W3C home > Mailing lists > Public > www-style@w3.org > May 2009

RE: Should implementors copy vendor prefixes from each other?

From: Alex Mogilevsky <alexmog@microsoft.com>
Date: Fri, 15 May 2009 01:52:18 -0700
To: "L. David Baron" <dbaron@dbaron.org>, "www-style@w3.org" <www-style@w3.org>
Message-ID: <7C2F64B551D8664AAD94A28DAC37D0207384DF75D9@NA-EXMSG-C103.redmond.corp.microsoft.com>
It seems there are two major reasons for a prefix
a) Specification is not finalized, so using any unprefixed version is risky and not even appropriate (property names may change). Meaning of properties is clear though and is unlikely to change.
b) The implementation is more specific than currently defined behavior, which may eventually differ from the final spec. UA may indefinitely support both early prefixed and final unprefixed behavior.

I think this thread is mostly about (a). A standards spec starts that as proprietary still falls into the same category IMO. In that case, why not come up with a common "not stable" prefix, e.g. corresponding to spec status:

	-ed-opacity:50%	
	-lc-opacity:0.5	
	-cr-opacity:0.5

Here I am making up a different syntax for "editor draft" version, I don't know if a percentage was ever proposed. "CR" doesn't really make sense either, as it's time when prefixes are OK to drop.

As far as directly copying exact syntax from other vendors, it seems fine at the first sight, but aside from company's ego, such as Microsoft using "-webkit-" prefix instead of "-ms-" (which I personally have no problem with), I think cloning prefixes leads to commitment on both sides:

	- vendor who copies another vendor's prefix commits to following that other vendors spec (which may or may not be stable, and even status of which may not be known to anyone other the few people working on the feature)
	- vendor whose syntax is being copied gets additional constraints on further changes. A simple change of renaming a group of properties to be consistent with another emerging spec suddenly becomes more complicated.

I think the key for decision to use somebody else's prefix is a statement on which spec it is referring to. When I use "-ms-block-progression", if there is any difference between IE implementation and CSS3 box model, the property means whatever IE happens to do. However if it was "-ed-block-progression", the correct behavior is whatever the latest editor draft says.

Alex

-----Original Message-----
From: www-style-request@w3.org [mailto:www-style-request@w3.org] On Behalf Of L. David Baron
Sent: Thursday, May 14, 2009 4:42 PM
To: www-style@w3.org
Subject: Should implementors copy vendor prefixes from each other?

When implementors are implementing experimental properties that have
already been implemented by other vendors, should they copy vendor
prefixes from each other, or not?
Received on Friday, 15 May 2009 08:53:05 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:18 GMT