Yes, that was actually the case (some Android-App filled my storage with a lot of files called ._backup-something.tmp all of them 5GB
).
Yes, I would really like to do that ![]()
Yes, that was actually the case (some Android-App filled my storage with a lot of files called ._backup-something.tmp all of them 5GB
).
Yes, I would really like to do that ![]()
Tips on diagnosing âFailed to setup Signal clientâ error are appreciated.
There is lots of available storage on the device. Jolla C2 Sailfish OS 5.0.0.77.
I uninstalled Whisperfish
Ran
rm -rf ~/.local/share/harbour-whisperfish ~/.config/harbour-whisperfish ~/.local/share/be.rubdos/harbour-whisperfish ~/.config/be.rubdos/harbour-whisperfish
Reinstalled whisperfish via Storeman
Then try to register
Output from
harbour-whisperfish --verbose >wlog.txt 2>&1
below:
e[2m2026-04-03T14:35:36.529360Ze[0m e[34mDEBUGe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m Using Wayland-EGL
library âlibutils.soâ not found
library âlibdmabufheap.soâ not found
library âlibbase.soâ not found
library âlibz.soâ not found
library âlibcutils.soâ not found
library âandroid.hardware.graphics.mapper@4.0.soâ not found
library âlibgralloctypes.soâ not found
library âlibhidlbase.soâ not found
library âlibhardware.soâ not found
library âlibc++.soâ not found
e[2m2026-04-03T14:35:36.627787Ze[0m e[32m INFOe[0m e[2mwhisperfish::guie[0me[2m:e[0m QmlApp::application loaded - version v0.6.0-beta.39
e[2m2026-04-03T14:35:36.628350Ze[0m e[33m WARNe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m src/sailfishapp.rs:211:int __cpp_closure_14317703314545309294_impl(QmlApplicationHolder* const&)(): Failed to load translator for (âen-GBâ) Searched â/usr/share/harbour-whisperfish/translationsâ for âharbour-whisperfishâ
e[2m2026-04-03T14:35:37.104797Ze[0m e[33m WARNe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m âFailed to attach key memory region for qtcontacts-sqlite-np: QSharedMemory::attach (shmget): doesnât existâ
e[2m2026-04-03T14:35:37.104957Ze[0m e[33m WARNe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m âFailed to open contacts transient storeâ
e[2m2026-04-03T14:35:37.105262Ze[0m e[33m WARNe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m âUnable to open synchronous engine database connectionâ
e[2m2026-04-03T14:35:37.105482Ze[0m e[33m WARNe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m QSqlQuery::prepare: database not open
e[2m2026-04-03T14:35:37.105585Ze[0m e[33m WARNe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m âFailed to prepare distinct display label group selection query: \n SELECT DISTINCT DisplayLabelGroup FROM DisplayLabels ORDER BY DisplayLabelGroupSortOrder ASCâ
e[2m2026-04-03T14:35:37.127893Ze[0m e[33m WARNe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m qml/harbour-whisperfish-main.qml:443:5: QML Connections: Cannot assign to non-existent property âonRingingChangedâ
e[2m2026-04-03T14:35:37.232674Ze[0m e[32m INFOe[0m e[1mrune[0me[2m:e[0m e[2mwhisperfish::worker::setupe[0me[2m:e[0m identity_key found, assuming registered
e[2m2026-04-03T14:35:37.232815Ze[0m e[32m INFOe[0m e[1mrune[0me[2m:e[0m e[2mwhisperfish::config::settingse[0me[2m:e[0m Setting default settings.
e[2m2026-04-03T14:35:37.235872Ze[0m e[31mERRORe[0m e[1mrune[0me[2m:e[0m e[2mwhisperfish::worker::setupe[0me[2m:e[0m ACI UUID not set after registration or opening storage!
