OpenSearch – open-source search and analytics based on Apache 2.0 Elasticsearch

I doubt it. Their forum seems pretty dead, Elasticsearch has the known name, etc etc. Seems a bit like MariaDb/Mysql where Mysql are still the default choice for most companies (over MariaDb).

I know some people on hackernews etc are upset with the decisions of Elastic. Suspect most people couldn't care less, even if they are even aware.

We’re looking to migrate from Elastic to OpenSearch for these reasons. We’re a really huge elastic user.

Also the EOL support on elastic versions are pretty aggressive now. It’s been a challenge for us to keep up with it.

FD: I have a friend who works at Elastic, though he doesn't really colour my opinions of things.

> Firstly, dick moves like this: https://github.com/elastic/elasticsearch-py/pull/1623

I understand that this is unpopular, but you can make a very strong argument that it's to prevent weird errors in the future. I'm also guilty of littering my code with Asserts to ensure the universe is working fine.

The alternative is to allow it to work and then you end up with weird issues like when you connect mysql client to mariadb server (and vice-versa): https://stackoverflow.com/questions/50169576/mysql-8-0-11-er...

> Secondly, I don't buy the argument from Elastic any more. Yes, the ethical thing to do when you're making money from someone's work is at least contribute back. At the same time though, they're making money from packaging it up and selling it _as a service_. That "as a service" part is where they're making the bucks.

That's just an opinion, yes they have a service, and yes it competes with Amazon. Is it cool for Amazon to take a body of work and sell it without supporting it? Are amazon actually supporting it? Is it the same as Elastic using Lucene? (not really because Elastic submits a the majority of fixes to Lucene, but, you get it).

it's kinda gray, I'm sure Amazon thinks they're the good guy, but it's hard for me to look at Elastic as the bad guy in all this.

We switched to OpenSearch from Elastic, as we have an open source platform. It's actually better than open source elastic - with the security plugin and k-nn plugin (an embedding store):

https://www.hopsworks.ai/post/open-distro-for-elasticsearch-...

Totally agree, and that PR seems pointless as AWS OpenSearch has an advanced configuration option that makes OS pretend to be ES for backward compatibility so not really sure what they're achieving with it.

Maybe don't use an overly generous license if you want to open source, but not really...

Amazon's thing is the network effect, not margins on a particular feature.

I'd doubt there's huge money being made by Amazon, but if they give a profit cut to one or two companies that make products AWS offer as a hosted service, then a bunch more will probably also start changing their licensing to match. Putting resources into building an alternative that discourages this is probably not as much as it might appear (there are already plenty of developers at Amazon, and it's unlikely they're all working on other things that need to take priority), so it may be a good investment.

Elastic-the-company have over 200 million USD in revenue per quarter and its growing. It's not a small company that needs protection.

I can't say enough nice things about Typesense. The team is super responsive, the hosted version is performant and affordable. There's some features missing when compared to other similar projects. But nothing terribly show stopping.

There aren't really restrictive open source licenses since any restriction you would actually want for business protection makes it not Open Source.

I have zero problem with companies making proprietary software and we should call it as such. If the way your business makes money is by selling or hosting your software then, ya know, not a good fit for OSS -- totally fine. If you develop a piece of software because it solves a problem for you or your business and you think it might help others and you want community contributions then OSS is for you.

Feature bloat, bad documentation, API models that are nearly inscrutable. Backwards-incompatible changes are constant, requiring reindexing.

I’d like to see the architecture re-thought to include an indexer, like Apache Druid. I’d like to see a proper command line tool (‘searchctl’?), and API clients a-la gRPC (to be less Java-centric). I’d generally like the project to go back to basics and design for 2022 applications.

I am no fan of a lot of the design choices or the effort required to operate large ES clusters, but there are easily several hundred man-years of development effort behind ES. It would be a really long time before a team starting from scratch could produce something better.

I can’t say I disagree. Lucene is a beautiful, scientific thing; but ES is an increasingly-complex mess of opaque API’s and broken promises. Search engines are really, really hard. There is so much room for something better, more full-featured, more tunable, more reasonable.

Oh, of course. But an open source implementation of ElasticSearch is (based on my experience of open source generally) likely to end up in core distributions again, and have longer support lifetimes/more backports. Plus it makes it easier for SaaS platforms in a lot of ways.

This is true, but it is a huge proxy for investment and survivability. In the beginning, it was hard to tell how much Amazon would invest in OpenSearch, but it appears now (based on activity) that Amazon is committed to OpenSearch.