This allows for easy support for HEAD requests without any code on our
side.
Differential Revision: https://phabricator.services.mozilla.com/D34631
--HG--
extra : moz-landing-system : lando
This has different behaviour from the regular update in a couple of
ways:
* It doesn't try to preserve existing conditionals that didn't match
anything in the results
* It removes subtests that are no longer present in the data set.
As such this is an appropriate mode to use when you have a try push
covering all the possible platforms, or when you are confident there
are no per-platform differences.
As part of this work, the code for updating the conditionals was
refactored to be clearer, and a bunch of tests were added to ensure
that it continues to work correctly.
Differential Revision: https://phabricator.services.mozilla.com/D34735
--HG--
extra : moz-landing-system : lando
Previously the wpt metadata update code generated rather conservative
conditionals, simply inclusing every possible condition unless all the
tests had the same outcome or similar. This isn't very nice to read
and makes the metadata rather fragile to changes in the configuration.
A better approach is to build a decision tree from the metadata,
choosing the properties to split on using a simple greedy algorithm,
as follows:
* Given a set of test results, associated run_info_properties and
properties that we want to use, partition the test results by each
property we are using in turn.
* For each partition, generate a score for how uniform the results
are in each subset after partition. The score should be good if the
partition results in small numbers of groups with mostly uniform
test results and bad if it results in a large number of groups or
groups with a mix of test results. For this we adopt a metric based
on the Shannon entropy.
* Chose the partition resulting in the best score, and use that as a
condition in the tree.
* Recursively reapply the algorithm to each group that's been
created.
Differential Revision: https://phabricator.services.mozilla.com/D34734
--HG--
extra : moz-landing-system : lando
Correctly handle clamping to 1 behavior of grayscale(),
invert(), opacity() and sepia().
Differential Revision: https://phabricator.services.mozilla.com/D35509
--HG--
extra : moz-landing-system : lando
This restores our previous and per-spec behavior. Comparing only ratios was not
correct in the case one of the dimensions was zero and thus not scaled.
Differential Revision: https://phabricator.services.mozilla.com/D35571
--HG--
extra : moz-landing-system : lando
This test is known intermittent on windows. It's already annotated to be
disabled for some categories of Windows testers, but that annotation was overly
specific.
I've already fixed the "upstream" version of this test (the version in
layout/reftests), so the intermittent failure will stop once that fix has been
synchronized around. But in the meantime, let's adjust the annotation so that
it includes Win10 64-bit testers, so we don't get intermittent failures for
those testers.
--HG--
extra : amend_source : c815ae1200e6a615f932ddf36fba6f8a7a039500
This patch produces the following serialization:
```
input | computed value
------------------------------
1. "auto" "auto"
2. "auto auto" "auto"
3. "15px auto" "15px"
4. "15px" "15px"
```
i.e. If the second value is 'auto', then it's omitted from our serialization,
because it's implied.
Besides, we update the wpt to address this spec issue:
https://github.com/w3c/csswg-drafts/issues/2574
Differential Revision: https://phabricator.services.mozilla.com/D35510
--HG--
extra : moz-landing-system : lando
This test is supposed to check CSS property support, but using
hasOwnProperty() is causing a lot of false positive because
hasOwnProperty() doesn't go up the proto chain even if the property
is actually supported.
Differential Revision: https://phabricator.services.mozilla.com/D35512
--HG--
extra : moz-landing-system : lando