e[2m2026-04-03T14:35:37.236282Ze[0m e[34mDEBUGe[0m e[1mrune[0me[2m:e[0m e[2mwhisperfish::qtloge[0me[2m:e[0m qml/harbour-whisperfish-main.qml:431:expression for onClientFailed(): [FATAL] client failed
e[2m2026-04-03T14:35:37.293891Ze[0m e[34mDEBUGe[0m e[1mrune[0me[2m:e[0m e[2mwhisperfish::qtloge[0me[2m:e[0m qml/pages/FatalErrorPage.qml:25:onCompleted(): [FATAL] error occurred: Failed to setup Signal client
e[2m2026-04-03T14:35:37.297689Ze[0m e[34mDEBUGe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m Starting favorites query at 233 ms
e[2m2026-04-03T14:35:37.300439Ze[0m e[33m WARNe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m qml/harbour-whisperfish-main.qml:41: ReferenceError: PeopleModel is not defined
library âlibion.soâ not found
e[2m2026-04-03T14:35:37.503889Ze[0m e[34mDEBUGe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m Favorites queried in 439 ms
e[2m2026-04-03T14:35:37.593475Ze[0m e[34mDEBUGe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m Starting metadata query at 529 ms
e[2m2026-04-03T14:35:37.602686Ze[0m e[34mDEBUGe[0m e[2mwhisperfish::qtloge[0me[2m:e[0m All queried in 538 ms
Thanks.
Canât really help as you seem to have followed instructions from here: Frequently Asked Questions ¡ Wiki ¡ Whisperfish / Whisperfish - Signal on Sailfish OS ¡ GitLab ? Maybe try reboot after running the command before reinstalling?
(This is the output from after registration, ideally youâd supply logs during registration.)
But: secondary device registration/linking is current broken. Matti and I plan to work on some more things over the weekend, and weâll get back with a new version beginning next week. Meanwhile, the latest main branch version should work and is safe to install.
Just a quick side note for anyone experiencing issues recently:
A couple days ago, Signal started silently unregistering acounts that did not yet use SPQR. In my case, my primary device is signal-cli, and Whisperfish (as well as an official Signal client on a Linux desktop) connected as linked devices. Since my version of signal-cli was older than 4 weeks, apparently my account got unregistered as well (can only verify after Iâm back with my main device after the holiday, so all messages in the meantime will be lost).
While this is not at all an issue with Whisperfish, it seems I will have to re-link my device soon, too. So, looking forward to a new release, and thanks for your work and patience
itâs really a shame to see how Signal developers treat unofficial clients.
Maybe they were rushing this change in response to the security research, but itâs great to see the acknowledgement at least from the researchers here!
Heh interesting, Iâd have expected signal-cli to fix those issues before us. We had that in in 7d95d4aa / beta.39. Thanks for sharing!
I have mixed feelings about that statement. It would be really neat if Signal published some list of breaking changes for the server/API, but meanwhile they already went from âbeing hostileâ (in times of Moxie, mostly) to âignoringâ, to "being mostly friendly" but maybe not proactive.
And posted April 1st⌠not a joke?
Haha sadly not a joke ![]()
Happy Easter to those whom celebrate! I found some source code in a chocolate egg, and thought Iâd dump it here. (Thatâs a joke)
Thereâs plenty of changes in this version! The biggest changes are improvements all around messages handling logic and profile fetching âengineâ which should result in a lot more fluent operation not slowing down the startup anymore if you have a large number of recipients. This should make maintainability easier in the future as well. There are the typical library updates to keep up with upstream, both for gaining new features and simply providing a functional application.
37 files changed, 2152 insertions(+), 1327 deletions(-)
Incredible!
Quite a lot of code found from an easter egg ![]()
Super. Thank you. Runs wonderfully.
Is there a fixed timeout for attachement download retry? I used up my mobile data so loading images fails at 89% or so
It should practically immediately offer you the opportunity to retry after a download failure. Or am I interpreting your message wrongly?
There was another report of some attachment download timeouting as well a while back.
Short answer; yes. We could set it longer (not good) or see if supporting and continuing partial downloads (better) is possible.
The percentage is still increasing when it âfailsâ so I think best would be to restart the timeout timer whenever some data is received (if possible)
That mostly sounds like the attachment downloading state machine needs some rework (on top of continued downloading)
EDIT: I donât see any timeout code on the Rust side. Thereâs a Timeout error that is handled, but never constructed; stale code from before a refactor. Moreover, that timeout is at connection-time, not during the download.