Well I'm just speculating anyways and R2 is still in active development 😄

It's less constraints around geo-replication and more about the underlying architecture of both solutions. But no benchmarks have been published yet, so we'll see. 🤷‍♂️

m

is it possible to use custom domains like r2.my.company in r2?

R2 is through a Worker so yeah.

Oh

nice

Just create a Worker like normal and then use the bindings.

Sorta like KV.

But different 😄

o

👀

Interesting

when I get acess I will try it out :)))))

there's a lil bit of docs here

Did i get a notification like a email on invite?

from what I understood you will see the tab in between pages and workers tomorow

or bellow workers

can't remember

Yeah, below I beleive.

Not in the sub-tab, just as it's own tab.

does r2 provide an s3 api?

forget it

Are you saying Streams are not count against compute time? Or just a better method with workers? I'm still new to Cloudflare. As for limiting the ranges, I did a terrible job explaining. The ideal scenerio for me is that multi-range requests are bundled into a single request (thinking cost and usage). For that to happen I assumed that you have to limit how many ranges you can add into a single multi-range request. Would this also simplify the complexity of setting up the service?

the R2 vs. KV conversation is interesting

would love to read a deep dive on the performance tradeoffs between both some day

that said, given the value size limit of KV, this comparison is only interesting if you're dealing with objects <= 25MB range requests are off the table with KV, i'm sure and you have the rich ecosystem of tooling and libraries around the S3 API standard that you don't really get with KV

What's the difference between

etag

and

httpEtag

in the

R2Object

bindings?