Why I won't be switching to Disque(charlesleifer.com)
charlesleifer.com
Why I won't be switching to Disque
http://charlesleifer.com/blog/why-i-won-t-be-switching-to-disque/
3 comments
Hi, I'm the author of this post. First of all thank you for responding to my comments, I'm humbled that you took the time and I apologize if you felt compelled to defend your work.
I was in part responding to the comment in the "Limitations" section of the README about how Disque was developed in "astronaut mode". Mostly, though, my point was that I think many people use Redis as a queue because they want to use Redis as a queue. Maybe it's already part of the stack, or they like the simplicity, whatever.
I am a huge fan of your work and hope to see Disque succeed, again apologies if you felt I was attacking your project.
I was in part responding to the comment in the "Limitations" section of the README about how Disque was developed in "astronaut mode". Mostly, though, my point was that I think many people use Redis as a queue because they want to use Redis as a queue. Maybe it's already part of the stack, or they like the simplicity, whatever.
I am a huge fan of your work and hope to see Disque succeed, again apologies if you felt I was attacking your project.
The author of the article declares his incompetence at building and maintaining software projects for which he's not the target user, and thus concludes the same applies to the creator of Redis and Disque. Based on that he instills a lot of FUD around a project that has proven very useful for me so far.
Also, from the article:
The other reason I don't foresee using Disque is alluded to in the author's own comments. He observes that many people are using Redis as a message broker, and decides that maybe there is a need for a "Redis of messaging". I would say the opposite is true, and that instead of another message server, people want to use Redis!
I disagree. I've been using Redis for queues for a very long time, and I'm in the process of migrating all those queues to Disque.
Also, from the article:
The other reason I don't foresee using Disque is alluded to in the author's own comments. He observes that many people are using Redis as a message broker, and decides that maybe there is a need for a "Redis of messaging". I would say the opposite is true, and that instead of another message server, people want to use Redis!
I disagree. I've been using Redis for queues for a very long time, and I'm in the process of migrating all those queues to Disque.
As much as I like Redis and Antirez, I have to completely agree with this article.
I would just point out that Pivotal (current Redis sponsor) runs RabbitMQ; no doubt that they have in-house knowledge about message brokers. No idea if they are willing to put time into Disque.
First version of RabbitMQ was released in 2008 and there have been more that 50 releases so far[2]. No doubt that maintaining a product like this is a lot of effort.
[1] https://www.rabbitmq.com/
[2] https://github.com/rabbitmq/rabbitmq-server/releases
I would just point out that Pivotal (current Redis sponsor) runs RabbitMQ; no doubt that they have in-house knowledge about message brokers. No idea if they are willing to put time into Disque.
First version of RabbitMQ was released in 2008 and there have been more that 50 releases so far[2]. No doubt that maintaining a product like this is a lot of effort.
[1] https://www.rabbitmq.com/
[2] https://github.com/rabbitmq/rabbitmq-server/releases
I think I can repeat the same with Disque, since I'm not an user, but the design is an effect of what users experience and complain about Redis as a message queue.
Moreover both for Redis and for Disque in the future I'll be working more with people having large deployments and use cases in order to get more feedbacks about what could be useful things to add.
I don't want to sound immodest, but the rule that you have to be an actual user of the software to create something decent does not necessarily applies to anyone, there are people that can synthesize the needs and design APIs without actually implementing anything serious. However note that don't implement anything serious does not mean don't implement anything at all. I created several "toy projects" using Redis over the years, and I'm doing the same with Disque.