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

RE: A way forward

From: Thomas Lord <lord@emf.net>
Date: Fri, 24 Jul 2009 15:47:21 -0700
To: Sylvain Galineau <sylvaing@microsoft.com>
Cc: Håkon Wium Lie <howcome@opera.com>, John Hudson <tiro@tiro.com>, www-font <www-font@w3.org>
Message-Id: <1248475641.7681.46.camel@dell-desktop.example.com>
On Fri, 2009-07-24 at 21:59 +0000, Sylvain Galineau wrote:
> > From: Thomas Lord [mailto:lord@emf.net]
> > Sent: Friday, July 24, 2009 2:41 PM

> > I suggest a different framing of the issue as two
> > questions:

> You and I can frame in different ways, it does not follow
> that others will follow you framing.

Indeed, it does not follow but we can still
explore my suggestion, as you go on to do:

> > (a) Does there exist some (any) reasonabler variation
> > of EOT-lite for which, if the other browsers implement that
> > support, the other browsers and existing versions
> > of IE will all do the same, useful thing?

> I have seen no evidence that there isn't. 

Neither have I.  I believe the opposite is the case
but I can't prove it.

> I am seeing concern over
> known IE interoperability bugs and I absolutely acknowledge authors
> will deal with them for some time. But as this concern is orthogonal
> to the underlying file format being served, I don't understand why
> it matters in assessing Ascender's proposal.

It matters this way: it *might* be the case that
restricted license fonts would mainly be licensed
mainly for a form of use that relies on misfeatures
(or bugs) of older versions of IE.  People on the
"other side" could not in good conscience sign for
a Recommendation that immortalized those bugs.  As
a practical matter, the other browser implementers
don't want to be put in a position of having to 
implement those bugs as a practical matter, even if they
are not officially a Recommendation.  Håkon, et al.
are not evading your questions at all - they are 
responding by voicing that concern.   If there is a good
answer to them, I think it is by more clearly identifying
the interop standard that is both downwards compatible
with IE and then going around the table and obtaining
a general consensus (especially among the type vendors
in question) that everyone can live with that standard.

It's not fair to overburden MSFT in this and so I 
don't want to say MSFT is *obligated* to do the following
but if MSFT were to do the following, it might help:
work up some Firefox patches (or Opera or whatever) and
demonstrate the reasonable proposed Recommendation.
To be clear, I don't think it is in any way incumbent
upon MSFT to do so - only that it *might* be something 
MSFT doesn't mind doing and that doing it can 
certainly help the process along.



> Hakon can't both tell me our quirks will force him to emulate legacy
> behavior if he support EOT-Lite then suggest we implement raw TTF/OTF
> support as if the exact same constraint did not apply.


His suspicions appear from my perspective to be based
on repeated experience in other areas of browser 
behavior.  He is asking, I think, why should I expect
fonts to go any differently?  

I don't know that any concept of which side has
"burden of proof" really applies here.  I do think
that your side can demonstrate proof for your position,
albeit with a little work to create some patches for
one of the other browsers (and some pages to show off
the consequence).  Just a thought.


> If this were the real issue blocking adoption of EOT-Lite then no
> other proposal on the table will address it. Changing the file format
> does not fix IE parsing bugs and interop quirks.

The main motive for EOT-lite from the other side
is the potential for downward compat. with existing
versions of IE.  If that compatibility is tainted by
quirks, the motive is gone and Håkon is right.


> > (b) Will restricted license type vendors agree to
> > license in that variant?
> 
> That is also John's question and it is a fair one. That is up
> to font vendors to answer. At least one - Ascender - has committed
> publicly so far and even developed tooling.\

Yes, that looks interesting promising.  It is 
part of why I'm trying to help out your side 
here.



> Authors have expressed interest on this list and others. A fair hearing
> is all I ask for.

This dialog is not chopped liver, so to speak.




> > If the answer to both questions is "yes" then
> > that is a strong argument for that variant of
> > EOT-lite.
> >
> > If the answers are (a) - yes and (b) - no
> > then that is a weaker but still positive argument
> > for that variant of EOT-lite.
> >
> > If the answer has (a) - no, then there is no
> > point to even considering any variant of EOT-lite.
> 
> That's a fine argument. Note that all I oppose here is Hakon's latest argument against EOT-Lite.
> I find flat-out misleading, and quite possibly misinformed if not dishonest.


Let's all get religion here.  I called MSFT bullies, and
that was not helpful.  You call Håkon evasive and perhaps
dishonest, and that is not helpful.  Let's both do better.

I think there's a solid argument that if MSFT works
up a patch for firefox to demonstrate an EOT-lite option
and if your colleagues on the other side can study that
and convince themselves they won't get caught in a 
"quirk trap" then it will be hard to justify refusing
the patch and hard to justify not making the feature
part of a Recommendation.  And I think there are solid arguments
that that is a rational strategy for improving the value 
of IE vis a vis MSFT shareholders.   It's an outrageous
thing to ask and so I'm not asking - I'm just mentioning
that the option exists for you.




> But however wrong Hakon's latest shtick may be, it does not follow EOT-Lite can work in practice.
> If, as John pointed out, the font vendors' EULAs do require same-origin checks then IE can't comply.


That's one of the reasons we have to go around the
table here, with special attention on type vendor's
feelings about the option.


> Which throws all potential interop benefits out of the window. Now *this* is an argument against EOT-Lite
> I can live with. Provided, of course, we have the evidence to call it.


"Talk amongst yrselves," I think (meaning MSFT and it's
close 3rd party  type vendors).  I agree that that's 
a potentially very solid argument against EOT-lite.


> 
> Claiming EOT-Lite is bad for the web because Opera's
> @font-face would have to match IE's behavior is
> plainly nonsensical. Or theater. I don't really know which.


Neither.  It's reasoning by sensible analogy to past experience
in different but similar areas.

-t
Received on Friday, 24 July 2009 22:48:01 GMT

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