Re: New Unicode input utility

From: Jukka K. Korpela (jkorpela@cs.tut.fi)
Date: Sun Jan 01 2006 - 03:10:15 CST

  • Next message: Raymond Mercier: "Re: New Unicode input utility"

    On Sat, 31 Dec 2005, Raymond Mercier wrote:

    > Doug Ewell writes
    >> BTW, the writer goofed by saying that Alt+plus+0165 would give YEN SIGN
    >> (¥). U+0165 is actually LATIN SMALL LETTER T WITH CARON (t).
    >
    > Of course these are both correct (note the 'plus' !).

    I'm afraid there's a notational confusion here. We are used to
    denoting key combinations either as Alt-x or as Alt+x, when we
    mean "press the Alt key down, type x, release the Alt key".
    This works up to a point, and we can usually interpret each other's
    notations; some people and publishing houses prefer Alt-x, some Alt+x.
    When a hyphen or a plus sign is part of the key sequence, things
    easily get confusing. Moreover, in more complicated sequences, it
    is not at all obvious when one should release the Alt key.

    This is why I used Alt(x) in a previous message (but forgot to explain
    why). This notation could be used for simple Alt sequences as well
    (especially since it lets us distinguish between Alt(x) and Alt x, i.e.
    pressing and releasing Alt, then x) but it would be particularly useful
    for complex sequences.

    The confusion quoted above would be solved by noting that Alt(0165)
    yields the yen sign, Alt(+0165) yields t with caron when EnableHexNumpad
    is enabled, but when it is not, the use of the "+" key is ignored, so
    Alt(+0165) reduces to Alt(0165).

    (When EnableHexNumpad is enabled, leading zeroes are not significant,
    and we can use e.g. just Alt(+165).)

    The Alt(x) notation is useful when describing the Quick Unicode tool, too.
    For example: to produce U+0141, type Alt(.141). This also makes it clear
    where the sequence ends, even when a punctuation mark follows.

    However, there's still the problem how to indicate that some characters
    mean that the numeric keypad (or its equivalent) must be used. I have
    no good solution to this in plain text.

    > Jukka "Yucca" Korpela finds that QuickUnicode doesn't work in Windows 98, but
    > I would not expect it to, since only the NT systems can be expected to work
    > with unicode.

    The page http://www.cardbox.com/quick/technical.htm says:

    "System requirements
    The Quick Unicode tool is designed to work on all versions of Windows from
    Windows 95 upwards and from Windows NT 4.0 upwards. We have tested it on
    Windows 98, Windows 2000, and Windows XP (Home and Professional
    Editions)."

    When I try it on Windows 98, it works up to Alt(.ff), but then things
    become strange (though not so unexpectedly): I get e.g.
    Alt(.100) A
    Alt(.101) a
    Alt(.102) A
    Alt(.103) a
    Alt(.104) A
    Alt(.105) C

    -- 
    Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/
    


    This archive was generated by hypermail 2.1.5 : Sun Jan 01 2006 - 03:12:17 CST