W3C home > Mailing lists > Public > public-webfonts-wg@w3.org > February 2011

RE: Minutes, 16 February 2011 WebFonts WG telcon

From: Sylvain Galineau <sylvaing@microsoft.com>
Date: Sat, 19 Feb 2011 18:18:59 +0000
To: "Levantovsky, Vladimir" <Vladimir.Levantovsky@MonotypeImaging.com>, Håkon Wium Lie <howcome@opera.com>
CC: "public-webfonts-wg@w3.org" <public-webfonts-wg@w3.org>
Message-ID: <045A765940533D4CA4933A4A7E32597E2AB6269C@TK5EX14MBXC120.redmond.corp.microsoft.com>
Thinking about this a little more, and since the web's 
consistency has been brought up several times lately, 
it is probably worth elaborating upon.

The reason From-origin:same will be set on licensed 
commercial fonts hosted by the licensee has nothing to do 
with 'the web'. As such, the so-called exception for font 
resources will occur at run-time on the web whatever browser 
vendors do. So the issue is not really about 'the web', it's 
a simple practical one: given that this header value will 
have to be set so often for font resources, where/when is 
it easiest and cheapest to set it ?

I don't believe the web to be cleaner when every author, 
hosting service or app server maker have to expend additional 
time, energy and effort configuring a default that browsers 
can implement at minimal cost (especially since the bulk of 
the engineering cost is in implementing From-origin for all 
file types, not in setting a default for @font-face). Putting 
this burden on the community sounds more like arbitrary web 
bureaucracy than architecture. 

At the very least, it can be argued to contradict other web 
design principles such as HTML's [1]:  

"In case of conflict, consider users over authors over 
implementors over specifiers over theoretical purity". 

(Note: Anne and Maciej co-authored this document)

If implementations' runtime cleanliness and architectural 
purity are the priority for the web then raw font resources 
loaded from anywhere are ideal. And from that point of view 
it will be argued - and it has been, most often by Opera 
staff - that overall WOFF *is* an exception to the general 
pattern. Agreeing to such an exception and then arguing
we should pretend it isn't one feels like sitting between 
two chairs (as in the furniture, not the WG folks; although
we probably are doing that too).

I guess I'm really wondering why Opera bothers with WOFF at all. 


[1] http://www.w3.org/TR/html-design-principles/#priority-of-constituencies


> -----Original Message-----
> From: Levantovsky, Vladimir
> [mailto:Vladimir.Levantovsky@MonotypeImaging.com]
> Sent: Saturday, February 19, 2011 8:25 AM
> To: Håkon Wium Lie; Sylvain Galineau
> Cc: public-webfonts-wg@w3.org
> Subject: RE: Minutes, 16 February 2011 WebFonts WG telcon
> 
> On Friday, February 18, 2011 7:37 PM Håkon Wium Lie wrote:
> >
> > Also sprach Sylvain Galineau:
> >
> >  > > It seems cleaner to treat all URLs the same way.
> >  >
> >  > Cleaner for whom ?
> >
> > The web.
> >
> 
> "The web"? The web are the people (authors, developers, and yes, font
> vendors too) who create content. A solution that makes them jump through
> hoops all the time could hardly be considered _clean_; to the contrary -
> the default behavior where things just happen the right way with no
> additional efforts is what I believe "the web" would want.
> 
> I've heard it on many occasions and from different people that, in the
> hindsight, SOR should have been implemented from the beginning of the web,
> and that not restricting links to the same origin was a mistake. Now, the
> argument that is frequently brought up is "consistency" - I can hardly see
> why consistently making the same mistake would be a benefit, especially in
> the situation when we have a chance to do things the right way (as the
> existing implementations already proved) without impacting anything else
> on the web.
> 
> Thank you,
> Vlad
> 
Received on Saturday, 19 February 2011 18:19:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 19 February 2011 18:19:36 GMT