RE: [My feelings about Vrinda] was: Bogous Microsoft's "Vrinda" font (v1.40) for Bengali+Latin in Vista RC1

From: Shariqul Islam Azad - Omi (omiazad@msn.com)
Date: Mon Oct 16 2006 - 06:30:38 CST

  • Next message: Simon Daniels: "RE: Bogous Microsoft's "Vrinda" font (v1.40) for Bengali+Latin in Vista RC1"

    Thanks for bringing this issue to attention. Since XP SP-2 when I first saw the font, I'm saying that the font is not user friendly for Bengali speaking people. But Microsoft is not interested to change the look. The font is not visible indeed, but they said everything will be fixed with Vista. The fact is they didn't update the font's look, they just change the visual engine, so that it can present ugly looking fonts in a better way (as Linux does).For experimenting, just turn clear type off and compare Vrinda texts besides Mangal. :)The look of Bangla is not user friendly at all. Once someone from MS told me that the look of Vrinda is going to be changed and after seeing the sample I was very happy. (You can check that out from here) but later I learn that they are not going to pay MonoType for using the glyphs and whatever you like it or not you HAVE TO use the UGLY Bangla for ever.I'm sad to know that there are more issues with Vrinda. Let's wait and see!!!--Shariqul Islam Azad - OmiMicrosoft MVP - Windows Shell/UserContributorBangla Computing and Localization Projects:Ekushey: http://www.ekushey.org> From: verdy_p@wanadoo.fr> To: unicode@unicode.org> Subject: Bogous Microsoft's "Vrinda" font (v1.40) for Bengali+Latin in Vista RC1> Date: Mon, 16 Oct 2006 05:28:27 +0200> > Microsoft is currently distributing with Windows Vista RC1 a really bogous update of the "Vrinda" openType font for supporting Bengali+Latin.> > The problem is not much in the Bengali script itself, but in the many incorrect mappings this font makes in other scripts: Latin, Greek, mathematical symbols.> > There was no such bugs in version 1.03 of this font (for Windows XP, copyrighted 2004), but the problem appears immediately with the version 1.40 of the font (currently shipped with Windows Vista RC1, copyrihted 2006, and signed by Ragunath Joshi and Vinay Saynekar).> > For example, this new version 1.40 maps:> * a Latin small letter thorn glyph where it should be a Latin small letter dotless i (U+00FD), or no mapping at all> * a spacing candrabindu glyph where it should be a spacing caron (U+02C7)> * a whitespace where it should be a spacing ogonek (U+02D8), or no mapping at all> * a vertical line glyph where it should be a spacing dot above (U+02D9), or no mapping at all> * a vertical line with a cross at top end glyph where it should be a spacing ring above (U+02DA), or no mapping at all> * a double danda glyph where it should be a spacing ogonek (U+02DB), or no mapping at all> * a danda glyph where it should be a spacing double accute accent (U+02DC), or no mapping at all> * a Latin capital letter E with diaeresis glyph where it should be a Greek small letter pi (U+03C0), or no mapping at all> * a Latin capital letter I with diaeresis glyph where it should be a Ohm unit symbol (U+2126), or no mapping at all> * a Latin capital letter E with grave glyph where it should be a partial differential mathematical symbol (U+2202), or no mapping at all> * a Latin capital letter U with acute glyph where it should be an increment mathematical symbol (U+2206), or no mapping at all> * a Latin capital letter E with circumflex glyph where it should be a family product mathematical symbol (U+220F), or no mapping at all> * a Latin capital letter E with acute glyph where it should be a family summation mathematical symbol (U+2211), or no mapping at all> * a Latin capital letter O with diaeresis glyph where it should be a square root mathematical symbol (U+221A), or no mapping at all> * a Latin capital letter A with circumflex glyph where it should be a infinite mathematical symbol (U+221E), or no mapping at all> * a Latin capital letter I with grave glyph where it should be a integration mathematical symbol (U+222B), or no mapping at all> * a Latin capital letter U with grave glyph where it should be a approximately equal to mathematical symbol (U+2248), or no mapping at all> * a reversed question mark where it should be a not equal to mathematical symbol (U+2260), or no mapping at all> * a Latin capital letter A with diaeresis glyph where it should be a less than or equal to mathematical symbol (U+2264), or no mapping at all> * a Latin capital letter A with ring above glyph where it should be a greater than or equal to mathematical symbol (U+2265), or no mapping at all> * a Latin small letter i with acute glyph where it should be a lozange geometric form (U+25CA), or no mapping at all> * a Latin capital letter O with tilde glyph where it should be a Latin small ligature fi (U+FB01), or no mapping at all> * a Latin capital letter O with diaeresis glyph where it should be a Latin small ligature fl (U+FB02), or no mapping at all> > There may be other incorrect mappings that i did not check ; if ever this fonts is within some documents or stylesheet, it will make havoc in documents that use the the codepoints above.> > I hope this is just a beta, and that this font won't be released as such with the superfluous extra mappings within the final release of Vista, because it creates nightmares in CSS stylesheets for multilinguage texts. It tooks me some time to figure out what was wrong in some of my documents rendered in Vista, after scrutinizing the complex stylesheets that contain references to lots of fonts.> > In addition, it seems that this font declares the wrong support for various scripts, and this causes lots of caveats with CSS. The workaround is to replace the Vista RC1 version (1.40) by the XP version (I have version 1.03 of this font).> > >



    This archive was generated by hypermail 2.1.5 : Mon Oct 16 2006 - 06:39:28 CST