Whilst I agree that the amount of useful advice on this topic is woeful (including from Bubble themselves), the number of posts in this forum over the past few weeks from people panicking and speculating about what and how to optimize their apps is equally perplexing to me.
It’s really not that hard, and doesn’t take too long, to test these things for yourself - then at least you’ll know for a fact what is, and isn’t, the most efficient way to do certain things, rather than relying on other people’s advice (much of which I’ve read over the past few week, including things from Bubble themselves, is simply factually incorrect or not efficient at all).
I’ve spent the last week or so doing just that - testing various ways of storing and loading data for a number of fairly common situations, so I know exactly how I need to optimize my apps for the future and, in some cases the differences between the least, and most efficient ways, in WU costs are enormous (up to a factor of 20).
I may choose to publish my findings here in the forum at some point for the benefit of others - but honestly, the same tests can (and should) be conducted by anyone, and the results I got are specific to the tests I conducted and data I used, and may be different with different data and uses etc.
In any case, I’ll try to answer your questions, although they are very vague and the answers will, as always, be dependant on many many factors which you haven’t shared, so they can only be generalisations.
Are Satellite databases still more efficient with the new WU calculation?
It depends on exactly what you mean by ‘more efficient’, and in where and how you’re intending to load the various data. But, in most cases, I would say ‘yes’ they are ‘still’ if not ‘more’ relevant now than ever before (until that is Bubble allow the option to only return the required fields from search queries, but I don’t think that will be happening any time soon).
Is it more efficient to host the database elsewhere considering the cost of APIs calls?
I haven’t tested this yet myself (it will be one of the next things I test) - but my instinct tells me ‘no’ it won’t, just based on what I’ve seen so far in my own testing. But again, it depends on a whole range of factors, and I could be wrong about that. There is already some discussion elsewhere on the forum about this, but I haven’t seen any evidence put forward either way.
But, as I’ve said before - test it for yourself if you really want to know right now.
Are using URL parameters “good” or “bad” (again in Bubble’s own terms vs WU)?
I’m not too sure what you mean by this question? Using URL parameters as opposed to what? Why would you think using URL parameters have anything to do with WU (unless you’re using them to pull data from the database, which you probably are, but that’s not specific to URL parameters). It IS better to use unique ID’s in URL parameters (as that way you can get the database item directly, rather than searching for it, which IS less costly in WU), rather than using a search (and unlike, URL Path Parameters, which can use the unique ID or the Slug to retrieve a thing from the database, query parameters can only use the unique ID).
So my question to you all: where are those informations? Have I missed some important announcements about where I can find those critical information about Optimization best practices “officially” or through the community (ideally both…)?
Mostly, you’ll have to find it for yourself - as the specifics of your app and how it’s built will largely dictate the most efficient way to do things (and besides, why trust someone on the forum when you can test these things for yourself).
