Re: Proposition to change the prefixing policy

On Sun, 06 May 2012 15:21:21 +0200, Tab Atkins Jr. <jackalmage@gmail.com>  
wrote:

> Florian later suggested unprefixing at some point in the WD cycle,
> when the WG agrees to it.

I haven't, you either misread me, or attribute me a mail written by someone
else.

>  Maciej proposed that we unprefix when we
> have two roughly-interoperable implementations.  Both of these are
> approximately where I want to see this land, but both suffer from some
> ambiguity that I don't think is necessary.
>
> Tantek's proposal, which he brought to the group late in the last
> Paris FtF, hits approximately the same sweet spot but with less
> ambiguity - it's nearly a mechanical process.  As a reminder, his
> proposal is that, at the moment anyone can prove two interop
> implementations of a feature with a WG-approved testsuite, we cut that
> feature into an LC->CR draft.  In effect, we have a constantly-moving
> ED, with snapshots of testably-interop features calved off as
> necessary.

I think that there are siginficant problems left under this approach.

Tying unprefixing with passing TCs provides a strong incentive to
write TCS that pass in all browsers, rather than TCs that exercise
the spec in the areas where browsers are most likely to be different,
even though the later kind is significantly more useful.

Also, just because a new browser implemented the feature and demonstrated
interoperability, it doesn't mean the first implementer will instantly drop
its prefixes. Merely that it can do so in the next version. But that next
version might not ship for quite some time, possibly a year or 2. And even
when that version is released, it could take an even longer time until
all users of the old version have upgrade. In the case of iOS safari, IE,
or the android browser, they might not even be able to upgrade without
upgrading their OS or even their hardware.

And during all that time, we will have a prefix situation fairly similar
to what we have now, with blogs and tutorials being written with a single
or a few prefixes in mind, authors writting sites that works only with some
prefixes, and a lot of content accumulating.

It would speed up things a little bit compared to what we have now, but
I don't think it would do so nearly fast enough to resolve the problem
we're dealing with.

  - Florian

Received on Monday, 7 May 2012 13:03:33 UTC