W3C home > Mailing lists > Public > www-style@w3.org > November 2011

Re: Unprefixing CSS properties

From: Bjoern Hoehrmann <derhoermi@gmx.net>
Date: Sat, 19 Nov 2011 00:57:22 +0100
To: Sylvain Galineau <sylvaing@microsoft.com>
Cc: "robert@ocallahan.org" <robert@ocallahan.org>, "www-style@w3.org" <www-style@w3.org>
Message-ID: <puqdc79s84gv0oq1klnr3ec0co8obi21ui@hive.bjoern.hoehrmann.de>
* Sylvain Galineau wrote:
>So I'd love to hear more about specific examples and supporting data. I
>am sympathetic to the argument that a bad versioning scheme can hurt the
>web. Or that it has caused pain and may do so again. But however
>attractively principled the case for change may appear, I think it is
>reasonable to go from the general - 'prefixes hurt the web' - to the
>specific. Some examples have been brought up; more data would be
>helpful. It's a big web. CSS is not that small either.

Well, the problem starts with the fact that nobody bothers to collect
information about this in detail in a central place. There are some
sites that try to list supported features in some limited way, but I'm
not aware of a site that would keep track of information the Working
Group would need both for regular prioritization decisions and do make
changes to the vendor prefix scheme in general. So people here rather
share their general philosophical ideas or rant about all the ills in
the world.

In order to come up with a sound policy that's not just based on gut
feelings, we would have to look at what vendor prefixed properties
there are, when they started to be supported, how they evolved, what
bugs there were, whether and how long implementations were incomplete,
like supporting only a few out of many proposed keywords, how the spe-
cification evolved alongside that, how severe the changes made to it
were, when those changes were adopted, what the policies for dropping
and modifying prefixed versions are, where modifications have been on
the timeline, how the test suites progressed, how long it took to find
major flaws, where the group decided to make a change that was then
found to break too much content, and so on and so forth.

We could then tell, for instance, how long it took implementations to
go from implementing some new selector to implementing it correctly in
cases where documents are changed dynamically. Did they all fix this
kind of bug before removing the prefix? Did it take years to fix them?
Are there drafts where most of the features are widely implemented but
some minor features that nobody currently implements are holding the
specification back? As it is we can share anecdotes about that, but it
seems the bigger picture is largely unavailable. If someone comes up
with good data, it would be far more reasonable to discuss changes.
-- 
Björn Höhrmann · mailto:bjoern@hoehrmann.de · http://bjoern.hoehrmann.de
Am Badedeich 7 · Telefon: +49(0)160/4415681 · http://www.bjoernsworld.de
25899 Dagebüll · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/ 
Received on Friday, 18 November 2011 23:57:59 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:46 GMT