Re: Browser implementations, prior to rec, used for justification

On Tue, Jan 5, 2010 at 5:58 AM, Julian Reschke <julian.reschke@gmx.de> wrote:
> James Graham wrote:
>>
>> On 05/01/10 14:31, Shelley Powers wrote:
>>
>>> And autobuffer is from which released specification, where we have to
>>> worry about legacy use?
>>
>> Usage is a matter of quantifiable fact, not a matter of W3C Rec track
>> status. If it is agreed that the legacy implied by released Firefox makes
>> the attribute name "autobuffer" unsuitable to resuse at this time it is
>> strictly irrelevant whether the legacy came from following a "released
>> specification", proprietary invention, an unintended bug, or an amazingly
>> improbable set of cosmic-ray induced bit flips on the build machine.
>
> Agreed by whom?
>
> Firefox can be updated easily.

I'm surprised by your level of confidence here. What are you basing
the above statement on?

What we can and can not put in a dot release is a very complicated
matter. For example we are responsible towards distributors that have
very conservative views on what is appropriate to put in a dot
release. Our by far over reaching goal with dot releases is to make
people more secure. If there's a risk that a behavioral change breaks
even a small number of websites we risk that people choose not to
install a dot release in order to keep their used websites working. We
already have much bigger problems than we'd like to get people to
upgrade to the latest dot release.

So unless you have talked to firefox people out of band about this
specific issue, I would change the terminology from "firefox can" to
"we should check if firefox can".

/ Jonas

Received on Tuesday, 5 January 2010 18:15:47 UTC