As noted in the scope section along with some other omissions:

> Only free and open source time series databases and their features have been compared. Therefore if someone asks “have you tried Kdb+ or Informix?” the answer will be no. They are probably awesome though.

Article could have been titled "Review of the Top10 FOSS Time Series Databases" since that's what it is.

The 32-bit version is free.

I agree that any comparison skipping it is meaningless, it is the "standard" in this space, to the extent there is one.

> The 32-bit version is free.

Only for non-commercial use: https://kx.com/2015/09/19/32-bit-kdb-for-non-commercial-use-...

KDB+ does have some disadvantages though.

Requiring a schema makes it ill-suited to many use cases since in many situations you don't know the schema ahead of time. And the biggest for me is poor Hadoop/Spark integration. JDBC is not a great choice since you have poor predicate pushdown support and it prevents say Spark from going specifically to the node with the data.

You always have a schema. The only choice you have is whether your tool helps you validate it or not.

You always have N>0 schemas. The question of whether you have N+1 or not is governed is whether your db requires you to have a second one other than the one in your code.

I dont understand the how/why it is, though. anyone have more information?

Performance is not the #1 criteria.