Recent updates reintroducing various bugs

[Artim96]

As it seems, the last few rapid releases did reintroduce quite a few bugs that were already closed. I can't tell at what point, but I guess something around 0.5.5d was the last issue-free version (for me). Although it's not impossible that one or the other issue may be caused by the Android 15 betas. #1115 for example was solved for me for quite a while, yet especially with Telegram there seems to be quite some interference, not only with background notifications but actually connecting to the servers when opening the app, often showing "Connecting..." for quite a while when opening the app. Also, I remember either reporting or replying to an issue (might be #1275) that connecting to the set DoH server is quite slow. It's not as bad as it was, but it does seem to be slower than it used to. But what's really unfortunate is that it currently at least once per day will just stop working, showing the typical notification of always on VPN not being connected, so I have to open the app for it to reconnect. I'll be emailing the bug report zip.

EDIT: and sometimes it may be interfering with the ability to connect to a Wi-Fi. Signal is strong, but unless you go into settings to have it connect manually, it won't.

[ignoramous]

connecting to the set DoH server is quite slow

Are you using WireGuard or any other proxy?

showing the typical notification of always on VPN not being connected, so I have to open the app for it to reconnect.

This is a crash; and there's quite a few of them that we're fixing as they get reported to us with sufficient information (now that we've implemented #1455 in v055n and continue to build upon it).

There's also issue with PR_TAGGED_ADDR_ENABLEDcrashing Rethink (on apparent memory corruption) on GrapheneOS and Samsung Androids thus far (which we suspect is emanating from the Go runtime which runs our network engine): #1539

As it seems, the last few rapid releases

There's a reason we did not graduate these to the Play Store ;) WireGuard has substantially complicated the app, and these bugs you're seeing are various efforts to close the other ones. It has been quite ... thankless.

[Artim96]

connecting to the set DoH server is quite slow

Are you using WireGuard or any other proxy?

No

showing the typical notification of always on VPN not being connected, so I have to open the app for it to reconnect.

This is a crash; and there's quite a few of them that we're fixing as they get reported to us with sufficient information (now that we've implemented #1455 in v055n and continue to build upon it).

There's also issue with PR_TAGGED_ADDR_ENABLEDcrashing Rethink (on apparent memory corruption) on GrapheneOS and Samsung Androids thus far (which we suspect is emanating from the Go runtime which runs our network engine): #1539

Now you can add Android 15 betas on Pixel to that list.

[ignoramous]

Now you can add Android 15 betas on Pixel to that list.

• Root cause PR_TAGGED_ADDR_ENABLE failures on Pixels  #962

We don't have a device with ARM MTE (address tags), which makes it hard to debug what's going on from just the submitted stacktraces (as the stacktrace from the go runtime is super unhelpful: golang/go#25035). We do plan to get one. Let's see.

Re: slow doh: I'm taking a look at the logs you sent to see what may be going on.

[ignoramous]

Update: We've identified the "slowness" bug but fixing it has introduced another bug...