Re: [css3-fonts] font-size-adjust auto issue

On 26/8/13 19:01, Levantovsky, Vladimir wrote:

> My
> objection to having <auto> value is because it encourages authors to
> use it blindly, not knowing (and having no chance to know precisely)
> what the effect will be on any given platform.

I agree with Vlad here. I think the <auto> value, as currently defined, 
is of very limited utility (given that the aspect value of the default 
font does in fact vary considerably between browsers and platforms); and 
I suspect it will be blindly (mis)used far more often than it is helpful.

A much more intuitive and useful - for authors - meaning of <auto> 
would, I think, be "the aspect value calculated by the UA for the first 
available face in the current font-family list"; in other words, it 
would never cause any rescaling of the first available font, but if 
there is any text that falls back to a later font in the list (or an 
arbitrary system font), *that* font will be adjusted so that its 
x-height matches the first.

A possible objection to this would be that if the first font is a 
webfont, its aspect value cannot be calculated until it has been 
downloaded, and so <auto> would initially have to be based on a font 
that's later in the list. This means that text rendered using a fallback 
font might suddenly change size when the webfont download completes and 
the <auto> aspect value is recalculated. This would perhaps be more 
disconcerting than the re-layout that can already happen as a result of 
a font download finishing.

Of course, an author who wants to avoid this can always do so by 
providing an explicit value for font-size-adjust, rather than relying on 
the <auto> magic.

Alternatively, I still think my earlier suggestion of allowing 
font-size-adjust:<generic-family> would be an improvement over <auto>, 
in that it would be clearer to authors that they're explicitly asking 
the browser to adjust the sizing of whatever font(s) are used to better 
match one of its defaults.

JK

Received on Tuesday, 27 August 2013 02:59:37 UTC