#dev 2023-01-20

2023-01-20 UTC

sknebel so for stuff where you repeatedly talk to the origin, something micro-cache-y feels appropriate (not sure to what degree that fits app engine models) - limits your rps, serve stale cache values if you got 50x on last try