lucifer: Let me know when you have time to talk about the slow loading dashboard issue, I'm getting confused about what we're improving and where 😅
2025-05-29 14904, 2025
monkey[m]
I'm testing on test.LB, but still seeing up to 40 seconds *initial* page load in some cases. Happy to help implement multi-stage loading (initial load returns whatever it can in minimal passes, then front-end does another API call to get the rest)
2025-05-29 14957, 2025
mthax joined the channel
2025-05-29 14933, 2025
monkey[m]
I'm somewhat caught up after a quick chat with mayhem, will put down what I understood in a Gdoc
2025-05-29 14936, 2025
mthax has quit
2025-05-29 14907, 2025
lucifer[m]
[@monkey:chatbrainz.org](https://matrix.to/#/@monkey:chatbrainz.org) can you share where you see the 40s loading time?
2025-05-29 14923, 2025
mthax joined the channel
2025-05-29 14940, 2025
lucifer[m]
[@monkey:chatbrainz.org](https://matrix.to/#/@monkey:chatbrainz.org), [@mayhem:chatbrainz.org](https://matrix.to/#/@mayhem:chatbrainz.org) another option to speed up loading time, cache the latest 25 listens for every user in redis.
2025-05-29 14958, 2025
monkey[m]
Looking at the test.LB logs for user Kanau (fetch listens Kanau 39.95s (5 passes))
2025-05-29 14925, 2025
monkey[m]
This is for their second or third pagte of listens (i.e. API calls, not initial load)
2025-05-29 14927, 2025
lucifer[m]
I see. So the fix shouldn't do multiple passes at all for initial load.
2025-05-29 14938, 2025
lucifer[m]
So it's working as expected.
2025-05-29 14944, 2025
monkey[m]
Let me check but I think the initial load was also slow
2025-05-29 14949, 2025
lucifer[m]
Subsequent loads will take longer as nothing has changed.
2025-05-29 14941, 2025
lucifer[m]
It shouldn't take more than 10-12s, which is not great but at least the page loads sort of.
2025-05-29 14914, 2025
lucifer[m]
We can shorted the one year search window to a few months and it would be faster.
<fettuccinae[m]> "Yes please" <- He agreed, made the change and pushed it to production. 😎
2025-05-29 14934, 2025
BobSwift[m]
<mayhem[m]> "https://playvinyltap.com..." <- Interesting that (at least today's puzzle for classic rock) seems to be based on the digital release which is the only one with that artist.
2025-05-29 14906, 2025
mayhem[m]
can of worms, cans of worms as far as they eye can see!
2025-05-29 14932, 2025
BobSwift[m]
Still pretty neat.
2025-05-29 14915, 2025
mayhem[m]
yeah. the goal is to do "wordle for music"
2025-05-29 14957, 2025
BobSwift[m]
I'm just bummed out because I would have gotten it on the first guess if they used the more common album artist. 😜
2025-05-29 14942, 2025
mayhem[m]
I think the author would appreciate hearing from you. want me to dm the email?
yvanzo: yes, it's a dependency of `recording_first_release_date`, so the script tries to build it in that case (and it's expected to be skipped if the table was previously built)
