HACKER Q&A
📣 craxel

Next Generation Database


If you had the next generation database that may make existing databases obsolete, how would you go to market?

What do I mean by next generation database? How about a high performance, encrypted database that never has the encryption keys on the servers? How about an encrypted database where every single tuple/record can be encrypted by the application with a different encryption key? How about an encrypted database that supports spatial, time series, graphs, and much of SQL while minimizing inference risk? How about a distributed database with massively parallel consensus for encrypted, ACID transactions (million+ per second) and replication for high availability? How about a distributed database that supports trustless/immutable transactions - in effect is also a distributed ledger?

You probably don't believe me yet - but our probabilistic spatial hash/filter with some cryptography turns out to be the huge breakthrough for searchable encryption (more specifically securely indexing encrypted records) -- not fully homomorphic encryption.

How would you go to market? Developer adoption? Enterprises? Government?

  👤 oldandtired Accepted Answer ✓
I have seen all sorts of different database technologies being touted as the "next generation" database technology over 4 decades. Each has not superseded any of the others. What ends up happening is that a "new" database technology finds an application area for which it is the "best fit" and is then used accordingly. Some of theses application areas can be broad and some can be very specialised.

If you can build a usable system and can make a market for it, do so. If you think it has broad application and want it to be ubiquitous with a variety of players (including FOSS systems) providing implementations then release all the relevant implementation and research information.

But you first have to decide if you want to give the basic ideas away or whether you want to hold them close to your chest. Until you make that decision, a lot of the questions you have asked are not going to be so answerable. There are consequences for giving the information away and there are consequences for keeping it close to your chest (including patents, trade secrets, etc.)

The application fields where you think you technology may be benefited from would require you actually building full-blown demonstrations systems against which you can compare existing technologies. I have seen many different systems over the decades which were simply "flash in the pan" technologies and as such take your comments with a couple of tonnes of sea-salt.

If your technology is both viable and worth the price and convincing, you might make it a successful venture.