- From: Cosimo Lupo <cosimo.lupo@daltonmaag.com>
- Date: Thu, 21 May 2015 15:19:06 +0100
- To: Levantovsky, Vladimir <Vladimir.Levantovsky@monotype.com>
- Cc: Jonathan Kew <jfkthame@gmail.com>, Esfahbod Behdad <behdad@behdad.org>, "public-webfonts-wg@w3.org" <public-webfonts-wg@w3.org>
Hello, thank you both for the clarifications. I just made a “woff2_hmtx” branch of Behdad’s fonttools to test the hmtx transformation. (Doing this in Python is much quicker than C/C++!). If you want to try it out, just clone it from here: https://github.com/anthrotype/fonttools/tree/woff2_hmtx I followed Jonathan’s suggestion. In the WOFF2 table directory, for hmtx I write only the transformLength, instead of the origLength. Then the decoder tests whether the reported lenght is equal to the calculated length of a “full” hmtx table (based on maxp.numGlyphs and hhea.numberOfHMetrics), or if it is equal to the length of a "short" hmtx table (without sidebearings). I will run some tests on the corpus of fonts which I have available here at DaMa, and will let you know how much the average gain is in terms of file size. All best, -- Cosimo Lupo, Font Design, Dalton Maag Ltd 9th Floor, Blue Star House, 234-240 Stockwell Road, London, SW9 9SP, UK Mobile: +44 7825 324360 London Office: +44 20 7924 0633 Registered office: Mutfords, Hare Street, Buntingford, SG9 0ED, UK Registered in England and Wales: 3103619
Received on Thursday, 21 May 2015 14:19:36 UTC