Re: Run-time checking of fonts for Sinhala support

From: Roozbeh Pournader (roozbeh@htpassport.com)
Date: Wed Aug 26 2009 - 14:29:03 CDT

  • Next message: announcements@unicode.org: "[Unicode Announcement] 33rd Internationalization & Unicode Conference - Keynote Speaker Announcement"

    On Thu, 2009-08-27 at 02:56 +1000, Harshula wrote:
    > How do you suggest an operating system determine if a font is a valid
    > Sinhala font? Level 1 compliance, as specified in SLS1134:Part2,
    > prescribes the minimum set of requirements for a SLS1134 compliant
    > Sinhala font. I would think it is only logical to comply with it.

    This is very solution-specific. It would really depend on what an
    operating system (or a font system, or a rendering engine) consider its
    model to be for language-support detection.

    For example, let's say a text rendering engine of an operating system is
    asked to display a specific piece of Sinhala text in a certain font. It
    can just go and try to render them as is. Or if it wants to be more
    intelligent, it can first check if the font (with its extra information
    like OpenType tables) supports all the characters and the character
    combinations it's been asked to display. If not, it may use a fallback
    method, like using another font it knows to support Sinhala (I believe
    Microsoft Windows hardcodes some such fallback fonts), or it may display
    missing glyphs as boxes. Both look acceptable to me as far as they are
    properly documented.

    On a side note, I believe that certain parts of SLS 1134:Part 2:2007,
    are in contradiction with certain parts of SLS 1134:2004. But that's a
    different story.

    > So, if SLS1134:Part2 stated something like 'Operating systems shall only
    > recognise Level 1 compliant fonts as Sinhala fonts.', would that be
    > sufficiently explicit in your mind?

    Is there any such text in the standard? I may have missed it.

    You had written:

    "Operating systems should, at a minimum, check that a Sinhala font meets
    these requirements before using it."

    And I countered that this is your personal opinion.

    Re-checking the actual standard, a correct claim could have been:

    "Fonts claiming to conform to SLS 1134:Part2, should support all the
    characters and character combinations mentioned in the standard for
    Level-1 support."

    Or:

    "'Computer systems preloaded with Sinhala' claiming to conform to SLS
    1134:2004 should include at least one Sinhala font, preferably
    conforming to SLS 1134:2004 Level-1 font requirements."

    Or:

    "Software that is used for testing fonts for SLS 1134:2004 conformance
    should check fonts for at least SLS 1134:Part 2 Level 1 support."

    Nowhere in the standard there is even a hint that running software, to
    be used by an everyday user of Sinhala, even if the software claims SLS
    1134:2004 conformance, should check for Level-1 support before using a
    Sinhala font it's been asked to use. (Or at least that's my opinion.)

    And the standard's scope doesn't even cover operating systems
    explicitly. The nearest it goes, is 'computer systems preloaded with
    Sinhala'. Then reading section 4.3 of SLS 1134:Part2:2007, which is
    about those preloaded systems, one finds that they are not supposed to
    check for the conformance of fonts at run time. They are only required
    to have one Sinhala font that conforms to the fonts requirements (of
    which Level-1 is a minimum).

    Roozbeh



    This archive was generated by hypermail 2.1.5 : Wed Aug 26 2009 - 14:34:50 CDT