Message ids repeat across accounts, so a hook payload naming only an
emailId could not tell a plugin which mailbox the message lies in when
it is a shared or group account's.
onBeforeEmailStarToggle and onEmailStarToggle now carry accountId when
the starred message is not in the user's own account, as the
message-row-actions click context does. Every star (the row's button,
the reader's star, the key, the menu) goes through toggleStar, one
message at a time, so one accountId names it; the account is resolved
before the plugins are asked. An own message's payload is unchanged.
onAfterEmailSend now carries accountId, the JMAP account the sent
message lies in (a mail sent from a shared or group identity lies in
that account), so api.email.get finds it. sendEmail returns that account
as emailAccountId, the S/MIME send too. The hook also carries
composerId, the id the sending composer's slots and actions gave
plugins, so a plugin can act on what the user chose in that composer.
A plugin opens a message as a click on a list row or on a search result
entry naming one does, also a message of a shared or group account,
reached through the login that reads it. On the mail page it opens in
place, in the Pro shell in an email tab as a message link does there;
elsewhere the tab loads the message's link, at most five times a minute.
A message no login reads, or one its account does not have, is refused
with an error the plugin sees, as is one the mail page could not open.
Needs email:read.
A shared account the calendar fan-out probes on suspicion may list the
calendars capability and still refuse CalendarEvent/query (Stalwart 1.0
answers accountNotSupportedByMethod for a share that grants no calendar
access). queryCalendarEvents remembered such an account and skipped it
from then on, but getCalendarTasks failed as a whole on it, so the task
list and the plugin task API stayed empty after a reload until the
calendar page had run once. Reading an account's tasks now takes that
refusal as no tasks and remembers the account the same way, through one
helper both share; any other failure still fails the read.
Input inside an iframe never reaches the host window, so the idle timeout
signed users out while they worked in a plugin page, dialog or panel. The
sandbox runtime now reports trusted input in its frame (the guard's own
events, focus left out) through a MessagePort it hands the host with
sandbox-ready, before any plugin code runs, at most once per 15 s. The
host bridge passes it to a small notifier the session timeout guard
counts like its own window's input, so it reaches the other tabs too.
Linus Rath
VasilisNtovolos
Jacob Zeitz