stumpapp / stump

A free and open source comics, manga and digital book server with OPDS support (WIP)
https://stumpapp.dev
MIT License
1.01k stars 45 forks source link

[BUG] koreader sync not working #506

Open LnLcFlx opened 2 hours ago

LnLcFlx commented 2 hours ago

First of all, thanks a lot for implementing this feature!

Description of the Bug

I followed https://github.com/stumpapp/stump/blob/4055495caf7ce3fcedca14a484fdc73aa32531c0/docs/pages/guides/integrations/koreader.mdx#koreader-sync on how to set up koreader sync. I am running koreader version 2024.07-8802 on a Tolino Vision 5. When I click on "Push progress from this device now" I get "Something went wrong syncing progress, check your network connection and try again later".

I used an arbitrary username and password as it wouldn't let me use an empty one, but it says the login was successful.

Yes, I have

Build Details:

aaronleopold commented 2 hours ago

Thanks for reporting this issue!

First I just want to double check that when you say "unstable image" that you are referring to the experimental image? The commit you specified looks to be right, but there is an old unstable that should probably be removed now.

Something went wrong syncing progress, check your network connection and try again later

If you can bump the verbosity and try again, please share the logs to see if perhaps something helpful can be found from there.

KoReader-compatible hashes enabled

Can you confirm there are hashes being generated? I'm not sure how comfortable you are with SQLite browsers, but you can just query for a media record you expect to have one and check that the koreader_hash column is set. I will say that if you set the field on an existing library, it probably isn't set. The TLDR; for that is some remaining work to improve the scanner to granularly opt-in to some operations to "force" a visit to a particular book / set of books. It was already started with a small set of options, and while it is exposed to the API the UI does not yet surface it. So you can use the API directly to make those hashes get set, if needed

I'm just getting back from DC for a work trip, so won't be too responsive the rest of the night. However I'll try and dedicate some time this weekend to debug further