Proton refresh tokens are single-use. store_tokens replaced the tokens of
whatever blob was stored, so after `pdfs login` restarted the daemon, the
old daemon's shutdown persist wrote its spent tokens into the new
session's blob, and the new daemon failed with InvalidRefreshToken
(issue #35). `pdfs unlock` and the legacy migration had the same shape.
Each session now writes through a TokenSink that only touches a blob
with its session id that still holds the refresh token it last left
there. Shutdown only flushes a rotation the keyring refused. unlock and
migrate_legacy re-read the blob after verifying.
docs/BUGS.md B201.
Offline or paused, a drain worker claims nothing and then slept until the
earliest queued op fell due. Every queued op was already due and nothing
attempted it, so the wait was zero and all 16 workers spun on the queue
for the whole outage (issue #35: 17 min of CPU in 2 min 16 s).
A worker that claims nothing now sleeps the full idle poll; a reconnect
or resume wakes it. earliest_due_at also uses the claim's exclusions, so
an op held behind a claimed op on the same node no longer counts as due.
docs/BUGS.md B200.
Stop Sharing removes the link, invitations and members but cannot delete
the share itself, and Drive keeps listing the empty share until its
backend removes it, minutes later.
The daemon now leaves a share with no member, invitation or link out of
the listing (unless a lookup failed, so a live share is never hidden),
and the app drops stopped items from the page as soon as the requests
succeed instead of waiting for the slow reload.
An unreadable subfolder, such as a container's root-owned cache, failed
every pass of its synced folder, so nothing else in the tree reached
Drive (#33, B196). The scan now leaves such a path out, together with an
entry that vanished mid-scan and a name that is not UTF-8, and the pass
drops everything under it from classification, the wipe guard and the
push pass's deletions. It is never read as deleted and never downloaded
into. An unreadable or absent root still fails the pass. The activity
feed names unreadable paths once each time the set changes.
Two related bugs on the way to online-only:
- The push pass read deletions from the raw baseline, so a synced path
that later became ignored was trashed on Drive (B197).
- The switch evicted everything once the folder was idle, including
ignored paths such as .git/, symlinks, unreadable folders and files
changed after the pass, none of which are on Drive (B198). It now
refuses and names the path.
Closes#33
The verification window now runs in its own process, so the app survives
when WebKit cannot start its sandbox on secureblue, but the CAPTCHA still
could not be solved there (#29). WebKit binds the journal socket with
`--ro-bind-try`, bubblewrap only skips a missing path, not an unreadable
one, and WebKit has no switch to leave the socket out.
Proton Mail Bridge never uses a web view: it opens verify.proton.me in the
system browser and retries with the original challenge token, typed with
the offered methods. Do the same as a fallback.
- core: `browser_verification_url` and `browser_verified` build the page
URL and the retry credential from the challenge
- gui: when the verification window fails, a dialog opens the page in the
browser and continues the sign-in once the user is done
- cli: `pdfs login` handles a CAPTCHA the same way instead of failing
- i18n: translate the new dialog in every catalog
- docs: B195, TROUBLESHOOTING.md, ARCHITECTURE.md, changelog
Fixes#29
On secureblue, WebKit cannot start the sandbox for the verification page:
bubblewrap is kept from the journal socket, xdg-dbus-proxy fails, and
WebKit ends its process with g_error. The webview lived in pdfs-app, so
the whole app crashed after the sign-in (#29).
- gui: `pdfs-app --human-verification` shows the page in its own window,
reads the URL on stdin and prints the token on stdout; its logs go to
stderr
- gui: a child that dies or fails without a token ends the sign-in with a
message that the page could not be opened, instead of "cancelled"
- i18n: translate the new message in every catalog
- docs: B195 in the bug ledger, the cause and a one-time workaround in
TROUBLESHOOTING.md, the child process in ARCHITECTURE.md, changelog
On a fresh install only the daemon and the CLI created
~/.local/state/proton-drive-linux. The app starts the tray before the
first sign-in, so the tray failed to bind its single-instance socket and
exited, and the Takeout page could not remember staged archives (#29).
- gui: pdfs-tray and pdfs-app call AppDirs::ensure at start, which also
makes the directories owner-only
- docs: B194 in the bug ledger, changelog entry
The ledger ran B193 down to B78, then B1 up to B75, with T1 to T4 in
the middle and a stray top-level heading inside B14. Status lines used a
dozen wordings for the same few states, so finding what is still open
meant reading 6,600 lines.
Add a header: what the numbers promise (permanent, never reused; B76
and B77 were never assigned), the status vocabulary, and how to add an
entry, with a template. Add an index sorted by status, with what is left
for every open or partly fixed entry. Order all entries newest first
under one Entries section, and move the acceptance-runner findings to
their own section. Titles are in sentence case; anchors are unchanged.
Entry text is kept as written, except B89: the file it described,
proton-dev/CLAUDE.md, no longer exists, so it is closed.
ARCHITECTURE.md described the 2.x write path, where namespace changes
waited for Drive and every mount had its own inode space. Rewrite it
around the local tree (lid, inode = lid + 1, local~<lid> stand-ins), the
queue and its 16 drain workers, sync issues, and the remaining split
between the mount and the mirror engine. Drop the pins.json claim; pins
moved into the database at schema v5.
TESTING.md gains the simulation runs and loses the per-case narrative
that --list and BUGS.md already carry. ROADMAP.md takes over the
post-3.0.0 items from the milestone status file.
Remove MILESTONE-3.0.0-STATUS.md (a working log) and UI_UX_PLAN.md (the
2.x design audit). Code comments and CI that cited the untracked
docs/MILESTONE-3.0.0.md now point at TESTING.md, ARCHITECTURE.md and
ROADMAP.md. Threat-model links follow the section to §10.
