Re: [http-state] non-ASCII cookie values (was Re: Closing Ticket 3: Public Suffixes)

[Maciej Stachowiak <[e-mail]>]

On Feb 3, 2010, at 12:41 AM, Adam Barth wrote:

> On Wed, Feb 3, 2010 at 12:39 AM, Maciej Stachowiak <[e-mail]> wrote:
>> On Feb 3, 2010, at 12:11 AM, Julian Reschke wrote:
>>> Adam Barth wrote:
>>>> ...
>>>> Those cases being charset0001, charset0002, and charset0003.  One
>>>> possible explanation is that Safari rejects cookies that have
>>>> non-ASCII characters in their values.
>>>> ...
>>> 
>>> ...which I think is not really a bug, according to the specs.
>>> 
>>> Unless that affects Safari's interoperability, I'd strongly recommend to consider making this the recommended behavior.
>> 
>> I don't have information one way or another on this. I don't see any bugs that are obviously this issue, but it could be the cause of less obvious compatibility bugs. Perhaps the thing to do is a study to determine the frequency with which non-ASCII cookies are sent.
> 
> Do you see a particular benefit to dropping non-ASCII cookies?  I
> don't understand what we're trying to gain here.

I don't care much either way, and it would not be hard to change our implementation. I'm just trying to clarify the level of evidence we have on whether this affects compat. Other things being equal, I would prefer to see Safari change to match the other browsers.

I will note that Safari (and Mac OS) are pretty rarely used in some asian markets, such as China and India, so there could be issues we are unaware of. iPhone has significant penetration into some of those markets however.

Regards,
Maciej