> Please make sure that use of the old syntax triggers some kind of
developer-facing warning, so that it won't be too hard to catch the cause
for the behavior change in the few sites that will be affected.
Sounds good, I will keep the existing console message in there, at least
for a while.
On
LGTM3, and +1 to Yoav's suggestion.
On 5/29/24 1:00 AM, Yoav Weiss (@Shopify) wrote:
LGTM2
Please make sure that use of the old syntax triggers some kind of
developer-facing warning, so that it won't be too hard to catch the
cause for the behavior change in the few sites that will be
LGTM2
Please make sure that use of the old syntax triggers some kind of
developer-facing warning, so that it won't be too hard to catch the cause
for the behavior change in the few sites that will be affected.
On Tue, May 28, 2024 at 10:33 PM Joey Arhar wrote:
> > How should I interpret the
> How should I interpret the results from your investigation? That none of
the 8 investigated sites would be negatively affected?
There is only one site which is actually affected: https://resolv.com.br/
The buttons in the carousel there won't light up anymore, but they will
still work. I made
To be particular about the usage number, it's about a magnitude more
than our informal limit. That doesn't mean it can't be shipped, but it
means that we want to be fairly certain that >90% of the users are
unaffected by the change.
How should I interpret the results from your investigation?
On Tue, May 21, 2024 at 2:20 PM Joey Arhar wrote:
> > Do you know what the breakage looks like
>
> I pushed to make sure that using the old syntax with CustomStateSet.add()
> wouldn't throw exceptions when in the new mode in order to reduce breakage.
> When websites use the old syntax after it's
> Do you know what the breakage looks like
I pushed to make sure that using the old syntax with CustomStateSet.add()
wouldn't throw exceptions when in the new mode in order to reduce breakage.
When websites use the old syntax after it's removed, they will just have
styling differences because
Hey,
0.04% seems like a fairly sizable number. Do you know what the breakage
looks like or whether this usage is limited to a library/a small set of
large websites or something else?
Ideally, this is feature detected with some fallback syntax
On Tue, May 21, 2024 at 1:39 PM Joey Arhar wrote:
Contact emailsjar...@chromium.org
ExplainerNone
Specificationhttps://github.com/whatwg/html/pull/8467
Summary
The CSS custom state pseudo-class is being renamed from :--foo to
:state(foo). The new syntax, :state(foo), has been enabled by default, and
now we have to deprecate and remove the
