Additional thoughts on TRANSITIONAL

Shawn Steele Shawn.Steele at microsoft.com
Fri Dec 4 23:27:52 CET 2009


The concensus amongst "the browser folks" seems to be that you'd have to impliment IDNAbis AND TR46 together.  That is necessary because IDNAbis isn't very compatible with mappings, regardless of those 4 characters.  Whatever behavior for those 4 characters is agreed upon for UTS46 is what we would implement.

I would not implement IDNAbis without UTS46 (as it stands), and it sounds like the other "browser folks" agree as well.  That still seems standards compliant to me.  If the other browser folks changed their positions I'd have to reconsider, but I think consistent mapping is critical.

In fairness, I should note that there's one small thing I don't like.  Migration from IDNA2003 to IDNA2008 is a pain (as mentioned), however most of the problem could be solved entirely by updating our tables.  There is the small matter of the BIDI rules being hard-coded however, which is MUCH harder for me to argue for servicing.  It is likely that I'll remove the extra hard-coded check to make future servicing easier (just data updates)  The slight additional exposure is mitigated by the fact that any lookup for those strings should fail, and IE's SmartScreen to catch any troublemakers that slip through.  Given that BIDI produces interesting results in an IRI I think it's likely that these rules, or at least BCP for BIDI IRIs will change.  Certainly what IE does with mixed RTL/LTR labels isn't very good :)

-Shawn

________________________________________
From: idna-update-bounces at alvestrand.no [idna-update-bounces at alvestrand.no] on behalf of Erik van der Poel [erikv at google.com]
Sent: Friday, December 04, 2009 12:21 PM
To: Shawn Steele
Cc: Andrew Sullivan; idna-update at alvestrand.no; Gervase Markham
Subject: Re: Additional thoughts on TRANSITIONAL

If the current drafts of IDNAbis and TR46 were published as RFCs and a
UTS, respectively, which would MSIE implement? They treat the infamous
4 characters differently.

(Maybe I'm pushing too hard and should stay out of this conflict
between IETF and Unicode. Maybe I should go back to
reverse-engineering the browsers and see what actually happens in the
real world. I am very, very tired of this WG process.)

Erik

On Fri, Dec 4, 2009 at 11:59 AM, Shawn Steele
<Shawn.Steele at microsoft.com> wrote:
> ??? I wasn't suggesting that IE would avoid the standard.  I'm saying it will take a while.  FF, Safari, Chrome & Opera don't patch previous versions, they expect users to install the latest & greatest versions.  IE does too, but you can see from the adoption rates that it's difficult to get everyone moved very quickly.
>
> In particular I very much advocate standards, so I'm a bit surprised by your remark.  Just to be clear:  None of my positions have been just because that's what I think is "easy" for IE.  I am well-known within the company for arguing what's right for the end-user, regardless of how difficult it is on our existing code base.
>
> I can't patch IDN and get it to all our customers in a month or two, it's just not possible.
>
> -Shawn
>
> ________________________________________
> From: Erik van der Poel [erikv at google.com]
> Sent: Friday, December 04, 2009 11:48 AM
> To: Shawn Steele
> Cc: Andrew Sullivan; Gervase Markham; idna-update at alvestrand.no
> Subject: Re: Additional thoughts on TRANSITIONAL
>
> In the past, we have seen browsers like Firefox, Safari, Chrome and
> Opera implement various specs rather than trying to be 100% compatible
> with MSIE. I wonder if this (IDNAbis) is one of those occasions where
> Firefox will stick to the "official" spec (IETF), hoping to eventually
> drag MSIE, kicking and screaming, into the "modern age"? Gerv?
> Anybody?
>
> Erik
>
> On Fri, Dec 4, 2009 at 11:26 AM, Shawn Steele
> <Shawn.Steele at microsoft.com> wrote:
>> Just sent, but I think transitional that breaks current IDNA2003 sites that work is really bad.  I also think anything that builds-in a delay of 5-10 years to get correct language support is also really bad.
>>
>> -Shawn
>>
>> ________________________________________
>> From: Erik van der Poel [erikv at google.com]
>> Sent: Friday, December 04, 2009 11:11 AM
>> To: Andrew Sullivan; Shawn Steele; Gervase Markham
>> Cc: idna-update at alvestrand.no
>> Subject: Re: Additional thoughts on TRANSITIONAL
>>
>> Yes, I'm hoping that we are willing to explore different ways out of
>> this mess. We currently have two different proposals for IDNA:
>>
>> https://datatracker.ietf.org/drafts/wg/idnabis/
>> http://www.unicode.org/reports/tr46/
>>
>> In particular, I'd like to hear from Microsoft and Firefox folks, on
>> their thoughts regarding the TRANSITIONAL ideas.
>>
>> Erik
>>
>> On Fri, Dec 4, 2009 at 10:58 AM, Andrew Sullivan <ajs at shinkuro.com> wrote:
>>> On Fri, Dec 04, 2009 at 09:16:02AM -0800, Erik van der Poel wrote:
>>>> some point in the future, when there is consensus that it is the right
>>>> time (and the right thing to do).
>>>
>>> Perhaps I'm am growing cynical with age, but given the difficulty of
>>> converging now, I have doubts about the chances of converging in the
>>> future, when there is even more deployed stuff depending on existing
>>> behaviour.  Still, it might work, and I'm by no means saying no.  Any
>>> route out of this twisty little maze of arguments, all repeated, is
>>> one I'm willing to explore.
>>>
>>> A
>>>
>>> --
>>> Andrew Sullivan
>>> ajs at shinkuro.com
>>> Shinkuro, Inc.
>>> _______________________________________________
>>> Idna-update mailing list
>>> Idna-update at alvestrand.no
>>> http://www.alvestrand.no/mailman/listinfo/idna-update
>>>
_______________________________________________
Idna-update mailing list
Idna-update at alvestrand.no
http://www.alvestrand.no/mailman/listinfo/idna-update


More information about the Idna-update mailing list