Fork me on GitHub

@wilkerlucio Getting really interested in pathom very impressed by the documentation. Have you used placeholders in a fulcro app? Because that is an issue we have run into multiple times.


@mitchelkuijpers thanks for the interest 🙂, and yes, putting a placeholder is a default for me on new projects, I'm currently using it as (p/placeholder-reader ">"), so it kind looks cool when you use: [{:>/alias [:query]}]


I'm still putting together a decent strategy for server side reads. I can cheat and just dispatch on query root keyword and ignore the rest of the query. So now I'm looking into pathom, but I'll see if it covers my usecases.


So I went through the first part of the Getting Started documents and have a well-functioning site with figwheel. I’m having some trouble converting this to a stand-alone server setup for heroku. I’ve compared what I have to the fulcro-template but that is so full and heavy with dat mutation and so forth. Anyone out there start with the getting started section and convert to a heroku setup?


@eugekev I remember it just being a config file with :env.edn/PORT for the server port


And a simple Procfile


You can switch config files with command line option -DconfigFile I think.


Might just be -Dconfig


Thanks @tony.kay - I think my server.clj and cljsbuild are borked up. Trying to reverse engineer.


@eugekev Yeah, there’s nothing all that special about it…just need the compiled js in the jar file, a config file that can pull desired server port from environment, and a Procfile.


You might consider adding some special sauce to you project.clj like this:

:aliases {"jar"       ["with-profile" "with-cljs" "jar"]}
  :profiles {:with-cljs {:prep-tasks ["compile" ["cljsbuild" "once" "YOUR-PROD-JS-BUILD"]]}}


so when you run jar you automatically get a cljs compile of the correct build