Search Rocks — If the Database Can Too
My colleague Jason and I were talking about search — which spilled into a 3-day exchange of emails between us — with me trying to work from the metaphor of dance and synchronization and he with engines and cars. (Talk about rigid data structures).
Then like a bluebird, I received this gem from him.
Plan on seeing bits of his prose creep into our marketing.
Google shows us that search rocks
Wouldn’t it be great if everything worked like that
Why doesn’t it?
Because most tech uses relational databases
So are search engines the answer?
No because all they do is search
You need infrastructure
A place to store data too — safely, surely, securely
A way to do rich query not just text search
An understanding of structure, but forgiveness of messy structure
Immediate indexing not the delays you usually see with search engines
With all the benefits of search
What you want is a database built on search technology
A NoSQL database that thinks in terms of documents
Enterprise-grade so the data is safe, secure, always available
With ACID transactions
Updates happen immediately
But built using search-style indexes
To support the hardest queries, whether they’re search or traditional database
All the speed and scaling you expect of search
But not limited to just search
And way better than relational
Which, let’s be honest, sucks.