Re: Much better Latin-1 keyboard for Windows

From: Cristian Secarã (
Date: Sat Jul 24 2004 - 17:43:19 CDT

  • Next message: Philippe Verdy: "Re: Umlaut and Tréma, was: Variation selectors and vowel marks"

    On Sat, 24 Jul 2004 18:51:04 +0200, busmanus wrote:

    > My intuition would suggest, that the language setting should be
    > independently modifiable if necessary from within the editor you are
    > working with, rather than having different language versions for
    > otherwise identical keyboard layouts.

    There are several aspects that can be discussed (for example the
    capabilities or limitations of the system, or the editor), but let's
    limit the discussion to the keyboard itself.

    As I said, I have inspired myself from MS Windows OS, where one can
    first choose a (keyboard) input locale for the same base language (for
    example English (Australia), English (Canada), English (Ireland),
    English (etc.), or approx. 6 French categories, or approx. 5 Germans
    categories, or approx. 20 Spanish categories, etc.), then for that
    selected language several keyboard layouts can be choosen
    simultaneously, to switch between them at any moment during an
    application work.

    In my case, there is only one language defined (Romanian).
    Then, for this Romanian, the standard defines the overall character set
    that *should* be generated using this keyboard (basically the ISO/IEC
    8859-16 character set). Then defines all the dead keys number and
    position and recommends the character set that *may* be composed by
    using those dead keys. I avoided defining as mandatory the list of dead
    key composed characters, because of real-life limitations on existing
    systems (i.e. codepage limitations).

    Then, based on that overall character set, I have defined the two
    different character arrangements (keyboard layouts).

    And still, I am convinced that I cave not break the ISO/IEC 9995-x
    rules (which I have read very carefully, in order to understand what
    can be done to override its limitations and still keep my national
    standard compatible with it).


    This archive was generated by hypermail 2.1.5 : Sat Jul 24 2004 - 17:38:28 CDT