> This project […] contains only the necessary parts for realistic benchmarking

> The test program is a simple […] return string

I understand how this is required to measure the effects of sandboxing in isolation. And the result is impressive.

In what ways would you expect performance to be affected when workloads are more realistic as well?

https://sciencebehindpixar.org/pipeline/rendering

I would guess there is more than one computer :)

Pixar's stuff famously takes days per frame.

iOS app to discover dog-friendly places in Amsterdam.

https://www.instagram.com/doggable.app/

We have over a 100 places now, and are getting ready for closed beta in September. You can sign up by emailing hello@doggable.app or DM-ing us on Instagram @doggable.app.

The app is developed live over at Twitch.

https://twitch.tv/elmigranto

So if most of the issues with GraphQL are the same as with the other approach, what is the point of GraphQL that justifies all the complexity?