W3C home > Mailing lists > Public > www-font@w3.org > July to September 2009

Re: EOT-Lite File Format

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Thu, 30 Jul 2009 18:36:36 -0500
Message-ID: <dd0fbad0907301636t4780e4aavf4603997165c45fc@mail.gmail.com>
To: Thomas Lord <lord@emf.net>
Cc: Sylvain Galineau <sylvaing@microsoft.com>, "robert@ocallahan.org" <robert@ocallahan.org>, John Daggett <jdaggett@mozilla.com>, www-font <www-font@w3.org>
On Thu, Jul 30, 2009 at 6:28 PM, Thomas Lord<lord@emf.net> wrote:
> On Thu, 2009-07-30 at 18:22 -0500, Tab Atkins Jr. wrote:
>> On Thu, Jul 30, 2009 at 6:11 PM, Thomas Lord<lord@emf.net> wrote:
>
>> > That suggests a SHOULD requirement.  UAs SHOULD ignore
>> > non-nil root-strings but are not obligated to do so.
>> > Authors can't count on them being ignored on the one
>> > hand but UA makers are encouraged to ignore them
>> > entirely.
>>
>> Nope, it has to be a MUST requirement - UAs MUST ignore non-nil
>> rootstrings.  IE <= 8 browsers will just be nonconforming (which is
>> fine, since they were produced before this standard was produced), and
>> authors can take advantage of that to hack something resembling
>> same-origin into it if they wish.
>
> You are a braver man than I, in this matter.
> Only... I'm not at all sure you're wrong.
> Still, I'll offer my case:
>
> I suggest "SHOULD" because that would allow
> "IE<=8" to be retroactively conforming.  Of course,
> if they are not retroactively conforming, then
> what's the point?  Take EOT-lite off the table,
> in that case.

Technically conforming and practically conforming are two different things here.

We want a MUST requirement so that we have interop for future
browsers, and users don't put in rootstrings expecting them to work.
A SHOULD invites non-interop here, which creates the possibility that
an author will put a rootstring in with a different domain and try to
load it from that domain, test it in a browser that *does* pay
attention to rootstrings, and conclude everything is fine.

IE<=8 will not technically conform to this, but that's okay.  A
well-formed EOTL file (no rootstring) will work great in them, and the
chance that an author will test only in one of the bad IEs will
continuously decline as those browser versions phase out.  Authors
update versions and use non-IE browsers in *much* higher proportions
than the average populace.  So as a practical matter, those browsers'
nonconformance to that particular line isn't very important.

~TJ
Received on Thursday, 30 July 2009 23:37:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 11 June 2011 00:14:03 GMT