RE: Fwd: Re: (SC2WG2.609) New contribution N2705

From: Kenneth Whistler (kenw@sybase.com)
Date: Wed Feb 18 2004 - 19:49:33 EST

  • Next message: Peter Constable: "RE: Fwd: Re: (SC2WG2.609) New contribution N2705"

    Michael responded:

    > At 15:02 -0800 2004-02-18, Peter Constable wrote:
    >
    > >- What is the potential that later on someone
    > >will start using h[a~e], or perhaps h[aŮe]!
    > >(say, to indicate an a-coloured laryngeal that
    > >in a certain context has become e-coloured)?
    >
    > Of course such a question cannot be answered.

    But such prospects need to be evaluated in terms of their
    likelihood and their potential for causing further problems
    in the encoding in the future.

    > There are billions of people on our planet.

    Non sequitur.

    > >- If such usage should arise, how would our
    > >decision to encode [/] affect how we decide
    > >about [] or [Ů]?
    >
    > Ignores that parentheses and hyphens are already encoded as subscripts.

    And your retort ignores the fact that such characters
    were added in the first place as *compatibility* characters
    for round-trip mapping to Asian character sets that
    had defective models about what should or should not
    be encoded, because of the rendering constraints of
    East Asian glyph-in-a-square-box typography on early
    PC's.

    The subscript-/ has no such roundtrip mapping justification,
    so adding such things is setting new precedents that will
    surely be used (by yourself, if no one else) to argue
    for more such characters in the future when found somewhere
    in somebody's linguistic data.

    *That* is the point of talking about a potential
    for a subscript tilde right now.

    > >The point is, we don't (or, at least, shouldn't)
    > >just encode things because we saw them being
    > >used. We should establish principles (however
    > >formally or informally stated) that we use to
    > >guide our decisions. *That* is why Ken cares
    > >about a possible subscript tilde.
    >
    > The boundary is what people need to encode, not
    > whether it fits in with one particular linguist's
    > view of what merits veto.

    This is ironic, coming from an historically-trained
    linguist who has been viewed in other contexts as
    "Mr. Veto" himself.

    You are stubbornly refusing to address the potential costs
    here, in favor of a simple, repeated assertion that people
    "need to encode" this.

    There are costs to precedents set in the standard, particularly
    when they eat away at the principled boundary between
    plain text and rich text. Every such precedent, based on
    whatever good or flimsy context it orginally got into the
    standard, has been used as a reason for encoding more such
    characters in the future, no matter how much they violate
    the ideals of the Unicode text model. The archives of this
    list are replete with people citing such and such a character
    in the standard as justification for bringing in some new
    harebrained collection of stuff.

    Look how the mere discussion of this problem elicited, for
    about the fifth time in Unicode history, the proposal to
    encoding a COMBINING SUBSCRIPT MARK, which once again had
    to be shot down for the really bad idea it is.

    Furthermore, a "need to encode" should be demonstrated,
    not merely asserted.

    The text in question is all easily representable *now* in
    Unicode. It simply requires stylistic markup. Given that
    that is the case, a proponent of encoding a particular style
    of a character as an encoded character itself needs to
    make the case for the urgency of having a *plain text*
    encoding of a style variant of a character. That is what
    is missing from the proposal, particularly for subscript-/.
    The proposal text just hand waves over this question.

    There is no "need" demonstrated here -- only a "want".

    What the committees should have here is a demonstration of
    why Indo-Europeanists cannot get by, in these kinds of materials,
    with simply subscripting such special expressions. After all,
    they have to use other styles in setting such text anyway.
    You should be making the case that such expressions *must*
    be expressible in plain text. I don't see it, and am having
    trouble, even as a historic linguist myself, in seeing what
    the real requirement is here.

    --Ken



    This archive was generated by hypermail 2.1.5 : Wed Feb 18 2004 - 20:21:10 EST