Firefox v68 is the _previous_ Extended Support Release. We are
increasing this because our current minimum version doesn't support
async iterators, so some of our dependencies cause the extension to
crash.
Our metrics show that usage of Firefox versions older than this is
quite low (under 0.7% of Firefox users in the past month). These older
versions are also _very behind_ on security updates. Using the Extended
Support Release also makes it easier for us to test the minimum
version, and ensure our extension remains compatible with it.
Relates to #6805
This restores support for versions of the inpage provider prior to v8.
This is intended to support dapps and extensions that directly
instantiated their own provider rather than using the injected
provider.
* Forward traffic between old and new provider streams
* Ignore publicConfig stream for non-legacy muxes
* Transform accountsChanged notification for legacy streams
* Convert publicConfigStore to singleton
Co-authored-by: Mark Stacey <markjstacey@gmail.com>
* Migration 51: ensure chainId is set in network controller provider state for all infura/default networks
* Clean up
* Migrate incorrect as well as falsy chainIds
Co-authored-by: Erik Marks <rekmarks@protonmail.com>
Our automatic token detection was hard-coded to only work on our built-
in Infura Mainnet endpoint. It now works with custom Mainnet RPC
endpoints as well.
Relates to #6992
The `disable-console` script introduced in #10040 used an arrow-
function no-op function to replace `console.log` and `console.info`.
This replacement function was early-bound to the `this` context of the
`disable-console` script, because that's how arrow functions work.
This violates an assumption baked into Sentry, which also replaces the
`console` functions. It wraps them in a function it uses to track
console logs as breadcrumbs. This wrapper function blows up for some
reason if the "original" `console` function is early-bound to a `this`
value of `undefined`.
This resulted in various UI freezes. One example is during onboarding,
when using Firefox with Enhanced Tracking Protection set in "strict"
mode. After submitting a password in the 'Create wallet' flow, the
Sentry `console` wrapper would throw and leave the user stuck on the
loading screen.
By replacing the no-op arrow function with a no-op function
declaration, the problem has been resolved.
Relates to #10097
The `disable-console` script introduced in #10040 used an arrow-
function no-op function to replace `console.log` and `console.info`.
This replacement function was early-bound to the `this` context of the
`disable-console` script, because that's how arrow functions work.
This violates an assumption baked into Sentry, which also replaces the
`console` functions. It wraps them in a function it uses to track
console logs as breadcrumbs. This wrapper function blows up for some
reason if the "original" `console` function is early-bound to a `this`
value of `undefined`.
This resulted in various UI freezes. One example is during onboarding,
when using Firefox with Enhanced Tracking Protection set in "strict"
mode. After submitting a password in the 'Create wallet' flow, the
Sentry `console` wrapper would throw and leave the user stuck on the
loading screen.
By replacing the no-op arrow function with a no-op function
declaration, the problem has been resolved.
Relates to #10097
Failure to persist state will now only report to Sentry if the last
attempt to save state succeeded. This ensures that if anyone is stuck
in a state where state can't be saved (e.g. low disk space), we aren't
flooded with repeated errors on Sentry.
Failure to persist state will now only report to Sentry if the last
attempt to save state succeeded. This ensures that if anyone is stuck
in a state where state can't be saved (e.g. low disk space), we aren't
flooded with repeated errors on Sentry.
This update comes with a breaking change to the Approval controller. It
now requires a `defaultApprovalType` parameter.
I don't think we have any use for a default approval type, but I've
added a "NO_TYPE" one for now because it's a strict requirement. We
should consider making this parameter optional in the future, for cases
like this where it's not needed.
This update will hopefully address some caching issues we've been
seeing with our phishing configuration. See here for more details:
https://github.com/MetaMask/controllers/pull/297
`eth_getProof` is an unpermissioned, read-only RPC method for getting account-related Merkle proofs, specified here: https://eips.ethereum.org/EIPS/eip-1186
It's been supported by major Ethereum clients, and Infura, for some time. By adding it to the safe methods list, we enable this method for our users.
* Maintain console logging in dev mode
Co-authored-by: kumavis <aaron@kumavis.me>
Co-authored-by: Erik Marks <rekmarks@protonmail.com>
Co-authored-by: Mark Stacey <markjstacey@gmail.com>
On Firefox 56 and Waterfox Classic, our `runLockdown.js` script throws
an error. This is fine on the HTML pages, as the next script tags still
get run without issue (though they don't benefit from the SES lockdown
sadly). But in the `contentscript`, an exception thrown here appears to
halt the execution of subsequent scripts.
To prevent the `contentscript` from crashing completely, lockdown
errors are now caught and logged. They are also logged to Sentry on the
pages where Sentry is setup.
`eth_getProof` is an unpermissioned, read-only RPC method for getting account-related Merkle proofs, specified here: https://eips.ethereum.org/EIPS/eip-1186
It's been supported by major Ethereum clients, and Infura, for some time. By adding it to the safe methods list, we enable this method for our users.
* Maintain console logging in dev mode
Co-authored-by: kumavis <aaron@kumavis.me>
Co-authored-by: Erik Marks <rekmarks@protonmail.com>
Co-authored-by: Mark Stacey <markjstacey@gmail.com>
This PR introduces the new approval controller to the extension codebase. We use it for the permissions controller's pending approval functionality.
The approval controller sets us up for a new pattern of requesting and managing user confirmations in RPC methods. Along with the generic RPC method middleware, the approval controller will allow us to eliminate our message managers, and decouple various method handlers from our provider stack, making the implementations more portable between the extension and mobile.
On Firefox 56 and Waterfox Classic, our `runLockdown.js` script throws
an error. This is fine on the HTML pages, as the next script tags still
get run without issue (though they don't benefit from the SES lockdown
sadly). But in the `contentscript`, an exception thrown here appears to
halt the execution of subsequent scripts.
To prevent the `contentscript` from crashing completely, lockdown
errors are now caught and logged. They are also logged to Sentry on the
pages where Sentry is setup.
The Chrome minimum version has been increased from v58 to v63. We found
that we had very few users on versions below v63, and v62 is
incompatible with our SES lockdown dependency.
This also makes us compatible with Object rest/spread syntax, so we
might not have to transpile that anymore. I'll revisit that separately.
When you load an extension `.zip` file in Firefox, it fails to load
scripts with the `.cjs` file extension. However, it works if you load
the extension via the `manifest.json` file instead.
After renaming the `lockdown.cjs` file to `lockdown.js`, it works in
Firefox in all cases, regardless whether it's loaded by manifest or by
`.zip`.
The new metrics controller has a `trackEvent` function that was being
called unbound, so `this` references were undefined. It is now bound
early in both places where it is passed in as a parameter.
The SES lockdown added in #9729 had the effect of obfuscating our error
messages. Any messages printed to the console would have the error
message replaced with the string "Error #" followed by a number. The
stack was also updated to point at `lockdown.cjs`, though the original
stack was preserved beneath the top stack frame.
Marking the `console` API as untamed seems to have fixed both issues.
The original error message is now printed to the console, along with
the original stack.
David Walsh
Mark Stacey
Erik Marks
MetaMask Bot
Yahsin Huang
gasolin
Dan J Miller
Wanshan
Brad Decker
Etienne Dusseault