[BUG] - Crashes on one of my Stash instances, but not others

[scruffynerf]

Description

FireTV, installed fine...

Setup up with my primary Stash, the 'big' one, and it connected fine during the test, but crashed completely (dropped back to home) when opened, repeated on each try.

Then I uninstalled (since I couldn't get back to settings), and reinstalled, and tried it again. Same. Reinstalled, and set it up to use a different install of Stash (same PC, different port, newer version of Stash (v24 vs V23.1), and it worked FINE.

Tried the original, no go... (used App manager to clear 'data' to reset the settings)... Tried yet a 3rd instance of Stash (there are advantages to have lots of different setups, as both a dev and a tester)... this one is also v23.1 and it worked fine.

so the only broken one is my main Stash instance. At first i wondered if was the front page stuff (I wasn't using the stock filters), but the other instances weren't stock either, and settings the main Instance's filters to the 3 in the app didn't help.

Maybe I'm hitting a limit...

Movie instance (which works):
10 TB SCENES SIZE
8,851 SCENES
6,204 MOVIES
21,039 PERFORMERS
1,757 STUDIOS
7,800 TAGS

Other instance which works:
5.2 TB SCENES SIZE
32,336 SCENES
0 MOVIES
1,545 PERFORMERS
125 STUDIOS
142 TAGS

Main instance which crashes:
71.5 TB SCENES SIZE
53,174 SCENES
444 MOVIES
7,874 PERFORMERS
2,988 STUDIOS
21,746 TAGS

given the first 2 work, I'm gonna say it's not the Movies or Performers, so either Scenes, Studios, or Tags cross the line and cross the app? I'm gonna guess Scenes... over 32,768?

Reproduction steps

No response

Device

FireTV

Logs

not sure how to get logs for the crash...

[damontecres]

@scruffynerf Hmm, interesting. Over 32,767 would certainly exceed a short in Java/Kotlin, but it looks like all of the queries use ints not shorts.

The error handling is not great in the app yet. It more-or-less assumes everything is going to work which is obviously a terrible assumption. I will add some error handling to help catch and maybe explain this crash.

As far the logs, I think the only way to grab them for now is using ADB & logcat. It's pretty technical, but if you're up for it, it would help debug this issue for sure.

[Darklyter]

It's not going to be a size limitation, I used it against my v23.1 instance, and it showed 264k scenes with no problem. Over 20k performers, and about 1500 tags.

[Darklyter]

At least on a Shield. Also I only have 1 movie, so pretty non existent. Not sure if movies are pulled though

[damontecres]

It's probably not movies since support was added in #33 and seems to work okay.

[damontecres]

I'm going to assume this no longer a issue, but if it is @scruffynerf, please reopen!