Using Mount in PROD ?

[amitvshah25]

Thankyou for a great library !

Any reason why I should not use mount in Production ? I have a fairly simple set of requirements, but touching lot of 'other systems' .. Monitor a socket and a mail box live, and push the updates through a websocket to browsers. Stores the data in database. Over all, the code totals to around 90 lines, so I am thinking of throwing it all in just one namespace .. ??

My developement is emacs+Cider REPL, and mount is working great. I am not using the reloaded workflow, or Components. And now I am thinking of just doing a (mount/start) in my main method, and uberjar it .. ?

Thoughts ?

[tolitius]

thanks for the feedback :)

Any reason why I should not use mount in Production

It usually comes down to personal preference, rather than any technical reasons.

mount is there to just kick it off and shut it down. Most of the runtime is spent on the code that you write.

Some people prefer other ways to manage state, and they are pretty vocal about it. Others spend that time creating things by using the tools that work and that they enjoy.

Here is what I know:

• There are quite a few production projects around the world that use mount
• Some of them are Luminus based, some just straight up mount
• I use mount in production for over 2 years in a dozen of projects

I don't claim that

mount is the best way to manage application resource state in Clojure

but I do claim that:

mount is the best way to manage application resource state in Clojure for me