3 ms·
I'm not sure I agree with this - there have been many postgres features over the past ~10 years that got me excited, some of the highlights that immediately spr
by mnahkies 5y ago
I'm not sure I agree with this - there have been many postgres features over the past ~10 years that got me excited, some of the highlights that immediately spring to mind:
- JSON / jsonb
- improved partitioning support
- identity columns
I agree that I wouldn't want to trust a new entrant to the space unless it was subject to rigorous testing, eg https://jepsen.io/ https://jepsen.io/
- Nextgrid 5y ago> postgres Postgres has the advantage that it's still open-source, can be ran locally and self-hosted (and has a healthy ecosystem of managed providers) and won't go away anytime soon. None of these are true for all these new DB startups.
- geophile 5y agoDoing well on the Jepsen tests is a very good sign. But it isn't anywhere near enough. Other things to consider: - The product has bugs not addressed by Jepsen. - Performance is inadequate. - The optimizer is immature, and you find yourself struggling to understand and work around its limitations. - Replication is missing or doesn't work well enough. - There is absolutely no substitute for N years of real-world experience, no matter how rigorous the company's testing. - Impossibility of finding people already familiar with the product, because they are all employed by the vendor. - Even aside from product issues, is the company profitable? If not, how long is it going to be in existence?
- redis_mlc 5y ago