Re: Proposition to change the prefixing policy

On Sun, May 6, 2012 at 1:02 AM, Maciej Stachowiak <mjs@apple.com> wrote:
> Here's another slightly more conservative version.
>
> Properties can be shipped in unprefixed form once both of the following are true:
> (A) The appropriate standards group (most likely the CSS WG for CSS properties) has agreed to take up the relevant specification as a work item; AND
> (B) At least two independent roughly interoperable (though not necessarily identical in all edge cases) implementations are publicly available.
>
> This would leave room for truly experimental work and proprietary extensions, and would avoid locking in syntactic quirks immediately, but would phase out prefixes much more quickly than the current approach of waiting for CR.

I'd like to chip in that I emphatically endorse this proposal.
Writing good tests takes a lot of work -- I know that very well from
personal experience.  It's also very easy to write extremely
incomplete tests and declare them adequate because no one objected, so
a test suite doesn't prove anything per se.

Good tests help a lot to gauge real-world interoperability, but
they're no guarantee.  Real pages do all sorts of things that tests
would never do, because they're written with totally different goals.
Tests necessarily will test features mostly in isolation, while real
pages will use them almost exclusively in combination with a host of
other features.

The way to gauge rough real-world interop is: do authors who use the
features extensively in practice have to work around lots of
incompatibilities, or not?  This is easy enough to tell by reading
through forums and tutorials -- or just asking some authors.  If
authors are *in practice* just copying the same markup for every
browser, it's time to unprefix regardless of whether we have any
formal tests at all.

Received on Monday, 7 May 2012 10:18:14 UTC