To read all comments associated with this story, please click here.
Comparing with ML, Go is just too primitive as it is basically C + Interfaces + Modules + GC.
Lacking many of the nowadays abstractions:
- enums
- generic types
- exception handling
- functional programming concepts
The oversold channels feature can be easily done with message queues in any current mainstream language.
My opinion of course:
- enums
Superfluous
- generic types
Interfaces with go's loose binding is good enough most of the time. You are right, this would be nice.
- exception handling
Go's take is better IMHO. C style programming by contract + panic/recover/defer leads to better code (for OK/competent and better coders).
- functional programming concepts
When I want this I will invoke my Haskell program... Most of the time this stuffis marketing/buzzword fluff, again IMHO.
I don't agree with what's written on the Opa blog.
Basically, the AGPL license allows you to mix it with GPL code.
So, applications written in Opa can be GPL licensed, which means that for instance you can host the applications online without providing the source code to users. Or am I wrong?





Member since:
2005-09-03
If doing something new on the server side I think I would be more tempted by Google's Go than either of these.
http://shootout.alioth.debian.org/u64/benchmark.php?test=all&lang=v...
I think a [Server-side Go / MySQL + Client-Side GWT] project sounds fun
+ No javascript! (written), unlike node.js...
+ I don't need a commercial license from Opa*...
+ All strongly typed
+ Should be fast to run and develop!
But this is coming from a person who's web projects are ASP.NET, JSF and PHP and really likes C.
* Firstly, "if I'm using AGPL Opa to develop an app does it need to be AGPL, too?". Long story short — yes, it does. http://blog.opalang.org/2011/08/opa-license-contributions.html