New Topic: We need a rant about
New Topic: We need a rant about
duplicate queries vs duplicate data3 Replies
The strong suggestion is "let convex do the work" and keep everything really clean, which usually means: local reasoning, attach queries directly to components, and therefore lots of small queries
teams end up happier with this
the counterargument is function calls -- are we advocating for this b/c it results in higher bills?
in the long run, no
we'll be introducing a "scale" plan that makes cache hits significantly cheaper than non cache hits, and then this architectural design gets very very cost efficient as well
so more is coming soon. but for now, we still recommend customers build for a the future, and just create a higher number of smaller queries, rely on convex's client libraries, smart caches, built-in deduplication to take care of things
vs. doing your own sophisticated client-side "state management"
@jamwt thanks alot. well i'm glad i would not have to redo my expensive code in future because on every step there is a new query in my code and i'm basically doing dirty with queries and keeping data clean. looking forward to a bright future for convex.
also to mention, i enjoy your videos too 🙏 :convex:
thanks!!