I’ve got a chart that displays the current user’s count of Answer things grouped by creation date. It takes over 6 seconds to load (on Personal plan) despite the current user having created only 35 Answer things.
The Bubble Manual has a section on connecting types with each other that offers three options for database design. I’m using the approach Bubble describes as Option 1+2. I could try Option 3 but that’s recommended when you “expect one thing to have a list of more than 100 values”.
With only 35 records I’d expect Option 1+2 to have much better performance. I’ve done some testing and the loading of this chart seem to scale linearly so at the 100 record threshold this chart would have a load speed of roughly 20 seconds.
Am I doing something wrong? Any advice would be appreciated.
yes, I’ve seen that thread and tried the loading thing. the issue isn’t the chart load but the data load. so my question is at its core another “why is bubble so slow?” question.
in response to these questions - there are a lot - bubble offers generic answers. they’re usually something like this: “bubble is fast enough; you just don’t know how to use bubble and you designed your database incorrectly”
it’s really frustrating and why I referenced the Bubble Manual that tells you how to structure your bubble database and described how my database is structured according to the Manual.
so why are my database queries so slow? I have no idea
my solution was to just remove this chart from my app. it was a nice to have but not critical feature. sucks but that’s bubble life
Not bad once you get the gist of it. I’d be happy to help you set one up sometime if you wanted to plan out a zoom call.
I probably won’t make documentation for them because they’re free but will do some forum posts if someone has specific questions. I’d also add features as requested if they can be feasibly implemented.
I’m using a fairly complex custom timeline chart for one of my apps that I coded. Loads very quick. Slow load times are likely due to a lot of complex data filtering.
Because calls to the dB often require the server to be woken up. The server takes a second to “come alive” then does the search and then sends it back.
Using Firebase or another real-time database or just using a backend that doesn’t need to “be woken up” with client side (not server side) api calls would make your data retrieval faster for sure. It takes more work though to configure a plugin the way you need it.
but will do some forum posts if someone has specific questions. I’d also add features as requested if they can be feasibly implemented.