Edit #2: This is now a bug-report. I changed the "reply as" option back to the default user, exported the thread, and it imported correctly.
Because I'm paranoid, I export my data frequently, but when testing to see how easy it was to re-import data, I keep running into this error.
I've tried exporting the thread multiple times, but it will not import properly. The imported thread appears and is interactable, but if you click on it, it starts an infinite loading screen that locks up the tab.
Considering the amount of time I've spent on some of these stories, it makes me uneasy that the official method of saving them just...doesn't work.
Windows 11, Chrome on desktop.
Anything I can do to save the thread aside from the soulless text export? (which I'm grateful for, don't get me wrong!)
Edit I tried a couple other threads that work, so it's only the current one I'm on that's messed up somehow. The only differences I can think of between it and the other threads are that I had added some styling in the character sheet, like a background and opacity to the text boxes. I also set user responses to respond as an imported character, instead of the default "user".
Is there any chance you could share an example file that isn't working? A bunch of users have reported similar issues, but I haven't been able to replicate it. I'll fix it almost instantly if anyone can provide either 1. an example file that doesn't import properly, or 2. a sequence of steps that can reproduce the issue (i.e. allow me to create my own broken file)
I really want to fix this, but for whatever reason, I'm not able to replicate it, despite testing on every OS (Windows, Mac, Linux, Android, iOS) and every browser (Chrome, Edge, Safari, Firefox). So I just kind of desperately need someone to share an example file, or reproduction steps.
You can share with me via DM if needed. You can also open the file in a text editor and use find+replace if you need to replace any text before sharing. But you may need to right click and decompress/extract if it's a
.json.gzfile - i.e. to turn it into a.jsontext file.I wouldn't mind sharing it, it's SFW, but the problem with this one seems to have been due to me changing the "reply as" option. Resetting it to the default user before exporting seems to have fixed this. I think all my other threads work okay as of now.
I'm new to Lemmy, not sure the best way to attach or send a file.
Ah, I found the bug, thanks so much!! I've fixed it and I'll be deploying the update within the next 30 minutes or so.
Edit: Just deployed the update - let me know if you have any issues.
Dude, you are the GOAT! I hope you're able to fix the other issues people are having, as well.
Just checked the previously non-working file, and it imported fine. Awesome work, thanks!
No worries, thanks for your help!
Please let me know if you run into any other issues, since I've just deployed a fairly big update which overhauls the summary system, and fixed a few other issues. There are already some bug reports due to those new changes that people have submitted via the feedback button, but there aren't enough details yet for me to track down the issue. Might have something to do with memory handling.
Is there a changelog somewhere I can read? I'm interested in learning more about how the memory and summaries are used. Like whether pruning duplicate memories would speed up response time, or how to keep the a.i. from bringing up the same (already resolved) issue over and over, lol.