Arch Linux User Repository

Run the following command in case you got errors during "Verifying source file signatures with gpg..."

gpg --recv-keys 1C61A2656FB57B7E4DE0F4C1FC918B335044912E

Alternatively, you can download Dropbox's public key from https://linux.dropbox.com/fedora/rpm-public-key.asc and import it with:

gpg --import rpm-public-key.asc

You can check whether keys are successfully imported or not using the output of gpg -k. You should find something like this:

pub   rsa2048 2010-02-11 [SC]
      1C61A2656FB57B7E4DE0F4C1FC918B335044912E
uid           [ unknown] Dropbox Automatic Signing Key <linux@dropbox.com>

If you can't run the dropbox@ service normally, try to create a read-only directory ~/.dropbox-dist and run again.

Some useful places for issues about Dropbox itself (not the package):

1. https://www.dropboxforum.com/t5/Desktop-client-builds/bd-p/101003016 Official Dropbox user feedback forum

2. Arch Linux discussion places: https://bbs.archlinux.org/, #archlinux on freenode.net, https://lists.archlinux.org/listinfo/aur-general

@Groumpf: Thanks a lot! That actually does the trick...

@gothicVI: I accidentally ran pikaur as root and got your error message. Try running pikaur as a normal user.

Running makepkg and makepkg --install on the snapshot did the trick. There seems to be something wrong with pikaur I'll need to investigate...

Hi, the update from version 92.4.382-1 -> 93.4.273-1 fails with:

:: Starte den Buildvorgang:
Job for run-u136.service failed because the control process exited with error code.
See "systemctl status run-u136.service" and "journalctl -xe" for details.

Ausführung des Befehls 'systemd-run --service-type=oneshot --pipe --wait --pty -p DynamicUser=yes -p CacheDirectory=pikaur -E HOME=/tmp -p WorkingDirectory=/var/cache/pikaur/build/dropbox makepkg --force' ist fehlgeschlagen.
:: Versuchen, wiederherzustellen?
[R] Build nochmal versuchen
[p] überspringe PGP-Signaturprüfung
[c] überspringe Checksummen-Prüfung
[i] ignoriere Architektur
[d] entferne Build-Verzeichnis und versuche erneut
[e] edit PKGBUILD
------------------------
[u] überspringe Build dieses Pakets
[a] bauen aller Pakete abbrechen

where

systemctl status run-u136.service
● run-u136.service - /usr/bin/makepkg --force
     Loaded: loaded (/run/systemd/transient/run-u136.service; transient)
  Transient: yes
     Active: failed (Result: exit-code) since Sun 2020-03-29 14:25:47 CEST; 4min 27s ago
    Process: 3231 ExecStart=/usr/bin/makepkg --force (code=exited, status=1/FAILURE)
   Main PID: 3231 (code=exited, status=1/FAILURE)

Mär 29 14:25:26 ${HOSTNAME} systemd[1]: Starting /usr/bin/makepkg --force...
Mär 29 14:25:47 ${HOSTNAME} systemd[1]: run-u136.service: Main process exited, code=exited, status=1/FAILURE
Mär 29 14:25:47 ${HOSTNAME} systemd[1]: run-u136.service: Failed with result 'exit-code'.
Mär 29 14:25:47 ${HOSTNAME} systemd[1]: Failed to start /usr/bin/makepkg --force.

