Re: Run-time checking of fonts for Sinhala support

From: Harshula (
Date: Sun Sep 06 2009 - 03:57:30 CDT

  • Next message: Asmus Freytag: "Re: Run-time checking of fonts for Sinhala support"

    Hi Asmus,

    On Thu, 2009-09-03 at 13:34 -0700, Asmus Freytag wrote:
    > On 9/3/2009 7:44 AM, Harshula wrote:
    > > On Thu, 2009-09-03 at 07:30 -0600, Doug Ewell wrote:

    > >> I agree in principle that it is pointless to use a less-capable font
    > >> when a more-capable font is available.
    > >>
    > >
    > > Glad we agree on that. :-)
    > >
    > Precisely, because it is pointless, requiring automatic validation is
    > pointless as well. Besides being expensive, and, like all technology,
    > subject to bugs and or unexpected side effects.

    Here are the two relevant scenarios:

    1) The operating system's UI needs to display a string in a particular
    script and defaults to a particular font.

    2) A user *chooses* to use a particular font in an application or in
    general across the operating system.

    I'm not concerned about scenario (2) because the user should be able to
    over-ride the defaults. They may even like a particular set of glyphs in
    an incomplete font and want to use them in a document.

    In scenario (1), there *must* be some mechanism to determine which font
    to use. Hard-coding a default font is one option, however, in the long
    term these actions will continue to occur:

    a) Incomplete and niche fonts will continue to get created.

    b) Users will continue to add and remove fonts from their operating
    system. Including the potential removal of any hard-coded default font.

    Taking these *realities* into account the operating system needs to have
    an *intelligent* way to determine if a particular font is sufficiently
    complete to be used for the UI. On some operating systems there is
    already a layer/mechanism that is responsible for doing exactly that.
    However, the mechanism is, unsurprisingly, Euro-centric.

    Just claiming it is *too hard* to address the issue is not a solution.
    There must have been some people who said it is *too hard* when they
    learnt about the goals of the Unicode Standard.

    > I've observed this exchange for a while. What I find so interesting
    > about it, is your unspoken assumption that Sinhala needs a verification
    > feature that no users of other script / language found necessary at all.
    > Yet they are all equally interested in getting the highest capabilities
    > to get their scripts / languages rendered.

    That's a bit presumptuous of you! :-) It is unspoken because I am not
    making that assumption.

    Look, I don't know how much experience, theoretical or practical, you
    have with South Asian scripts and fonts. I've gone through this process
    before with the Indic ZWJ behaviour model change. Sinhala required the
    change to generate mandatory ligatures, but the layout engines and
    applications were oblivious to the requirement. First they'd be
    sceptical and assume status quo was *good enough*, then I'd have to
    explain it to them, then provide the patches and convince them. They
    then implemented it as a Sinhala only feature. A year or two later some
    developers realised they needed the changes for other South Asian
    scripts. C'est la vie.


    This archive was generated by hypermail 2.1.5 : Sun Sep 06 2009 - 04:02:58 CDT