Defender app update (harbour-defender) - my very own ‘Sailfish Devember’

Applications
#121

I tried with one source (the main goodbye ads), first update goes very quickly (a few seconds) for > 100 000 blocked resources. It’s the update python script merging multiple sources that is so so slow.

#122

Seems like latest update broke Android-Support (X10ii, 4.3) :disappointed: Android-Support doesn’t start anymore, i tried to 1. downgrade and 2. deinstall defender and 3 reinstalled Android-Support, doesn’t help.

1 Like
Full wipe of Android support?
#123

After updating to the last version of Defender, AlienDalvik doesn’t boot anymore by being stuck in a state where I cannot stop/start it from the settings even after a reboot.
It’s probably related to the last change to mount /etc/hosts inside lxc?
I’m using an Xperia 10 ii on sailfish 4.3.

1 Like
#124

Same here! Xperia 10 ii SFOS 4.3, Defender 0.5.7.

#125

Sorry to hear this. :frowning:

And tbh cannot understand why such thing could happen!

As I do not have a 64bit device I could not test it. Will revoke that package (and better the i486 as well) for now from openrepos until clarified…

#126

Would someone please check journalctl
and also

systemctl status -l aliendalvik

systemctl status -l harbour-defender*
#127

And check

cat /var/lib/lxc/aliendalvik/extra_config

or in case (idk)

ls -la /var/lib64/lxc/
cat /var/lib64/lxc/*
#128

Best guess for a quick solution:

devel-su
mv /var/lib/lxc/aliendalvik/extra_config ~/
systemctl restart aliendalvik
#129

It is indeed extra_config that prevents Aliendalvik from starting.
The file looks as follows:

lxc.mount.entry = /system/etc/hosts system/etc/hosts none,bind,ro 0 0

After removing it, Aliendalvik starts as usual.

Here is some output from journalctl (however, it doesn’t seem to be related to this issue):

Nov 17 12:36:08 Xperia-10ii systemd[1711]: aliendalvik.service: Executable /usr/bin/start-aliendalvik-preinit.sh missing, skipping: No such file or directory
-- Subject: Process /usr/bin/start-aliendalvik-preinit.sh could not be executed
-- Defined-By: systemd
-- Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- The process /usr/bin/start-aliendalvik-preinit.sh could not be executed and failed.
-- 
-- The error number returned by this process is 2.
Nov 17 12:36:08 Xperia-10ii systemd[1712]: aliendalvik.service: Executable /usr/sbin/start-aliendalvik-preinit.sh missing, skipping: No such file or directory
-- Subject: Process /usr/sbin/start-aliendalvik-preinit.sh could not be executed
-- Defined-By: systemd
-- Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
-- 
-- The process /usr/sbin/start-aliendalvik-preinit.sh could not be executed and failed.
-- 
-- The error number returned by this process is 2.
2 Likes
#130

Looking at other entries in bsp_config, shouldn’t the extra_config file look as follows:

lxc.mount.entry = /system/etc/hosts system/etc/hosts none bind,ro 0 0

instead of

lxc.mount.entry = /system/etc/hosts system/etc/hosts none,bind,ro 0 0
3 Likes
#131

Oh no! :hide: :duck: :cover:

Thanks for investigating. Of course that cannot work!

Would you please edit this file “/var/lib/lxc/aliendalvik/extra_config” to

lxc.mount.entry = /system/etc/hosts system/etc/hosts none bind,ro 0 0

(mind the space between none and bind)
and put it back in place?
And restart Android Support once more.

This should be the cause.

P.S.: there is no /var/lib64 folder or more specific /var/lib64/lxc/aliendalvik?

1 Like
Android support does not start in 4.3.0 on XA2
#132

Yes, was writing answer above…

#133

Already did that, and can confirm it works. And there is no /var/lib64 folder on my device (Xperia 10 ii).

1 Like
#134

Thank you!

New packages to come …
(but not right now :frowning: but soon™ :slight_smile: )

–edit
And new RPMs are out (by courtesy of @orangecat, big thank you)

3 Likes
#135

@peterleinchen I have several suggestions for the new release, would you like to hear them?

#136

It seems I’ve messed something up and now I can’t block ads anymore (even though the list loaded successfully). Could someone help me out, please?

Edit: I fixed it in the end

#137

I’d like to hear them :slight_smile:

#138

Plus a counter instead of the indefinite progress. How many passed and how many remaining. Just the other day I had to once again quit updating because it had been going on for days and I had no idea where it was.

#139

One could obtain a Content-Length header and count?

#140

I don’t know. I surely hope so. It might have nothing to do with HTTP headers but rather the volume of sources if kept as objects.