When a suggested token was resolved in a different window, the popup
or notification UI could get stuck with an empty suggested token list,
where either action would throw an error.
This case is now handled by either redirecting or closing the window,
in the popup and notification cases respectively. This check is
performed on both component mount and update.
The "confirm suggested token" page allowed the confirm button to be
pressed even when there were no tokens to confirm. This can happen
sometimes when the page is in the process of redirecting.
The inline speedup and speedup cancellation buttons in the activity log
were broken. An exception would be thrown upon either button being
clicked, and nothing would happen from the user's perspective.
Both handlers were being passed a transaction id, which was a holdover
from before the transaction list redesign. The handlers passed for
these two actions now have the transaction id embedded, so it doesn't
need to be passed in anymore. They expect the click event to be passed
through instead.
The handlers passed also didn't handle closing the transaction details
modal when clicked. After fixing the first problem, they still didn't
work because the speedup/cancel dialog was shown behind the transaction
details modal.
Both issues are now fixed. Both buttons now close the transaction
details modal, and trigger the appropriate action.
Previously the `getMessage` function would throw if a substitution was
falsey. Now it will accept any substitution, including `undefined`.
A substitution of `null` or `undefined` will still be reported to
Sentry and printed to the console as an error, but it will not
interrupt execution. Any `null` or `undefined` substitutions will be
rendered as empty strings.
Ideally we'd never pass in `null` or `undefined` as a substitution, but
in practice this sometimes just occurs breifly between renders, which
isn't a severe enough problem to justify crashing the UI.
The detection of React component substitutions has been updated as
well, to ensure that `null` values aren't counted as React
substitutions.
The `static-server` script now exits with a code of `1` upon failure.
Previously it would print the error to the console but exit with a code
of `0`, indicating success.
This script is relied upon by our e2e tests. It was broken in #8888,
because `@metamask/test-dapp` was updated to a version with a different
directory structure. The website is now in the `dist` directory instead
of the `website` directory.
The CSS `opacity` rule accepts percentages on newer browsers, but some
older browser versions we support (e.g. Firefox v60) doesn't support
them. A number is now used instead, which is supported by all browsers
we support.
We were including the polyfill for the `Intl.RelativeTimeFormat` API,
but we weren't including any locale data. This polyfill doesn't work
without the locale data for whichever locale you're formatting.
The data for all locales we support is now included. The locale data
is loaded from disk as-needed (during app startup, and upon each change
in locale).
The `getMessage` function in `i18n-helper` was assuming that any
substitutions passed into the transaction function were used by the
corresponding localized message. However, some messages are
intentionally ignoring substitutions passed in. This was done to
simplify the UI logic, so the same substitutions could be passed in for
many different messages, even if some don't use them.
For example, `transactionCancelSuccess` is passed two substitutions but
only uses the second one. `transactionErrored` is passed in two, but
uses neither.
`getMessage` has been updated to no longer make that assumption. It
will now only throw an error if the localized message expects a
substitution that was not given. A given substitution that is unused
results in no error.
The `activeTab.id` property is relied upon in the connected sites modal
to prevent the user from manually connecting to the MetaMask extension
itself. Unfortunately the `id` property was never set.
`id` is now set on the `activeTab` state, so manually connecting to the
extension UI is now impossible.
There was a case where the `activeTab.origin` was not set, yet the user
could still navigate to the "Connected accounts" modal, which assumes
that `activeTab.origin` is set. This would happen in Firefox when the
user opened the popup on a page internal to Firefox (e.g.
`about:blank`). The connected status indicator would still be shown,
but the UI would crash when it was clicked.
The connected status indicator is now hidden whenever
`activeTab.origin` is falsy. The 'Unconnected account' alert has also
been made impossible to trigger in that circumstance.
The `activeTab` state is now set to an empty object if the `origin` of
the active tab is missing or invalid. It can be invalid if the URL
passed to the `URL` constructor is missing a scheme (e.g.
`about: blank`).
There are currently no cases where the rest of the data in `activeTab`
is useful in the absence of an `origin`. This will make upcoming UI
logic changes a bit simpler than they would be otherwise. Now we can
assume that if any property is set on `activeTab`, it must have a valid
`origin`.
The designs for the connect flow show the site `origin` below the site
icon rather than the site name. This was done for security reasons,
and because the site name is often set to an unwieldy long string.
This was accidentally undone in #8815 in the process of fixing a
separate bug. The origin has now been restored. More specific PropTypes
have been set on each use of the `domainMetadata` prop as well.
The account list on the unconnected account alert was in the wrong
order; they were in the order provided by the permissions controller
rather than by last active.
The accounts are now sorted correctly; first by last active, second by
the keyring controller order.
The `getPermittedIdentitiesForCurrentTab` selector was removed because
it is no longer used.
The mobile sync redirect was failing due to a typo in a prop. It would
fail to redirect correctly in the event of a timeout, or after pressing
'Cancel'.
Any changes that were development-focused, or that solely affected
newly-introduced features, have been omitted.
A duplicate '7.0' title has also been removed. It was accidentally
added in a merge conflict.
There were three cases where execution unintentionally continued after
an error was encountered. These cases likely are impossible to
encounter in practice due to recent validation improvements in the
`eth-json-rpc-middleware/wallet` module, but they were broken
nonetheless.
Execution inside the Promise constructor now halts immediately after
`reject` is called.
Erik Marks
Mark Stacey
MetaMask Bot
Brad Decker
Thomas Huang
Dan J Miller
Whymarrh Whitby
Esteban Miño