[artellectual]

Elastic is a company I’ve been following from the start when they just had elasticsearch opensource project. Watching this company grow to this point has been inspirational. I really want to meet / talk to the people who made this company. They made open source work.

I’m happy to use their products, filebeat, kibana, elasticsearch, and their cloud offering has been top notch. They got a lot of things right.

Now elasticsearch even supports SQL as a query language.

This is a company that you don’t hear often in the news, but they just kept their focus and shipped the goods.

As you can tell by now I’m a hardcore fan of Elastic.

[feeblebee]

I'm also a big fan and I'm really very happy for them and their success.

I was using elasticsearch 7 years ago in production already. And just for me, from the outside, it always looked like a crazy trip. I knew there was something going on when nearly half of my twitter timeline tweeted "am working for elasticsearch now". (they renamed to elastic later)

I was chatting in early 2017 with Shay and along the way he mentioned they were almost 500 then and now ~18 months later they are +800. Crazy!

> They made open source work.

Yes!

They have built and still build impressive (open source) projects and they managed to still earn money :) - this is the real big thing!

[ousta]

i would be surprised if they were supporting sql as query language. do you mean by using spark sql?

[vquemener]

Actually, they really do support SQL as a query language : https://www.elastic.co/en/blog/an-introduction-to-elasticsea...

[lnx01]

There's even a JDBC driver on the commercial tier

[ousta]

yeah a "subset of SQL" and on the page you gave me 0 example of a JOIN. so no JOIN no SQL sorry

[AtlasBarfed]

"We support SQL on our distributed data platform" has to be the new big lie of distributed systems salespeople.

I'm not saying that computing joins, aggegates, sorts, etc can't be done in a distributed fashion, it just takes time, a lot of network reliability/reattempts, and understanding of the approximate nature of results if there are ongoing "eventually consistent" updates rolling through the cluster(s).

[ousta]

well you cannot have a search engine query time + a SQL layer at least not at DB speeds