Re: Much better Latin-1 keyboard for Windows

From: Philipp Reichmuth (
Date: Sun Jul 18 2004 - 10:35:20 CDT

  • Next message: Jony Rosenne: "RE: Folding algorithm and canonical equivalence"

    "Donald Z. Osborn" <> schrieb am 18.07.04 17:17:49:
    > Thanks for this info. I've been used to thinking in terms of deadkeys or
    > Alt(Gr)- combos as alternative approaches, each with different advantages. For
    > me using an Alt(Gr)- combo as a deadkey seems to defeat the advantage of having
    > an AltGr key: In non-English European language keyboard layouts of course it
    > permits access to accented characters without use of deadkeys.

    I've had the problem when designing my Latin keyboard layout (without AltGr dead keys at first) that for my uses there were simply too many different kinds of diacritics in Unicode I had to support, so eventually there was no other option than AltGr dead keys. I mean, you can support characters like grave, acute, tilde, circumflex and ring accents without too much un-intuitiveness; dieresis is already awkward, and when you start to support macron, caron, hungarianaccent, macron below, cedille, dot below, dot above, slash across, breve, ogonek and an extra key for inverted shapes like schwa, it all gets a bit crowded and inconvenient without AltGr-style dead key combinations.

    I think the best solution is to use existing keys for frequent and intuitive combinations (circumflex, tilde, grave, acute) and use AltGr for other combinations. When you want to support *really* many different diacritics, you'd probably run out of intuitive AltGr combinations otherwise.

    > I'll cc to a12n-collaboration, where there have been discussions on the Alt-
    > combo vs. deadkey solution with regard to extended character & diacritic needs
    > of African languages (which go well beyond Latin-1 in many cases).

    You might take a look at the layout from . In addition, I've got a small program lying around for converting Windows keyboard layout files to Keyman Unicode format, if anyone's interested I could upload it somewhere.


    This archive was generated by hypermail 2.1.5 : Sun Jul 18 2004 - 10:36:40 CDT