Hoarding is good
It turns out that hoarding is good sometimes. Due to my inbox having more than 500 articles, I've discovered a bug in the archiving functionality. Unfortunately, there's no fix due to the limitations of Instapaper API, but I now have a workaround.
Testing finished
I've tested the on-device version for two weeks without any major issues, and the filesystem corruption issue seems to be fixed. Currently working on the final touches before merging the Go rewrite.
Go version released
Part of the Go rewrite of Instakobo, I've ported the Quarto library to Go and released it as quarto-go.
Minimal version working
I got a minimal version running locally that synchronizes with my Elipsa 2E. Currently it has:
- User authentication.
- Initial Kobo proxy implementation.
- Kepub conversion.
- Minimal UI for book listing and details.
Port to Go and running on Kobo devices
Finished the initial port from Deno to Go, and spent some time debugging on the Elipsa 2E. Hooking into the sync event provided some pretty janky and unstable UI, so I decided that the app will need at minimum NickelMenu and, for better experience, NickelDBus.
Spent some time working on the collections code, which had the unfortunate habit of corrupting the device filesystem due to how I was accessing the Kobo sqlite database. I think it's fixed now, but before releasing I will need to test this for a while to make sure the fix works.
Research
I looked into the possibility of extracting highlights from articles sent to a reMarkable device. There's not a lot of example code out there, but I did get a proof of concept working on some test articles.