Re: [OT] multilingual support in MS products (was Re: Kurdish ghayn)

From: Michael \(michka\) Kaplan (michka@trigeminal.com)
Date: Sun Apr 27 2003 - 10:35:02 EDT

  • Next message: Michael \(michka\) Kaplan: "Re: [OT] multilingual support in MS products (was Re: Kurdish ghayn)"

    From: "Thomas Milo" <t.milo@chello.nl>

    Some interesting suggestions in this mail -- most of which have little to do
    with Office (poor Chris gets picked on for a lot of things, including
    feature requeests of other groups!)

    > Would it be possible to make the IJ/ij available at last as a single
    > character IJ/ij for Dutch users?

    <snip a bit>

    Well, the spell checking issues are Office ones, for the ones who own the
    proofing tools. I do not know how flexible the spelling issues are here (I
    think spelling is always case insensitive), but perhaps one could leverage
    the grammar checking, which does notice case irregularities. The end result
    of flagging cases that are incorrect would be the same.

    > A simple way to correct this situation would be to convert a
    > US-International style keyboard as defined for MS Windows - which adds
    most
    > of the Dutch diacritics except c-cedilla and IJ/ij: by generating the
    correct
    > single Unicode from the sequence Ij-IJ and from ij as well as joining ,C
    and
    > ,c (comma c) into C/c-cedilla. In the exceptionally rare case where Ij/
    > needs to remain two characters in Dutch, <I> <space> <j> would suffice -
    the
    > same procedure to avoid "a to compose ä.

    It is never possible to change keyboard layouts in Windows once they have
    shipped, and it would in any case be inappropriate to use a Dutch-specific
    convention on an attempted multilingual keyboard.

    I am also not entirely clear on the mechanism you suggest here as there does
    not exist a way to change the input in the way you suggest other than
    treating "I" as a dead key, a "fix" that we know is worse than the problem
    as there is already a host of user complaints about the fact that input is
    not immediately visible and that some characters require the extra space
    character be entered.

    > May I suggest a similar improvement for all relevant Arabic-based
    keyboards
    > (like Arabic, Persian and Urdu and any other) that include U+0621 Hamza:
    >
    > Wherever U+0621 Hamza is preceded by either U+0649 Heh or U+06D5 AE to
    > substitute U+06C0 Heh-Hamza and, by analogy, wherever U+0621 hamza is
    > preceded by U+06C1 Heh Goal to substitute 06C2 Heh-Hamza Goal. A single
    > back-space from U+06C0 Heh-Hamza should default to U+06D5 AE.

    <snip a bit>

    Well, see the above -- there is not really a mechanism in Windows for
    keyboards that allows the backspace to change the data being edited.

    > Finally, it would be the Arabist's dream to get the subscript dot by
    typing
    > a regular period directly before the affected letter, e.g., Kurdish Ghayn
    by
    > .x to get x̣, just like the already existing practice of typing "a to get
    ä.
    > Such an Arabist's keyboard mapping should be adjusted for ^A and ^a to
    > produce Ā and ā (US-International is set up for  and â)

    As I mentioned before, it is not possible to change keyboard layouts. But
    this type of change with an altered (or added) dead key is a perfect
    candidate for the MS Keyboard Layout Creator, or any number of third party
    tools.

    MichKa [MS]



    This archive was generated by hypermail 2.1.5 : Sun Apr 27 2003 - 11:03:01 EDT