Did you test this on X11 too and find that it works there? If not, then there's no need for the "wayland-only" tag.

The "qt6" tag is likewise no longer needed.

I just tested this in KDE Neon Testing, and can reproduce. 

1. Opened Discover, confirmed there are updates. Closed it.
2. Clicking Start Update *seems* to trigger the update but doesn't
3. Opening Discover again showed the same packages as before
4. The tray icon came back indicating there are updates available

Raising priority as this was highlighted in the Plasma 6.4 release announcement.

Anyone able to reproduce it in 6.4.0?

I tried on KDE Linux and it worked for me.

Please provide logs. E.g. output in `journalctl --user`

🐛🧹 ⚠️ This bug has been in NEEDSINFO status with no change for at least 15 days. Please provide the requested information, then set the bug status to REPORTED. If there is no change for at least 30 days, it will be automatically closed as RESOLVED WORKSFORME.

For more information about our bug triaging procedures, please read https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging.

Thank you for helping us make KDE software even better for everyone!

Today notifier icon appeared as a cloud. I clicked "RMB -> Start Update" and appeared Discover's window, whereas I expected to see that updates starts. Please notice that after I was login to Plasma session notifier icon about updates was put in Notification section, so wasn't visible in systray. And after I closed Discover window it disappeared at all. I reported issue related this issue.

Below please find logs for command: 'journalctl --user':
<cut>
jul 20 01:01:30 Dell DiscoverNotifier[1909]: started unattended update QDateTime(2025-07-19 23:01:30.218 UTC Qt::UTC)
jul 20 01:01:30 Dell plasma-discover[8203]: qrc:/qt/qml/org/kde/discover/qml/UpdatesPage.qml:40:5: QML OverlaySheet: Binding loop detected for property "implicitHeight":
                                            file:///usr/lib/x86_64-linux-gnu/qt6/qml/org/kde/kirigami/templates/OverlaySheet.qml:139:5
jul 20 01:01:30 Dell DiscoverNotifier[1909]: Finished running plasma-discover 0 QProcess::NormalExit
jul 20 01:01:30 Dell DiscoverNotifier[1909]: skipping update, already updated on "Sat Jul 19 23:01:30 2025 GMT"
jul 20 01:01:30 Dell kded6[1699]: Service  ":1.104" unregistered
jul 20 01:01:44 Dell akonadi_imap_resource[2320]: org.kde.pim.imapresource: "XArchives/Randki/eDarling" : detected inconsistency in local cache, we're missing some messa>
jul 20 01:01:44 Dell akonadi_imap_resource[2320]: org.kde.pim.imapresource: "XArchives/Randki/eDarling" : refetching complete mailbox
jul 20 01:02:42 Dell plasmashell[1722]: kf.plasma.quick: location should be set before showing popup window
jul 20 01:02:42 Dell plasmashell[1722]: qt.qml.signalhandler: Property 'onPressed' of object CompactRepresentation_QMLTYPE_422(0x5fdcf1456e40) is a signal handler. You s>
jul 20 01:02:46 Dell plasmashell[1722]: file:///usr/share/plasma/plasmoids/org.kde.plasma.notifications/contents/ui/FullRepresentation.qml:59:13: QML PlasmoidHeading: Bi>
                                        qrc:/qt/qml/org/kde/plasma/extras/PlasmoidHeading.qml:34:5
</cut> 

When I get "normal" notifier icon I also will report what printed command 'journalctl --user'

I just tried this out on Fedora KDE and it worked as expected for me:

1. Right-click on notifier system tray icon
2. Click "Start update"
3. "Applying unattended updates..." notification appears, makes progress, and then eventually goes away

You say that different things happen when you do this? Sometimes Discover crashes, sometimes Discover opens its main window.

Since both you and Tracey can reproduce this on KDE Neon, and Harald and I cannot reproduce it elsewhere, this seems like it may be a Neon-specific issue. Moving to Neon.

