Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Intent to unship: Legacy -moz-transform syntax.

73 views
Skip to first unread message

Emilio Cobos Álvarez

unread,
Feb 14, 2018, 3:38:35 PM2/14/18
to dev-pl...@lists.mozilla.org
Hi,

In bug 1438297 I intend to remove the legacy syntax for -moz-transform,
which allows lengths or percentages apart from numbers in some transform
matrix components.

That will make -moz-transform a normal alias for the "transform"
property, just like -webkit-transform, following the usual parsing rules.

The reasoning for this is that this particular property is the only
alias that requires different parsing rules from the property they're
aliasing to. That means it needs to be implemented as a shorthand
property, which means in turn a bunch of special-cases to serialize it, etc.

I think the risk for this should be minimal given these properties are
unprefixed from a long time ago and we keep the property working with
the syntax that works on all browsers (plus we also have -webkit-
prefixes). Just as a data point, we didn't have a single reftest that
failed without it, only parsing tests.

Let me know if you think otherwise though.

Note also that there's bug 770560 for unprefixing them completely, I'm
more interested on removing the complexity from the code for now so I
haven't investigated that yet, so I think this change is less risky.

-- Emilio

Mike Taylor

unread,
Feb 15, 2018, 4:50:59 PM2/15/18
to Emilio Cobos Álvarez, dev-pl...@lists.mozilla.org
> On Feb 14, 2018, at 2:37 PM, Emilio Cobos Álvarez <emi...@crisal.io> wrote:
> The reasoning for this is that this particular property is the only alias that requires different parsing rules from the property they're aliasing to. That means it needs to be implemented as a shorthand property, which means in turn a bunch of special-cases to serialize it, etc.
>
> I think the risk for this should be minimal given these properties are unprefixed from a long time ago and we keep the property working with the syntax that works on all browsers (plus we also have -webkit- prefixes). Just as a data point, we didn't have a single reftest that failed without it, only parsing tests.
>
> Let me know if you think otherwise though.

LGTM.

--
Mike Taylor
Web Compat, Mozilla


0 new messages