In regard to improving tools. One of the most common criticism I hear about dynamically typed languages, is that you can't have precise code completion, and without this, it is impossible to know what methods can be used in given context. My usual response is, that you can, just ask the object what messages do you respond to. Then I explain what a live object is, and how coding inside the debugger work.
The problem is, neither the Debugger nor the Workspace take advantages of this. If I start typing a message inside the Debugger, the code completion won't take the type of the receiver into consideration. -- View this message in context: http://forum.world.st/Brainstorming-Pharo4-tp4761647p4761805.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.
