My attention drew back to scrobble life for a while, here's some things I did there.
Since wikipedia search is often lacking, either it won't find the movie/tv show at times or it does but won't load the image properly, I now added a manual entry for logging as well using tmdb and letterboxd. Meaning, if you can't find the movie/tv show you're looking for, you can now select "manual input" and add a link from those two sites to your log, then rate it, then add a comment/review and log it.
Becoming a bit impatient with waiting on a proper onboarding tool from someone who's too busy irl, I decided to go ahead and add gmail login to scrobble.life, for you hivers this is not something you need to worry about, but this is basically how it'll work:
you login with gmail, and the (upcoming extension update), will automatically sync your login to the extension, all your scrobbles, logs and comments are now server only and scrobble.life only, this means no immutability/censorship resistence and no hive transactions. Once you've used the site enough (I won't go into details), we will be manually offering gmail users a hive account if they wish to upgrade and link to their existing gmail login. Upon linking, their account will then receive enough RC to backfill the scrobbles they've done into hive blocks as well as movie/tv/other logs, however not comments/reviews, etc, that's only for hive accounts so we will naturally recommend them to try and get a hive account on their own if they want those features and potential rewards that may come from comment snaps/review curation.
This was quite a lot of work actually, now we wait 5-7 days for chrome to update the extension before we can start considering sharing scrobble.life with the outside world and them being able to test it even without a hive account.
Then, I finally managed to get myself a little bit of anthropic API credit so we can test our inhouse agent a little bit more, which is kind of one of the heavier points of the website. This will allow us to talk to an agent in the chat so we can ask it for recommendations of new songs based on our scrobble history.
I did some more work building out the agent a bit more, it can now web search and look for data outside if you ask it more complicated questions such as "what songs by billy talent that I haven't scrobbled yet are popular that I may wanna listen to?"
Those queries are a bit costlier however so we may have to figure out how to make that work for all our users. The simpler ones are questions like "what billy talent song has someone else scrobbled that I haven't?" as they're in-house and can be checked through our server or hive custom ops.
Anyway, quite a lot of work today, and I haven't even gotten started on what we've done on Holozing:Wilds today, think I'll save that for another day as it's getting late and I'm getting tired. :D
Thanks for reading and using scrobble.life! Hopefully soon you can start interacting with your personal scrobble agent as well!
