Re: vertical direction control

From: Asmus Freytag (
Date: Wed Mar 24 2004 - 18:08:31 EST

  • Next message: Peter Kirk: "Re: vertical direction control"

    At 02:58 PM 3/24/2004, Thomas Kuehne wrote:
    >Am 2004-03-23 20:23 schrieb Asmus Freytag:
    > > I don't think I know of a scenario where it is crtical for a
    > > resource limited device to display the kinds of texts you list
    > > below.
    >Reading the font data and processing it into a display representation
    >poses the same resource costs for mirrored, rotated and "normal"
    >The limitations are mainly processing speed and memory/storage. On the
    >display side there are few potential problems.

    Unless you eventually display text, you don't need to deal with direction
    control. That's why I used the term 'display' in posing my question earlier.

    Let me rephrase it: I don't think I know of a scenario where it is
    critical for a resource limited device to support the selection of vertical
    vs. non-vertical display of text.

    Unless there is some such scenario, there's no need to even speculate
    about alternatives to existing markup languages. I have not seen a
    credible scenario.

    Bidirectional controls are different: Even on the most limited device
    displaying text, for example for text messages on a cell phone, you
    need to run the bidirectional algorithm for bidi scripts. As the algorithm
    uses heuristics, some form of additional controls are needed to disambiguate
    certain cases. Many 8-bit sets support only a subset of these, i.e. RLM/LRM,
    but not the additional set of stateful controls.

    See for example UTR#20 ( why having
    control codes for these purposes can cause issues in migrating text between
    plain text and marked-up text.


    This archive was generated by hypermail 2.1.5 : Wed Mar 24 2004 - 18:52:01 EST