I've been using carica (https://github.com/sonian/carica) and have config map like {:dev <...> :prod <...>} and just using a single environmental variable to determine which of those config maps to use.
On Friday, January 17, 2014 5:06:31 AM UTC-8, Malcolm Sparks wrote: > > There's also Jig : https://github.com/juxt/jig > > Jig's recommendation is for configuration to be held in EDN files. > > I much prefer EDN over environment variables. Environment variables feel > awkward when there you have lots of configuration, usually in the form of a > tree. > > Usually it makes sense to have one configuration file per environment. > That can make it hard to keep configuration files up to date, so judicious > use of the #=() Clojure reader macro can allow you to programmatically > deviate from a default configuration - often the delta from a default > configuration is more informative. There are more details in Jig's > README.md file. > > I agree with the comments against using dynamic vars. They reduce the > ability to reason about functions and are precarious when dealing with > threads - binding conveyance is tricky and small mistakes can lead to > elusive bugs. > > I pass configuration as a single argument into types/records that declare > lifecycle methods. I have found this pattern to serve me well in over a > dozen projects. Since stumbling on this pattern I have had no need to > resort to dynamic vars for environment config, and my projects have been > smaller, more reliable and easier to reason about. > > > On Tuesday, January 14, 2014 9:04:45 AM UTC, James Trunk wrote: >> >> Thanks for all the great links and ideas you have all posted, now I have >> plenty of reading and thinking to do! >> >> > I am curious about what you mean by 'thread safety'. >> Perhaps "thread safety" is the wrong term, but what I meant was the >> limitations dynamic binding introduces around thread dispatching, which >> Stuart Sierra explains in this blog >> post<http://stuartsierra.com/2013/03/29/perils-of-dynamic-scope> >> . >> >> Cheers, >> James >> >> On Monday, January 13, 2014 7:10:38 PM UTC+1, Stefan Kanev wrote: >>> >>> On 13/01/14, James Trunk wrote: >>> > The downsides to dynamic vars seem to be: hiddenness, thread safety, >>> and >>> > more complex tests (binding before each test). >>> >>> I am curious about what you mean by 'thread safety'. As far as I know, >>> dynamic variables are thread-local, which means that they are >>> thread-safe, at least to some extend. I assume you mean something >>> specific? >>> >>> -- >>> Stefan Kanev ¦ @skanev ¦ http://skanev.com/ >>> If a program manipulates a large amount of data, it does so in a small >>> number >>> of ways. >>> >> -- -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