The confirmation run on 812eb3d (PDFS_SIM_KNOWN=fail, twelve seeds)
passed every profile, so no open bug is tolerated any more: the call,
hang and damage matchers for B125 to B134 go, and any of them now fails
the seed. B126, B128 and B133 were fixed with B138, B162 and B137;
B134 stays open with local_first off only, which the runs do not mount.
The known-bug confirmation run failed one_client_on_a_flaky_link seeds 6
and 11 on three bugs around a create whose answer was lost:
- B182: after a restart, a write over a file whose create landed but was
never read back forked a conflict copy. A revision of a file that has
no revision of Drive's now adopts the landed node first.
- B183: the SHA-1s a create sent were kept in memory, so after a restart
its trash or retry no longer knew the node Drive made. They are now on
the op's row (pending_op.sent_sha1, schema 42).
- B184: a rename rewrote the op's target, and the trash and the retry
looked for the node under the last name only. Each attempt's parent
and name are recorded (pending_op.sent_to); the trash looks under all
of them, and a retry adopts the node an earlier attempt made and moves
it to the current name.
The daemon logs it as "<reason>; local changes uploaded as <copy>" or
"...kept at <path>", which the Activity page showed as written, in
English. It now reads both and the three reasons, one sentence each.
A file deleted after its create was sent keeps the op as a trash. The op
kept the create's attempts and backoff, and the drain recorded the failure
of a create attempt still on the wire on the row that was by then the trash.
The trash waited out the create's backoff.
The trash starts due, as a create that becomes a revision does, and the
drain records a failed attempt only if the op is still the kind it ran as
(`Db::record_attempt_failure`).
An upload that landed let go of its blob before the tree had its node. A
write open in between took the base it sampled at its start, the revision
the upload replaced, found none of its bytes cached, and queued an
incomplete revision; the next write open failed with EIO.
The drain keeps the landing node aside until the tree has it, and a write
open with no queued revision takes its base from that node, or from the
tree as it is when the handle goes in. The fake Drive can hold the drain
in that window (`hold_next_landing`).
A stop that came while a mirror pass was uploading let the pass go on to
the next level of folders until the stop gave up on it and the runtime's
end cancelled the uploads under it. Once the daemon is stopping, a batch
starts none of its waiting ops, lets the ones sending finish, and ends
the pass; the remote walk and a new pass stop too.
A delete of a file whose create was on the wire withdraws the create into
a trash. When its folder was removed too and that trash landed first,
Drive refused the create, and the drain re-homed the deleted file in the
root; only its missing bytes stopped it. The withdrawn create's trash
then read the trashed folder back into the cache before finding nothing
to trash.
A withdrawn create is no longer re-homed, and its trash is done once the
folder it was sent to is trashed or gone.
The hydrate rebuilt each listed folder's listing with a pass over every
entry. With 4,600 listed folders and 37,000 entries that took 1.7 s at
every start before the mount came up; one pass grouping the entries by
parent takes 3 ms.
The account run's 0-byte read of a new file was B166 too: Drive's
listing had the node from before its create was committed, and the
queue no longer held it when read. The fake Drive can now hold a
folder's listing on its way back with one file listed unrevised.
A listing from Drive was overlaid with the queue as it was after Drive
answered. A create that landed in between was in neither, and a wide
folder on Wi-Fi dropped files that were never unlinked. The queue is now
read before Drive is asked too; a node that left it meanwhile is listed
as the DB has it.
A stop joined every background thread for as long as each took. The sync
engine's mirror pass calls Drive with no deadline, and a stop that waited for
it was killed by systemd at TimeoutStopSec. The stop now waits up to 10 s for
the background threads and FUSE workers together, names any still running, and
the runtime shuts down within 2 s.
A listing from Drive overlays what is still queued for the folder. A create
Drive has made but whose answer is on its way, or which the drain is still
reading back, was in both: Drive's node under its new uid and the queue's
stand-in, under one name. A node queued in the folder now holds its name there.
A folder removed and made again met the old folder while its trash was on
the wire and adopted it, so the new folder's contents went into the old
one and were trashed with it. The twin search now skips hidden nodes, as
listings do.
The fake Drive can hold a trash (hold_next_trash).
adopt_real_uid asked whether a move was queued only after reading the
node back. A move that landed in between left no op, and the answer still
had the old name, so the mount showed the file under it. Ask first.
The fake Drive can hold a read's answer (hold_answer_to_next_read) so the
test can land the move while the answer is on its way.
Queueing a folder's create left the workers asleep until their 30 s idle
poll; every other enqueue woke them. A file's create still waits for the
release that attaches its bytes.
A lookup in the root publishes the "Shared with me" entry only if the root
listing did not change while its plan was read, and answered EAGAIN when it
had. Right after a restart the root is listed again as the first lookups come
in, so a write could fail with "Resource temporarily unavailable".
ensure_children now takes the listing again, up to four times.
B141 drops a rename that fails with a newer one queued behind it. The
worker records the failure and releases its claim in two steps, and a
rename queued between them waited behind the old one, which was kept
with a backoff. Released, the old one was not due and the new one was,
so the new one landed first and the old one then put the first name
back.
Releasing the claim now drops the old rename too when a newer one is
queued, and both checks match the node by local id as well as by uid.
When the read-back after an upload went unanswered, the drain let go of
the staged blob without caching it. A partial write opened next had
nothing to fill its gaps from and queued an incomplete revision, so the
write open after it was refused with EIO.
The bytes are now cached under the size and mtime the tree still has,
which is the base the next write takes. Their SHA-1 stands in for the
revision id the read-back did not bring, so that write chains onto the
upload and does not fork a conflict copy of itself.
nvrl