Lately I have been observing the constant picking of Neon, mainly I presume for its performance from Planetscale’s CEO, and the X/Twitter community.
Here are some of the many tweets I have encountered, but I didn’t save them, so a simple search resulted on the following list, just to get the gist of it.
This post is going to be extremely short, and it was just something that struck me when I read this phrase from Project Hail Mary, when Stratt, Grace and Redell had a discussion around growing Astrophage at scale:
“As a breeder database system it’s horrible,” I said. “Way less efficient
and far lower yield QPS than my system … . But
he didn’t design it for efficiency. He designed it for scalability.”
There is profound wisdom in this imaginary discussion, and it is a reminder, there are no free lunches in distributed designs. Both are winners, Planetscale and Neon, in their own niche and for their own use cases.
I know negativity and drama in X/Twitter brings the eyeballs, but its a friendly reminder to my fellow engineers that there is no universal optimal solution.