Also remove temporary string from preferences.ftl (search-input). This was added
to avoid removing the string when running migrations against mozilla-central
MozReview-Commit-ID: IDgTNoANPyV
--HG--
extra : rebase_source : a2fe06f2214dfb6189351cd9a0eef4e41869b639
Patch originally by Ganesh, updated by Standard8
MozReview-Commit-ID: AihTLo5OyK1
--HG--
extra : rebase_source : f5a886326dd7f2fb2dba3309199ef277affe6b77
nsFrameLoader is on WebIDL bindings, so those QIs are no-ops anyway, unless the given object is no a frameloader to start with.
MozReview-Commit-ID: IPiW70H5NPc
l10n.toml files need to be valid across all channels, add back the
devtools/shim directory with a comment on when to remove it.
MozReview-Commit-ID: 8lOKRn3LwS
--HG--
extra : rebase_source : 035745f8ccf8a6542e364de3f6765ef11dfe6b25
extra : amend_source : 6c7221142591613b23cbca56cc3d3055fb2cede7
The transforms for turning an nsIScriptError into a payload that Sentry
understands were getting a bit complex for a single function, so they're
refactored into a list of transform functions that are applied in sequence to
produce the payload. This will make it easier to manage adding new transforms to
the list.
Refactoring this revaled a problem with the test code: it assumed that listeners
for console messages were notified in order of registration (since it used a
temporary listener to determine when the rest of the listeners had been notified
of a message). Changing the async evaluation of the code broke the tests, so
they had to be refactored as well.
Without a way to know when all console listeners have been notified, we can't
assert that a message will not be received by BrowserErrorReporter. We do two
things to get around this:
- Where possible, call `observe` directly on the reporter instance.
- Add constructor params for registering and unregistering listeners so we can
test that logic without relying on messages being received or not.
MozReview-Commit-ID: EEH6IROOuHD
--HG--
extra : rebase_source : a5af344c86e9756d4dbef761e4a6060515c87a61
extra : histedit_source : 5491c7359d2989a2735ec6d39de372154706c475
Now that WebExtensions run OOP in Windows by default, and we have no intention
of going back, there's no need to run an extra set of in-process tests for
them.
MozReview-Commit-ID: t8ryuuNFHe
--HG--
extra : amend_source : 90e006ca618206a943991b14cb11779baa7b8934
Calling into the Photon PageAction API to update a property is orders of
magnitude more expensive than the simple DOM-based updates we used previously.
To make matters worse, a lot of our caching was removed during the migration,
and the Photon API introduces a lot of duplicated work when selecting icons.
This patch caches the last known state for each property to avoid calling into
the Photon APIs to update each property more than necessary, and removes the
extraneous preferred icon size calculations that the Photon code already
duplicates.
MozReview-Commit-ID: LjPPxolmcd6
--HG--
extra : rebase_source : 70bce5b1460c93cc738999b1e167eb17a39259b1
The amount of computational complexity and garbage array/string/object
generation for each update to a pageAction property went up astronomically
with the migration of WebExtension page actions to the Photon API. This
resulted in non-trivial talos regression when Screenshots attempted to switch
back to the built-in pageAction API.
These changes fix most of the garbage generation, and reduce a lot of the
duplicated work for each update.
MozReview-Commit-ID: 4uPLnAesdU2
--HG--
extra : rebase_source : 3f723f3f35abf032cf12e02ce38552e21ea4827f
Additionally removes a stray debugging line that I accidentally checked into the tree previously.
MozReview-Commit-ID: 7F2S8WBgKCj
--HG--
extra : rebase_source : e25e343f3e8f57a52678c65ded973bb91d73d90a
This keeps --disable-stylo working and --enable-stylo=build with the same
semantics, but it makes also --enable-stylo / and the default to not build the
old style system at all.
This also removes the stylo-only platforms, since they're now the default.
MozReview-Commit-ID: DL2eZZn9suE