A bit of a weird article.
It announces super streams-which are basically topic partitions in Kafka, acknowledges that, but then hand-waves “oh but they’re different” and then spends the rest of the article talking about all the features that are identical to how topic partitions work…
It’s a good feature, there’s nothing wrong with that, but there’s nothing wrong with saying “we’ve brought this feature to RabbitMQ too” rather than trying to pretend it’s totally-not-the-same.
Here's the full paragraph where they talk about Kafka:
> Let’s talk about the elephant in the room: how does it compare to Kafka? We can compare a super stream to a Kafka topic and a stream to a partition of a Kafka topic. A RabbitMQ stream is a first-class, individually-named object though, whereas a Kafka partition is a subordinate of a Kafka topic. This explanation leaves a lot of details out, there is no real 1-to-1 mapping, but it is accurate enough for our point in this post.
I think that's fine. This is a post about a new feature in RabbitMQ, from the maintainers of RabbitMQ. It's not intended as a RabbitMQ-Kafka comparison post.
They acknowledge that it's equivalent to Kafka, and then chose not to spend time digging into the details of how it's different - I'm willing to take their word for it that there are all sorts of interesting technical differences here, and I'm fine with them not addressing those in the post where they announce their new feature to the world.
I don't see this as them pretending that it's totally-not-the-same.
It’s a good feature, there’s nothing wrong with that, but there’s nothing wrong with saying “we’ve brought this feature to RabbitMQ too” rather than trying to pretend it’s totally-not-the-same.