF can be updated at the
    same time.

            Ron

    -----Original Message-----
    From: Hastings, Tom N [mailto:hastings@cp10.es.xerox.com]
    Sent: Tuesday, May 15, 2001 11:01 AM
    To: Bergman, Ron; Ira McDonald (E-mail); Don Wright (E-mail)
    Cc: 'ipp@pwg.org'
    Subject: RE: Revised ABNF per Monday's Phone Conference

    I believe that we agreed that the mm class names wouldn't need to be
    explicitly listed in the ABNF, since most new class names that are
    registered will be using the "mm" units. Then we don't have to update the
    ABNF, except in the unlikely event that a class name that uses "in" units is
    registered.

    Also we agreed that the registered class names couldn't have "-" in them,
    but could have "." (as in registry path names).

    So here is my suggested ABNF instead (which just changes the class2 line):

    media-size-self-describing-name =
              ( class1 "_" size-name "_" short-dim "x" long-dim "in" ) |
              ( class2 "_" size-name "_" short-dim "x" long-dim "mm" )

       class1 = "na" | "asme" | "oe"

       class2 = lowalpha *( lowalpha | digit | "." )

       size-name = ( lowalpha | digit ) *( lowalpha | digit | "-" )

       short-dim = dim

       long-dim = dim

       dim = integer-part [fraction-part] | "0" fraction-part

       integer-part = non-zero-digit *digit

       fraction-part = "." *digit non-zero-digit

       lowalpha = "a" | "b" | "c" | "d" | "e" | "f" | "g" | "h" | "i" |
                  "j" | "k" | "l" | "m" | "n" | "o" | "p" | "q" | "r" |
                  "s" | "t" | "u" | "v" | "w" | "x" | "y" | "z"

       non-zero-digit = "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9"

       digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9"

    -----Original Message-----
    From: Bergman, Ron [mailto:Ron.Bergman@Hitachi-hkis.com]
    Sent: Tuesday, May 15, 2001 09:18
    To: Tom Hastings (E-mail); Ira McDonald (E-mail); Don Wright (E-mail)
    Cc: 'ipp@pwg.org'
    Subject: Revised ABNF per Monday's Phone Conference

    Here is my proposed ABNF to document the agreed restrictions in yesterday's
    phone call. I may be missing some of the class names but this should
    correctly linl\k each class to only one set of units.

    media-size-self-describing-name =
              ( class1 "_" size-name "_" short-dim "-" long-dim "in" ) |
              ( class2 "_" size-name "_" short-dim "-" long-dim "mm" )

       class1 = "na" | "asme" | "oe"

       class2 = "iso" | "jis" | "jpn" | "prc" | "roc" | "om"

       size-name = ( lowalpha | digit ) *( lowalpha | digit | "-" )

       short-dim = dim

       long-dim = dim

       dim = integer-part [fraction-part] | "0" fraction-part

       integer-part = non-zero-digit *digit

       fraction-part = "." *digit non-zero-digit

       lowalpha = "a" | "b" | "c" | "d" | "e" | "f" | "g" | "h" | "i" |
                  "j" | "k" | "l" | "m" | "n" | "o" | "p" | "q" | "r" |
                  "s" | "t" | "u" | "v" | "w" | "x" | "y" | "z"

       non-zero-digit = "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9"

       digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9"

    Any comments?

            Ron



    This archive was generated by hypermail 2b29 : Tue May 15 2001 - 20:58:54 EDT