0.64-Test crashes in background
0.64-test What seems to be at the automatic update interval, if not connected to the WiFi, there would be a "Unfortunately, F-Droid has stopped." message. It is very annoying. This did not happen to previous versions.
Sorry if I did something wrong, I am bad at bug reports.
LG Optimus Dynamic II (L39C) (Tracfone/Verizon) Android 4.1.2 no custom ROM.
Activity
Hi @husseydevin,
Sorry if I did domething wrong, I am bad at bug reports
Don't be sorry, one sure fire way to get better is by actually making bug reports, so thanks for taking the effort. Usually when there is a crash like this, we will want to know a couple of key pieces of information. You've already started alright by telling us the F-Droid version and the Android version/device/rom, which is a good start.
Other helpful things to know would be:
- Does the crash happen every time the automatic update interval should occur? or only somewhat randomly?
- Are you able to capture a log which shows the error message?
Obtaining a log of what occured during the crash is the best way for us to investigate, and how you do this depends a lot on the software/platform you are submitting a bug for. In Android, you will want to use an app like CatLog or aLogcat. If you are unable to get a log from them, then we can still investigate, but it may take a little longer to figure out.
I found out that it crashes whenever the update interval occurs, EVERY time. I found out it doesn't matter if you are on WI-Fi, 3g, 4g, no connections or even disabled. I haven't yet tried to logcat, but I could if I needed to. It is pretty annoying to catch because I am usually not available to usb debug. (I have Jelly Bean, unrooted on that. Might see my no-name Pandigital Planet tabled which is rooted CM11/AOSP 4.1.1.)
I will reinstall 0.64-test and see what I can do. (I downgraded to 0.63 because it was annoying the Vista out of me. I had to clear data when I downgraded, maybe the end result would be different...)
Possibly related to #20 (closed)
-
@mvdan: I have a sneaking suspicion that it isn't todo with issue #20 (closed), as that doesn't cause a Force Close. Rather, it may be some sort of NullPointerException that occurs only when it updates in the background. I am thinking along these lines because there is indeed a small amount of code which only runs in background-update-mode. To be honest, I haven't had time to reproduce this yet, so I will experiment at some point, maybe on the weekend. I have the advantage over @husseydevin in that I can change the code to update in the background every minute, which should make reproducing much easier :)
@husseydevin have you experienced this issue again with 0.66 or later versions?
I am pretty busy now. I will logcat (lol autocorrect changed it to log at) later. Haven't noticed it much, on 0.68-test, but I have seen it crash. I can confirm that I can't logcat on my phone, damn JB, so I set it to hourly. I could compile it with AIDE on my phone but I'd have to uninstall, which might remove the problem or not. Signature conflicts are annoying.
