2020-06-15 12:59:39 UTC - Alexandre DUVAL: Hi, did open an issue about that? ---- 2020-06-15 12:59:42 UTC - Alexandre DUVAL: You* ---- 2020-06-15 15:00:33 UTC - Cathy Thompson: @Cathy Thompson has joined the channel ---- 2020-06-15 15:41:59 UTC - Addison Higham: it is already fixed in master (after I implemented the same fix :slightly_smiling_face: ) ---- 2020-06-15 17:46:20 UTC - Alexandre DUVAL: Oh okay :slightly_smiling_face: ---- 2020-06-16 00:56:46 UTC - Devin G. Bost: I just got this email from the author. Any thoughts on how I should handle it? ---- 2020-06-16 02:30:57 UTC - Ali Ahmed: Confluent has not been a good steward of truth , first they denied that an alternative existed to kafka, not they are denying the benefits while replicating the pulsar design, best avoid them. ---- 2020-06-16 02:42:46 UTC - Anup Ghatage: Which parts of the design are they replicating? ---- 2020-06-16 02:46:22 UTC - Ali Ahmed: they are trying to decouple storage in shades, before they use to defend it. ---- 2020-06-16 02:46:47 UTC - Anup Ghatage: Any place I can read more about this? ---- 2020-06-16 02:47:16 UTC - Anup Ghatage: Last I heard anything it was about getting rid of Zookeeper from them ---- 2020-06-16 02:47:34 UTC - Ali Ahmed: confluent has some blogs on there site about it. It’s not in open source. +1 : Anup Ghatage ---- 2020-06-16 02:48:51 UTC - Ali Ahmed: it’s because they misused zk , using it for metadata, now are going trying to replace it with a raft based state store. ---- 2020-06-16 02:49:11 UTC - Anup Ghatage: As somebody who’s worked in this space for sometime, i’ve seen closely how much of a pain it is to decouple existing legacy software from its storage.. takes forever, never lives up to ideal vision. Too much baggage. ---- 2020-06-16 02:52:35 UTC - Ali Ahmed: the capability gap in enormous despite there best efforts , they can’t can’t guarantee data durability, have no notion of mulitenancy, can’t support different message semantics can’t scale elastically. I am sure they will keep trying to retrofit kafka but I don’t see it working well. +1 : Anup Ghatage ---- 2020-06-16 02:57:08 UTC - Jerry Peng: @Devin G. Bost the blog is just pure marketing material for Confluent and makes no genuine effort to be unbiased thus. If you read through the blog, anyone can clearly see the recurring theme is to use half truths and shaky logic to explain away Pulsar’s strengths and make excuses for Kafka’s weaknesses. Thus, I don’t think it’s worthwhile to respond to the email. Though I think it it is worthwhile to write blogs to combat Confluent misinformation compaign. If you are interested in writing a blog to address the misrepresentations of this blog I am sure the community is will to collaborate with you :) ---- 2020-06-16 03:29:29 UTC - Penghui Li: @Penghui Li set the channel topic: The Apache Pulsar 2.6.0 Candidate 1 is now out for votes, please help verify and vote. <https://lists.apache.org/thread.html/rd99e22bed6c35ced08eff8c929ff5c094e1f6268af09c39bc3c45132%40%3Cdev.pulsar.apache.org%3E> ---- 2020-06-16 03:31:52 UTC - Penghui Li: Hi, all. 2.6.0 Candidate 1 is out for votes, please help verify and vote, thanks. ---- 2020-06-16 05:32:33 UTC - Patrik Kleindl: Coming from the Kafka world (not side) my first impression of Pulsar was that it copied a lot from Kafka eg IO vs Connect and Functions vs Kafka Streams. With KOP it actually copies the Kafka protocol. So you might want to go easy throwing these stones. Pulsar has some great features but features aren‘t everything. ---- 2020-06-16 06:21:30 UTC - sundar: Can you help me out a little in this? As in can you give me few starter small big fixes that I can do? I've used pulsar but I'm not familiar with the codebase and seeing issues is a bit overwhelming right now... ---- 2020-06-16 06:36:13 UTC - Patrik Kleindl: @Devin G. Bost Pick one thing that is wrong in your opinion and put an argument to it. Repeating “all lies, Pulsar is only good, Kafka/Confluent is only bad” does not help anyone and is not true either. You can even invite Kai over here to discuss this. ----
