@developerbayman this thread is for people located in countries where due to government restrictions or other censorship, the Ubuntu Touch servers are blocked at least some of the time.

@sixwheeledbeast said:

Overall 24.04-1.2 seemed to work perfect in my day to day usage.
Whereas 1.3 has some occasional quirks/regressions/differences? that are noticeable to me but not breaking enough to remember or make a note of each of them while out at work.
This signal strength was a noticeable one, I know the nearest mast is not far away and I know what is displayed now is very different to 1.2. Which is correct/accurate is another matter, if it's correct that's fine but ultimately people want to know at a glance if they can make a call and an empty triangle with even poor signal is not a helpful indicator if a call would be possible. Is there a different symbol for no signal as I haven't seen it? I believe you have to pull down and look at the carrier showing "Searching..." for this?
Another change I can remember is cellular data, before I could open the default browser anywhere and just browse straight away. Now intermittently with no WiFi available the browser will not have a data connection, the quick way I found to resolve this is shuffle cellular data off and on again.

I have noticed some similar things as well. Also having to reboot the phone way more often than normal because something gets funky. It's not like it's ALL the time, but it is a noticeable increase. I've also had the phone outright freeze completely when something too big in the browser was happening.

@RJDan I think a more immediate solution to this problem would be to allow apps like the activity tracker to not require the screen to be on. It should be able to track with screen off. Then you wouldn't have to go through this strange dance. Come to think of it, I am not really sure why it doesn't work with screen off? At least if suspend is prevented...

@nbdynl said:

@brenno.almeida great work!only use your app now since it is fast smooth and reliable.

only hickup, anyone else keeps missclicking the send button? its really hard to press for some reason (xiaomi redmi note 9 pro)

I have this issue on the Quintus. Maybe the mouse area is too small?

@brenno.almeida said:

@developerbayman your point just suggests that we need better confinement, not to remove it. For example, the alarm app could have apis to run confined

There is definitely an argument that current permissions are either too broad or too restrictive. Ideally, an app should not require review to use Bluetooth, for example. Because most uses of Bluetooth are not administrative. The reason review is required is because granting the permission allows full control of the BT stack.

But, like all things, someone needs to make it happen...

Question_Asker Well, I am not a lawyer. But in general the obligation of the GPL is that you must provide the source code upon request from your licensee (which is someone you have sold your GPL-containing product to). That means you must provide it, not point to some upstream source. But there are many ways to fulfill the obligations of the GPL. Technically you don't even have to put the code on an online repository. You can mail it to people on CD if you want lol.

So the simplest way to do this, is to maintain a mirror of the code on a git forge (GitHub, self-hosted Foregjo, GitLab, etc), and then when or if people ask for the code, point them at that link. Or just put the link in a manual. Which is exactly what Volla does.

Edit: and to be honest, maybe upstream source is enough, if your source is entirely upstream? I dunno. Again, not a lawyer.

@Question_Asker https://volla.online. More specifically, https://volla.online/en/operating-systems/ubuntu-touch/. Pretty sure, as far as license goes, you just need to provide the GPL sources and so forth. Trademarks of Ubuntu Touch and UBPorts foundation would be a separate matter, and I have no idea how that is handled for commercial sellers that sell UT devices. Maybe certain assets/names are trademark protected. But as far as code goes, you can take the entire UT stack, slap it on a phone, and sell it, so long as you follow the GPL or other applicable licenses.

@d4f0 said:

@sixwheeledbeast said:

Possible issue found in 1.3, maybe this could be checked it's not only myself?
Bluetooth can't be disabled or enabled from the battery menu. The toggle moves but it doesn't work and stays in the same state.
I use this regularly as a quick way to achieve some power savings from one place, part of discussions on usage here
Disabling from there own respective pull down works correctly.

It also happens on my FP4 with 1.3

Seems to be a bug, yes. Happens to me on a Volla Quintus with stable 1.3.

@PerlMax said:

@projectmoon Unfortunately this is not enough. The c library searches for config files etc. under the installation prefix (e.g. /usr/share/elementary). Perhaps it would be possible to compile efl at the UT XDG data-dir, but I didn't have luck with this so far That's why the solution would be to run in a chroot environment (where I could copy the files from the libertine container.. Best wishes, Max

I mean, if it uses an installation prefix, you should be able to change the prefix to anything you want. What happens if you try to compile it to an XDG data-dir?

@Moem ABN AMRO works fine in Waydroid + microG, with root even (although I have it hidden from Magisk).

@PerlMax said:

Hello everybody,

I am trying again to develop a simple pEFL test app. In libertine everything works great. Unfortunately the app won't start outside the libertine environment. Apparently, efl needs to be able to read files in the $HOME directory and some other directories (see https://sourceforge.net/p/enlightenment/mailman/message/58751741/).

My idea now is to run the app in a chroot environment. Would that be possible for a normal click app? Are there any examples of other apps that do this? Or is that not such a good idea?

Thanks in advance!
Max

You can make the application unconfined (not recommended), or use the read_path/write_path options in the apparmor file. Or, potentially better, change the library/application to read from the UT XDG data-dirs correctly.

@brenno.almeida said:

@projectmoon @amaurydbz found it. Spotify is not allowing using my keys to login into your account, which means that you'll need to login into Spotify Developers, generate a pair of keys and bring those to the app.

That's why it works for me, but not for you. It uses my keys.

As the app does not have a way to do that yet, I unpublished it. When I add a way to do it, I'll let you know and publish it back.

Thanks for the feedback!

Link for reference: https://developer.spotify.com/blog/2026-02-06-update-on-developer-access-and-platform-security

Hmm, so enshittifcation. Reading this, it seems one will need their pwn spotify dev account and keys?

@AmauryDBZ the redirect to 127.0.0.1 is to redirect back to the application for oauth finalization. The app runs its own server to capture the callback from Spotify. But it shouldn't crash :).

@RJDan it should be right under network, bottom right.

@brenno.almeida same result. Browser shows error empty response. And then connection refused on refresh (because app crashed).

@developerbayman dunno about a lot of that, but some things I notice that are missing: xdg handling, IPC over dbus for confined apps, auto fill API.

@brenno.almeida tried logging in. The app crashes after completing oauth flow and crashes on open every time afterwards. Oddly, there's no exception in the logs?

@Vlad-Nirky said:

@projectmoon
In this test, I have sent messages in the following order
UT message 1
Android message 2
Android message 3
UT message 4
UT message 5
Result on Indra

I have seen this before. It's not a latency problem but rather some kind of lock up with the async runtime. I haven't experienced it in a long timed. Do you have app suspension disabled in UT Tweak Tool? And how are you connected to the radio with Indra?

@pparent said:

@projectmoon

Ok well I don't know, are you testing with a specific wifi network with captive portal? Does it support RFC 8910 and RFC 8908, or is it legacy? Did you try with other wifi networks?

Ps: You can also retry with mobile data disabled, sometimes it's better.

Disabling mobile data didn't help. The app receives the captive portal URL, but I get connection refused. I don't think this has anything to do with the captive portal helper. I thought it might be uAdblock, but seems no. So, no idea. I also have no idea if the captive portal is RFC or legacy.

@Vlad-Nirky good to heaven. Can you be more specific about latency issues?