Today appeared cloud as notifier, I clicked "Start update" and in result I saw Discover window.
Below please find logs for command: 'journalctl --user':
<cut>
jul 21 21:21:57 Dell DiscoverNotifier[1912]: started unattended update QDateTime(2025-07-21 19:21:57.716 UTC Qt::UTC)
jul 21 21:21:57 Dell plasma-discover[3568]: qrc:/qt/qml/org/kde/discover/qml/UpdatesPage.qml:40:5: QML OverlaySheet: Binding loop detected for property "implicitHeight":
                                            file:///usr/lib/x86_64-linux-gnu/qt6/qml/org/kde/kirigami/templates/OverlaySheet.qml:139:5
jul 21 21:21:57 Dell DiscoverNotifier[1912]: Finished running plasma-discover 0 QProcess::NormalExit
jul 21 21:21:58 Dell DiscoverNotifier[1912]: skipping update, already updated on "Mon Jul 21 19:21:57 2025 GMT"
jul 21 21:21:58 Dell kded6[1703]: Service  ":1.70" unregistered
</cut>
Here (in Discover) I clicked "Update All", and when progress reached 99% Discover crashed. Described situation happens for me very often.

(In reply to Piotr Mierzwinski from comment #8)
> Here (in Discover) I clicked "Update All", and when progress reached 99%
> Discover crashed. Described situation happens for me very often.

Thanks for the journalctl logs. Can you please open a new bug report for the crash? It looks like a different but related problem. In the new bug report, please include a backtrace. The template has instructions on how to get this. Thanks.

Created attachment 183430 [details]
System logs before triggering update manually

update_logs.txt: before triggering updates manually

The updates icon says "applying unattended updates"
Right click -> open Discover
System upgrade - 560 packages will be upgraded
Close Discover
System tray icon changes to "Security updates available"
 - At this point, the only thing of note in the log is
 
> Jul 22 11:07:34 tracey-neon-dev-vm DiscoverNotifier[1725]: skipping update, already updated on "Tue Jul 22 16:07:34 2025 GMT"

Created attachment 183431 [details]
System logs after triggering update

> Jul 22 11:07:34 tracey-neon-dev-vm DiscoverNotifier[1725]: skipping update, already updated on "Tue Jul 22 16:07:34 2025 GMT"

manual_update_logs.txt: after triggering updates

Right click tray icon & click Start Update
Tray icon again  says "applying unattended updates" 
The log shows a failure for each package, ex:

> Jul 22 11:15:12 tracey-neon-dev-vm packagekitd[3914]: pk_package_set_update_severity: assertion 'update_severity == PK_INFO_ENUM_UNKNOWN || update_severity == PK_INFO_ENUM_LOW || update_severity == PK_INFO_ENUM_NORMAL || update_severity == PK_INFO_ENUM_IMPORTANT || update_severity == PK_INFO_ENUM_CRITICAL' failed

Near the end, the log has a success message for `get-updates transaction`, bu plasma-discover seems alarmed that PackageKit stopped running.

> Jul 22 11:15:20 tracey-neon-dev-vm packagekitd[3914]: pk_package_set_update_severity: assertion 'update_severity == PK_INFO_ENUM_UNKNOWN || update_severity == PK_INFO_ENUM_LOW || update_severity == PK_INFO_ENUM_NORMAL || update_severity == PK_INFO_ENUM_IMPORTANT || update_severity == PK_INFO_ENUM_CRITICAL' failed
Jul 22 11:15:20 tracey-neon-dev-vm PackageKit[3914]: get-updates transaction /1647_ceeeebac from uid 1000 finished with success after 2535ms
Jul 22 11:15:41 tracey-neon-dev-vm systemd[1]: systemd-localed.service: Deactivated successfully.
Jul 22 11:17:01 tracey-neon-dev-vm CRON[4024]: pam_unix(cron:session): session opened for user root(uid=0) by root(uid=0)
Jul 22 11:17:01 tracey-neon-dev-vm CRON[4025]: (root) CMD (cd / && run-parts --report /etc/cron.hourly)
Jul 22 11:17:01 tracey-neon-dev-vm CRON[4024]: pam_unix(cron:session): session closed for user root
Jul 22 11:20:14 tracey-neon-dev-vm systemd[1]: Starting sysstat-collect.service - system activity accounting tool...
Jul 22 11:20:14 tracey-neon-dev-vm systemd[1]: sysstat-collect.service: Deactivated successfully.
Jul 22 11:20:14 tracey-neon-dev-vm systemd[1]: Finished sysstat-collect.service - system activity accounting tool.
Jul 22 11:20:24 tracey-neon-dev-vm PackageKit[3914]: daemon quit
Jul 22 11:20:24 tracey-neon-dev-vm plasma-discover[3898]: PackageKit stopped running!
Jul 22 11:20:25 tracey-neon-dev-vm systemd[1]: packagekit.service: Deactivated successfully.
Jul 22 11:20:25 tracey-neon-dev-vm systemd[1]: packagekit.service: Consumed 9.244s CPU time.

