W3C home > Mailing lists > Public > www-style@w3.org > May 2012

RE: Vendor-prefixes: an idea

From: Brian Manthos <brianman@microsoft.com>
Date: Tue, 8 May 2012 01:57:59 +0000
To: Tab Atkins Jr. <jackalmage@gmail.com>, Brad Kemper <brad.kemper@gmail.com>
CC: François REMY <fremycompany_pub@yahoo.fr>, CSS 3 W3C Group <www-style@w3.org>
Message-ID: <4031779A10373C4AB7B9FBAD218A8BAC0242508A@CH1PRD0310MB391.namprd03.prod.outlook.com>
As much as Tab is apparently willing to take the heat on this one, there was more to the story.  We could all dig through the archives if we'd like but, relative to the current discussion...

One of the key parts of that decision making process was the CSSWG mindset at the time of the roles of and relationship between prefixed and unprefixed support in implementations.

For some of the proposals in this thread, if they had been the CSSWG model then Tab (and others) wouldn't have made the change in the same way (and perhaps not at all).

In short, looking at past examples is useful for evaluating *those models* not for re-litigating past decisions with a new set of rules.


That said...

> prefixed gradients (Firefox has updated theirs compatibly.)

Brad's point is one reason why I chuckled to myself when I read this, as I'm curious how they solved the unsolvable.

Another example is corner-to-corner gradients which have different renderings for the same input in the CR vs. -ms-linear-gradient.

-Brian

-----Original Message-----
From: Tab Atkins Jr. [mailto:jackalmage@gmail.com] 
Sent: Monday, May 07, 2012 5:34 PM
To: Brad Kemper
Cc: François REMY; CSS 3 W3C Group
Subject: Re: Vendor-prefixes: an idea

On Tue, May 8, 2012 at 2:31 AM, Brad Kemper <brad.kemper@gmail.com> wrote:
> On May 7, 2012, at 7:54 AM, "Tab Atkins Jr." <jackalmage@gmail.com> wrote:
>> The only failure-mode is if some of
>> the old syntax overlaps the new syntax, but has a different meaning 
>> (like what linear-gradients did with <angle>s).  The simple solution 
>> is "don't do that".  If you absolutely *must* do that, just change 
>> the property name.
>
> At the time, I begged that we just don't do that, without also changing the syntax enough so that the previous version could still be there in a style sheet and ignored in newer UAs. I was told that the rest of WG would not give any sort of consideration of what happened with experimental prefixed values when determining how the value should work in a later draft.

Call it a rookie mistake on my part.  (Because it was a mistake.)

~TJ




Received on Tuesday, 8 May 2012 01:59:50 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:53 GMT