Re: Latin-script keyboard layout (was RE: Quick Question About Korean Input Methods)

From: Karl Pentzlin (
Date: Tue Jan 05 2010 - 17:13:32 CST

  • Next message: "[Unicode Announcement] New lower individual member rates"

    Am Montag, 4. Januar 2010 um 21:44 schrieb Michael Everson:

        [... regarding keyboard standards ISO/IEC 9995 series:]
    ME> ... They use this weird terminology ...

    SC35/WG1 has its lingo, which looks some less weird if you have a closed look.
    For instance, A is a "shifted" version of a, but # is not a shifted
    version of 3 (nor §, which is is on the 3 key on German keyboards).
    Thus, "shifting" does not exactly describe the function of the key
    called "Shift key" colloquially, especially when the keyboard is
    designed for a non-bicameral script like Arabic.
    Thus, speaking of "Level 1 and "Level 2" instead of "unshifted" and
    "shifted" is more abstract and more universal.

    Now, if you have more than two characters assigned to one, you
    probably want to display them somehow on the keytop. One method is to
    arrange them in rows and columns (rather than e.g. arrange them in a
    circle, like digits on a clock face).

    E.g., if "unshifted" shall yield 3, "shifted" #, "AltGr/unshifted"
    shall yield "€", and "AltGr/shifted" shall yield ¥, you probably engrave:
     # ¥
     3 €
    Now, "level" in SC35/WG1 lingo simply refers to the row, and "group"
    to the column on the keytop. In this case, you have 2 levels and 2
    groups. "3" is Group 1, Level 1; "#" is Group 1, Level 2; "€" is Group
    2, Level 1. You use AltGr as Group Selector here.

    You can use the AltGr key (resp. the "Right Alt" or "Option" key, you
    see these namings are not standardizing) alternatively to select a
    "Level 3" and advise another key (or key combination) to act as Group
    selector, working like a dead key. By this means, you can use all
    levels in the selected group, as no simultaneous pressing is required
    for the group selection itself.
    ISO/IEC 9995-1 explains all this and more in details, e.g. stating
    that you do not need to engrave a lower case character on Level 1 if
    the corresponding upper case character is in Level 2 (not very surprising).

    ME> ... that doesn't fit with the real world ...

    I hope to have proven this claim being false, as I have described
    exactly what you do when you say:
    ME> I use Plain, Shift, Option (Alt if you prefer), and Shift-Option as my
    ME> main states (there are some other states, like Command and Control),
    ME> and I use deadkeys to do other things.

    Also, regarding the real world, you find a photo (taken by Alain LaBonté)
    of a real keyboard here:
    adhering to a real standard - the Canadian standard as e.g. found on:
    You see 3 levels and 2 groups employed, thus having up to 5 characters
    per key (as Level 3 of Group 2 is not used).

    ME> ... and I guess they hope that their standard will get rid of
    ME> what they claim to be "unergonomic" keyboards in favour of what's in
    ME> their standard.

    Not exactly. The standard offers more ergonomic solutions than
    selecting a Group 2/Level 2 character by Shift + AltGr/Option + Character-Key,
    which requires three arbitrary placed keys being pressed the same time
    (no problem for an experienced piano player but not for everybody,
    when it comes to fast everyday touch typing). But if you prefer this
    method, you may adhere to it, within the means provided by the standard.

    - Karl

    This archive was generated by hypermail 2.1.5 : Tue Jan 05 2010 - 17:18:10 CST