and

journalctl -xe
Mär 29 14:26:51 ${HOSTNAME} kded5[1357]: Registering "org.kde.StatusNotifierItem-5501-2/StatusNotifierItem" to system tray
Mär 29 14:26:51 ${HOSTNAME} plasmashell[1444]: QQuickItem::stackAfter: Cannot stack StatusNotifierItem_QMLTYPE_391(0x55a649fdff80, parent=0x55a6493beb80, geometry=0,0 0x0) after StatusNotifi>
Mär 29 14:26:52 ${HOSTNAME} systemd[1]: run-u136.service: Main process exited, code=exited, status=1/FAILURE
-- Subject: Unit process exited
-- Defined-By: systemd
-- Support: https://forum.manjaro.org/c/technical-issues-and-assistance
-- 
-- An ExecStart= process belonging to unit run-u136.service has exited.
-- 
-- The process' exit code is 'exited' and its exit status is 1.
Mär 29 14:26:52 ${HOSTNAME} systemd[1]: run-u136.service: Failed with result 'exit-code'.
-- Subject: Unit failed
-- Defined-By: systemd
-- Support: https://forum.manjaro.org/c/technical-issues-and-assistance
-- 
-- The unit run-u136.service has entered the 'failed' state with result 'exit-code'.
Mär 29 14:26:52 ${HOSTNAME} systemd[1]: Failed to start /usr/bin/makepkg --force.
-- Subject: A start job for unit run-u136.service has failed
-- Defined-By: systemd
-- Support: https://forum.manjaro.org/c/technical-issues-and-assistance
-- 
-- A start job for unit run-u136.service has finished with a failure.
-- 
-- The job identifier is 2492 and the job result is failed.
Mär 29 14:26:52 ${HOSTNAME} audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj==unconfined msg='unit=run-u136 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? ad>
Mär 29 14:26:52 ${HOSTNAME} kernel: audit: type=1130 audit(1585484812.965:108): pid=1 uid=0 auid=4294967295 ses=4294967295 subj==unconfined msg='unit=run-u136 comm="systemd" exe="/usr/lib/sy>
Mär 29 14:26:54 ${HOSTNAME} wpa_supplicant[1037]: wlo1: WPA: Group rekeying completed with 38:10:d5:32:86:99 [GTK=CCMP]
Mär 29 14:28:25 ${HOSTNAME} kwin_x11[1436]: Could not attach Keys property to:  PlasmaQuick::Dialog(0x558b70338c80)  is not an Item
Mär 29 14:28:25 ${HOSTNAME} kwin_x11[1436]: qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 22552, resource id: 37751024, major code: 3 (GetWindowAttributes), minor code: 0
Mär 29 14:28:25 ${HOSTNAME} kwin_x11[1436]: qt.qpa.xcb: QXcbConnection: XCB error: 9 (BadDrawable), sequence: 22553, resource id: 37751024, major code: 14 (GetGeometry), minor code: 0
Mär 29 14:28:25 ${HOSTNAME} kwin_x11[1436]: qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 22556, resource id: 37751024, major code: 3 (GetWindowAttributes), minor code: 0
Mär 29 14:28:25 ${HOSTNAME} kwin_x11[1436]: qt.qpa.xcb: QXcbConnection: XCB error: 9 (BadDrawable), sequence: 22557, resource id: 37751024, major code: 14 (GetGeometry), minor code: 0
Mär 29 14:28:29 ${HOSTNAME} sudo[2066]: pam_unix(sudo:session): session closed for user root
Mär 29 14:28:29 ${HOSTNAME} kernel: audit: type=1106 audit(1585484909.176:109): pid=2066 uid=0 auid=1000 ses=2 subj==unconfined msg='op=PAM:session_close grantors=pam_limits,pam_unix,pam_per>
Mär 29 14:28:29 ${HOSTNAME} kernel: audit: type=1104 audit(1585484909.176:110): pid=2066 uid=0 auid=1000 ses=2 subj==unconfined msg='op=PAM:setcred grantors=pam_unix,pam_permit,pam_env acct=>
Mär 29 14:28:29 ${HOSTNAME} audit[2066]: USER_END pid=2066 uid=0 auid=1000 ses=2 subj==unconfined msg='op=PAM:session_close grantors=pam_limits,pam_unix,pam_permit acct="root" exe="/usr/bin/>
Mär 29 14:28:29 ${HOSTNAME} audit[2066]: CRED_DISP pid=2066 uid=0 auid=1000 ses=2 subj==unconfined msg='op=PAM:setcred grantors=pam_unix,pam_permit,pam_env acct="root" exe="/usr/bin/sudo" ho>
Mär 29 14:29:45 ${HOSTNAME} kwin_x11[1436]: qt.qpa.xcb: QXcbConnection: XCB error: 3 (BadWindow), sequence: 36756, resource id: 37751044, major code: 20 (GetProperty), minor code: 0

Do you need any additional information?

When I get the error in

Starting prepare()

which has been the case for quite a while, I do

rm src/dropbox.desktop

and since I initiated the process with makepkg -si, I issue that command again. YMMV

Hi, for the last few versions, I’m getting the following error:

==> Starting prepare()...
[dropbox]                         Generating desktop file... no
dropbox.desktop already exists. Use -f as the first argument to gendesk to overwrite.
==> ERROR: A failure occurred in prepare().
    Aborting...

@jomority No problem, it would be unreasonable to require everybody to read hundreds of comments before posting anything. Maybe we (I, at least) should add a few notes about the Dropbox startup process to the Wiki.

@fra-san Ah, sorry, I was not aware that it was already proposed.

So, it looks like currently, no easy solution for all setups is possible. It could maybe be solved by a wrapper script, but really the starting process should be standardized upstream.

The %h is on me. I didn't test it with the system unit. I used it because homes can be in other locations than /home/$USER. It seems like there is no easy solution for instantiated units for that case.

And I forgot, that I symlinked .dropbox to .local/share/dropbox.

So next time I will think longer and search harder, before I post here.

@jomority Something like that has been proposed in late 2018, see this comment by soderstrom.

Unfortunately it doesn't work in some configurations, as I pointed out back then. The dropbox process only forks if it can find a new version online. This is usually true if you haven't opted out from the "Early releases" program. If a new version cannot be found, the process does not fork, systemd assumes something went wrong (when a timeout is hit), kills it and starts a new instance. In a loop. This issue may be possibly worked out, but any solution will have to be thoroughly tested before it can be added to the package.

As an aside: %h in system units (dropbox@) resolves to /root; (at least on my system) the pid file's full path has to be /home/%I/.dropbox/dropbox.pid.

I have another solution for using the dropbox service (referencing the pinned comment of yan12125 from 2018-08-01). Override the unit with:

[Service]
# Fix successful start not detected
SuccessExitStatus=SIGKILL
# Fix stopping and detection of running status and main process
Type=forking
PIDFile=%h/.local/share/dropbox/dropbox.pid

You can override this with systemctl --user edit dropbox or sudo systemctl edit dropbox@ depending on if you are using the user service.

Maybe the package sources could also be updated.