After all that, if I go into Discover and run updates there, I also get the crash.
The VM didn't have enough RAM to get a good backtrace. I gave it more RAM and ran updates from command line
That failed with this error (there is no indication which of the hundreds of packages failed)

> Fatal error: Error while installing package: trying to overwrite '/usr/lib/x86_64-linux-gnu/libweather_ion.so.7.0.0', which is also in package plasma-workspace

I currently can't get a new backtrace, because after the boot splash screen, the screen is black with a mouse cursor. If I get a backtrace, I'll attach it to the new report, after that's created.

From time to time I sent automatically crash report coming from Discover, but I'm not sure where it is placed, because here (link below) I can't see it.
https://bugs.kde.org/buglist.cgi?component=discover&list_id=3206123&product=Discover&resolution=---
Anyway yesterday's crash generated 17MB compressed (zst format) file.
I'm not sure if is possible put here such big file. Let someone confirm, please.

Today I get "normal" notifier about updates. I clicked RMB and "Start Updates". In result only icon changed and in logs I found:

jul 22 22:29:31 Dell plasmashell[1720]: kf.imageformats.plugins.iff: IFFChunk::innerFromDevice: unknown chunk "\x89PNG"
jul 22 22:29:35 Dell DiscoverNotifier[1908]: started unattended update QDateTime(2025-07-22 20:29:35.192 UTC Qt::UTC)
jul 22 22:29:35 Dell plasma-discover[5981]: libs QList("/usr/lib/x86_64-linux-gnu/qt6/plugins", "/usr/bin")
jul 22 22:29:35 Dell plasma-discover[5981]: org.kde.plasma.libdiscover: OdrsReviewsBackend: Fetch ratings: true
jul 22 22:29:36 Dell plasma-discover[5981]: adding empty sources model QStandardItemModel(0x5df524545140)
jul 22 22:29:40 Dell plasma-discover[5981]: error: QSnapdRequest::ConnectionFailed "Unable to connect snapd socket: No such file or directory"
jul 22 22:29:40 Dell plasma-discover[5981]: error: QSnapdRequest::ConnectionFailed "Unable to connect snapd socket: No such file or directory"
jul 22 22:29:41 Dell plasma-discover[5981]: qml: Waiting for updates

Updates didn't start.

I started snapd
$ sudo systemctl start snapd.service

I started also "plasma-discover"
libs QList("/usr/lib/x86_64-linux-gnu/qt6/plugins", "/usr/bin")
org.kde.plasma.libdiscover: OdrsReviewsBackend: Fetch ratings: false
adding empty sources model QStandardItemModel(0x616e0bfedb40)
qt.core.qobject.connect: QObject::connect: signal not found in QSnapdRefreshRequest

Notifier icon didn't appear.

(In reply to Piotr Mierzwinski from comment #13)
> From time to time I sent automatically crash report coming from Discover,
> but I'm not sure where it is placed, because here (link below) I can't see
> it.

If you chose an automatic report in Dr Konqui, it would have been sent to Sentry (https://crash-reports.kde.org) instead of bugs.kde.org. It will still get to us. Thanks for confirming.

Today I retested and seems started work, but in a bit different way. I mean after I clicked RMB into an icon in systray, appeared main Discover window, where I could click "Update All", so now I don't know what is difference between:  I double click on icon and selection an option from context menu: "Start Update". I'm not sure if this is intended action or not?

Another feature is that before I click I made update by double click onto discover icon and in main Discover window,  I clicked "Update All". All passed well, appeared message that installation will happen after restart, but icon in systray was present and I could click RMB on it.

Just confirming that the original reported bug ""Start Update" invoked from systray icon (click RMB) seems doesn't work" is still reproducible with Neon Testing and Plasma built from git-master on Solus.

Confirm. Today I repeated activity  and unfortunately happened the same what I described in original post, so icon changes, tooltip changed and nothing more - upgrade didn't start.