Why you should stop writing shiny apps

[JosiahParry]

Or, "No, you don't need a shiny app" or "Stop making Shiny Apps"

data scientists are not web developers. They should not be making web apps.

• high maitenance cost
• R is not a web development language
• it is not multithreaded async is weird
• shiny is more extensible to people who want to become a front end developer such as react js
• R is more extensible to those who want to build backend solutions a la go, rust, and node js

if you find yourself wanting to make a shiny app only do it as a proof of concept. find someone in your organization who specializes in making websites

[JosiahParry]

When people think of R the first few things people think of are niche statistical libraries, ggplot, and shiny applications. "RShiny" is all over LinkedIn. I think the focus on shiny apps has been detrimental to the development of R as a language. Maybe it is because shiny apps feel like a real manifestation of the work you've done. I don't think Shiny apps are intended for real large scale use. They're great for small applications that will not have a lot of traffic or for a proof of concept. I use shiny apps (mininmally) in my own work to demo proof of concepts for tools.

[JosiahParry]

Outline maybe:

• what is shiny
• why its dominated the R discourse
• why i think it has been detrimental to R's development
  • bad look for many half baked apps
• why i dont think R users should be so concerned with shiny
  • data scientists shouldnt become web devs
  • its not as scalable as other things
  • detracts you from real work that you are better suited for