Call for testing: Ubuntu Touch 24.04-1.3

messayisto

gpatel-fr

After looking at the MRs, there is one (1) change since 2 years. I don't see anything else but it's not a small change, rather a collection of logically unrelated changes.

@peat_psuwit
should not a 'stable' branch (something like 24-04-1.x that IIRC exist in some other repos) be in the repo and changes be backported only after having seen some testing in the 'unstable' branch ?

Linus67

@gpatel-fr said:

journalctl -b

Finaly the lines from journalctl -b -f

jornal.txt

gpatel-fr

@Linus67

well, there is a choice of error messages about undefined properties in gallery code.

gpatel-fr

Looking at the log with my FP5 back to stable 24.04-1.2 I see no such messages.
Messages about 'binding loop' whose fixing was one of the goal of the change.
I see one message about property of nulll but it's of 'isSelected' instead of 'show' and 'path'.

gpatel-fr

This morning after having tested that the new version gallery problem was indeed 'fixed' by returning to the previous stable, I wanted to upgrade to RC and again I see the lack of upgrade channels. For screenshot see https://forums.ubports.com/post/95155 . This may be a new problem, or not. I have never wanted to switch back and forth like that previously.

Edit: I restarted the phone again and the channels were back to normal. This is a transient problem.

Edit 2: checking again, the Update parameters screen is displaying 'Channel: candidate', the system state displays '1.2' and checking updates says that everything is up-to-date. This is a bit messed up

Edit 3: after 2 hours of trying of updating, failing, restarting, switching channel, restarting, failing, I finally discovered in a flash of insight into the minds of advanced developers that the update process is actually testing the update history, the things that are displayed below 'recent updates', and this case of getting RC and stable back and forth has never been really thought to be possible (after all, regressions are something that have never been seen for more than one time or two in all the eons of software development), and so unlocking the switch to RC again needed to delete the previous update history.
So I'm finally on RC3 again. Oups, edit 4: RC2

gpatel-fr

gpatel-fr said:

I was not able to use the Recorder

cancel that, it had absolutely nothing to do with the upgrade, it was just a problem in my phone configuration. Recorder works fine.

peat_psuwit

@gpatel-fr said:

After looking at the MRs, there is one (1) change since 2 years. I don't see anything else but it's not a small change, rather a collection of logically unrelated changes.

@peat_psuwit
should not a 'stable' branch (something like 24-04-1.x that IIRC exist in some other repos) be in the repo and changes be backported only after having seen some testing in the 'unstable' branch ?

Unlike system components and some system apps, Gallery app is a Click application where updates are shipped from OpenStore. And said change has not been released to OpenStore yet, so I'm pretty sure it has not made it to 24.04-1.3.

I'll look into this problem in more detail. Since this is a regression, I'll try to fix this problem before the release of 24.04-1.3.

gpatel-fr

@peat_psuwit said:

Unlike system components and some system apps, Gallery app is a Click application where updates are shipped from OpenStore.

color me surprised.

phablet@ubuntu-phablet:~$ ls /usr/share/click/preinstalled/gallery.ubports/
3.1.1  current

gpatel-fr

gpatel-fr said:

3.1.1

BTW this does not look normal: the version is not incremented while there is about 500 LOC added in the RC version.

BTW 2: I have done already 3 back and forth between 1.2 and 1.3 RC2 and I have trouble reproducing the problem, that is, I have seen the problem in my last reinstall of 1.2 . Color me baffled

peat_psuwit

@gpatel-fr said:

BTW this does not look normal: the version is not incremented while there is about 500 LOC added in the RC version.

How did you determine that? I see the exact number of LoC [^1] on both 24.04-1.2, 24.04-1.3 RC 2 and 24.04-2.x daily images.

[^1]: determined with find /usr/share/click/preinstalled/gallery.ubports/3.1.1 -type f -print0 | wc --files0-from=- | tail -1

peat_psuwit

@gpatel-fr said:

BTW 2: I have done already 3 back and forth between 1.2 and 1.3 RC2 and I have trouble reproducing the problem, that is, I have seen the problem in my last reinstall of 1.2 . Color me baffled

I have tested on my side and I can indeed reproduce the problem on both 24.04-1.2 and 24.04-1.3 RC 2. So this is likely a problem on the Gallery app and not on the Ubuntu Touch per-se. As such, we'll fix this issue on the Gallery app and will publish the update on OpenStore in the future. A GitLab issue is opened at: https://gitlab.com/ubports/development/apps/lomiri-gallery-app/-/work_items/130

Thank you you and @linus67 for bringing this issue up.

peat_psuwit

Update 4 May 2026: We've published Ubuntu Touch 24.04-1.3 RC 3 which fixes the regression regarding sensor support. We hope that this is the final RC; Ubuntu Touch 24.04-1.3 is still scheduled to be released on 7 May 2026.

Vlad Nirky

@peat_psuwit
Hello Ratchanan.
I noticed that with RC2, the phone’s USB-C port was powering my Lilygo LoRa board.
In RC3 (and in 24.04-2.x), the USB-C port no longer powers it (OTG?).
Forget it, it seams that the board has to be connected during reboot after a power off of the phone.
It's working now.

peat_psuwit

Ubuntu Touch 24.04-1.3 is released. Thank you everyone for testing.

https://ubports.com/blog/ubports-news-1/ubuntu-touch-24-04-1-3-release-and-updates-on-ubuntu-touch-24-04-2-0-3996

kristatos

@peat_psuwit said:

I've asked one of the developers in the relevant area to look into this. Although, given that this is a cosmetic issue, it might not be fixed before 24.04-1.3 is released.

I have updated this morning from 1.2 stable to 1.3 stable and have the same issue. Is there already a bugreport for that? All other things seems to work without any issues

gpatel-fr

@domubpkm said:

clock which indicates H-2 hours (France) until the time and date icon in the system settings is pressed.

FYI an issue has been created on Gitlab.

As the hypothesis exposed in the issue description needs some confirmation, the next moment you see the problem (before going in the system settings), could you please issue this command in terminal:

sudo systemctl start systemd-timedated

if it fixes the indicator, the theory will be vindicated and it will be a step toward solving the problem.

If not, it's nonetheless interesting to know. So if you can report the result of the test in the issue, it would be a good thing.

gpatel-fr

@kristatos said:

Is there already a bugreport for that?

I browsed the recent issues a bit and did not find anything about it so I'd say not.

sixwheeledbeast

This post is deleted!

sixwheeledbeast

@kristatos Same for me on FP5.
It depends on the cellular connection selected.
With 5G selected signal shows empty, with 4G it's either low or half and with 2G or 3G its full.
All of these previously were full signal I'm pretty close to my nearest cell mast which has 2G/4G/5G