- From: Jonathan Kew <jonathan@jfkew.plus.com>
- Date: Tue, 17 Aug 2010 10:12:42 +0100
- To: John Daggett <jdaggett@mozilla.com>
- Cc: Thomas Phinney <tphinney@cal.berkeley.edu>, "Tab Atkins Jr." <jackalmage@gmail.com>, Behdad Esfahbod <behdad@behdad.org>, www-style@w3.org, Brad Kemper <brad.kemper@gmail.com>
On 15 Aug 2010, at 15:49, John Daggett wrote:
> Thomas Phinney wrote:
>
>> Synthesized stuff (or transformations) are not the same as designed
>> condensed or extended fonts. I would ask two things:
>>
>> - that you do not *encourage* agents to synthesize font-stretch
>> ("allow" is okay I guess, though you should understand the results
>> will always be crap)
>>
>> - that for any agent that does do synthetic stretch, it is OFF by
>> default and turned on by some additional flag or option.
>
> I agree, I think any stretchy/squishy text effects should be in done via
> properties separate from the font selection process, properties clearly
> defined with testable behavior. Selecting a font should be different
> from applying an effect.
Ok, that seems fair enough. It's a pity "font-stretch" is taken for the font selection property ("font-width" might have been better?), but it's much too late to change that, no doubt.
In that case, I suggest we should have a new property called "font-extend" that would take a scaling factor to be applied to the glyphs and metrics in the x-direction (or rather, along the direction of text layout, so for vertical text it would apply in the y-direction). A font-extend value of less than 1.0 would actually compress the glyphs, of course. And negative values ought to reflect the glyphs and reverse the text direction, so that <span style="font-extend: -1.0">mirror writing</span> becomes easy to express.
Suggestions of better names are welcome.
JK
Received on Tuesday, 17 August 2010 09:13:29 UTC