> > * physically closer > > How can this be guaranteed? You have no control over the closest Edge function vs Cloud Server to the user. > Because the user does

> > * physically closer > > How can this be guaranteed? You have no control over the closest Edge function vs Cloud Server to the user. > Because the user doesn't get to talk directly to the database. The database could be on the second floor of the user's building, but it's still going to have to get routed via an API. But the API is colocated with the database in any sane setup. You're trying to argue "the extra hops can add up significantly", when using the plural is straight up illegal - it is at most 1 hop more than via Edge - and you still never satisfied how the Edge is "closer" to the database... > If you're using serverless functions for your API (presumably so, given this conversation) It's like you're not even listening to me... Cloud vendor lock-in, in this thread via serverless or Edge, is worthless. It has no benefits. You've tried for 4 comments to find 1 and you haven't. > A true Edge function has nearly 0ms latency to spin up from cold > Lambda as a pricing comparable You're right, I'm sure they just fixed the top issue with Lambda, did it with no extra costs, and are passing true pricing onto you as they try to aggressively grow. No risk here. Did you know it actually costs no extra money per request to an already running server instance? Incredible. > Trying to describe how webpage rendering works after you failed Not interested, you completely missed the point that BI applications are going to be long-running, therefore initial startup time - the literal only gain of SSR - is virtually irrelevant, and when you look at the median request issued by their app, by virtue of it being long running, your SSR only saved you time on the initial requests out of an arbitrarily large set of them. Your average request savings are epsilon defined on session length > A BI dashboard is an excellent use case for SSR, because it has multiple independent components that draw data from multiple API endpoints > This is precisely where SSR shines Again, you failed to defend how Edge / SSR are "closer to the database" (because they're not), and now you're trying to argue SSR is going to load API endpoint data faster again? Come on man I'm done with this, you were given explicit examples to work against and you just talk past my technical objection. I'm genuinely offended you are trying to say I'm the one with "cognitive dissonance" here when you sound like a cloud spokesman and can't actually defend clear technical details. Bye, enjoy giving your money to cloud companies and your time to fitting into their lock-in architectures