Is sindi 1.0 far ahead ?

20 views
Skip to first unread message

Francois Armand

unread,
Jul 19, 2013, 4:38:25 AM7/19/13
to sindi...@googlegroups.com
Hello guys,

I was recently interested in enhancing our service configuration in our open source software, Rudder.
 
We used to use Spring (yeah... Rudder started as a Java project 4 years ago...), but at least with a Java Configuration and only constructor injection.

We recently switch from Spring to a "AppConfig" object with a hundreds of lazy val. That is not nice (not modular, need reference to an object were services are needed), but at least runtime is fast again, and there is no magic involved.

So now, for the next step, I want to work on using a nicer DI library, especially to help on testing (now, for tests, we have to build services by hand, it would be nice to be able to reuse some part of the general config), be able to use a kind of "autowire" for all non ambigus services, and be done with all the lazy val (we need them because of initialisation order maze - perhaps we could use vals with a micro-managed instantiation order of services, but it won't be nice to maintain...)

Along with my search, I look to Shapeless HList (they could have done a really nice register, but they don't scale really well for compilation time) and the next item is Macro based ioc, and so Sindi.

So, my question are:

  • Sindi 1.0 looks really nice, is it far ahead ? Is it still developped ?
  • How Sindi managed initialisation order of service under the hood ? 

Thanks !

Reply all
Reply to author
Forward
0 new messages