Re: Full Unicode Computer Keyboard

From: Hans Aberg (haberg@math.su.se)
Date: Tue May 10 2005 - 17:08:29 CDT

  • Next message: JFC (Jefsey) Morfin: "Re: Full Unicode Computer Keyboard"

    At 23:42 +0200 2005/05/10, JFC (Jefsey) Morfin wrote:
    >>> - the first one being the keyboards. With real problems since
    >>>my AZERTY keyboard does not support all the legal French
    >>>characters, so I must rely on the goodwill of applications to
    >>>enter characters sequences correctly, what is totally impossible
    >>>(how do you want a program to know that COEUR has a ligature and
    >>>NOE has not).
    >>
    >>This last is essentially a parsing and rendering problem: If COEUR
    >>and NOE can be identified by the context they are in, the correct
    >>semantic information can be attached to them. This semantic
    >>information can then be used to provide a correct rendering. The
    >>problems is similar to the multiple uses of "-", as a number
    >>negation, and a number range which in correct typesetting is
    >>rendered differently. This difference led me to the notion of
    >>"input characters".

    >Here I am lost. Context has nothing to do with hardware?

    What I mean is that it might not healthy expecting te computer
    keyboard input to be able to supply the needed context in each key.
    It might be faster to enter French without accents, letting the
    computer software figuring them out, while retaining the ability to
    enter them in special case by more complicated key combinations.

    So hardware and software are intimately intertwined.

    >I just say that I am quite unhappy to miss on my keyboard keys
    >permitting me to properly type my own language.

    I just got private email, pointing out that on Mac OS X, anyone can
    create their own keyboard layouts using the program Ukelele:
       http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&item_id=ukelele

    Regardless whether this might solve your problem, it just illustrates
    the intimate connection between hardware and software.

    >What is interesting me is therefore to study the different charsets
    >(as defined by RFC 2277) in use and to see what are the industrial,
    >economic, technical, ergonmic, etc. parameters of their best support
    >for different languages. Which propositions can be made towards a
    >standardisation and political decisions making this issue to be
    >addressed.

    Even though it is important to design special keyboard input methods
    for certain special, often used, Unicode character subsets, it will
    probably not be sufficient for general use. The problem arise when
    character sets must be combined. Think of say a Russian mathematician
    writing some lectures in Russian then, but which must also use
    English words. Then that fellow needs to quickly switch between
    English and Russian character sets, as well having quick access to
    thousands of math characters. If the keyboards input methods are then
    tied to special keyboards, then that will be problematic.

    But on the hardware side, one can think of using as board essentially
    a touch screen. Then any keyboard layout can be displayed on that
    one. The problem with ouch methods is usually to provide sufficient
    feel feedback to the human user. So it needs to augmented with
    something such. Perhaps some kind of electrical current that gives
    the feel of proximity and key borders.

    -- 
       Hans Aberg
    


    This archive was generated by hypermail 2.1.5 : Tue May 10 2005 - 17:09:23 CDT