Personally, I think this is an extremely important topic, we need to change the way we program. It isn't the 70's anymore, why do we still program like it is.
Why does it take teams of developers to create and manage applications which do really simple tasks in the grand scheme of things (and I realize the amount of complexity in building applications is staggering, but large portions of it could be better automated). Where are the auto generated GUIs, where is the ability to ship execution control to arbitrary devices, where is a hypermedia layer with independent view and presentation code?
I'm approaching this from a different angle than the light table guys appear to be (I agree with everything they are saying). My angle is an attempt to build a cross platform module system (where platform includes runtime and programming language as well operating system and architecture): https://github.com/OffByOneStudios/massive-dangerzone
My argument is: before we can build the next generation of useful tools, we need a framework for managing both generations. massive-dangerzone is an attempt at bringing next generation usage (like that described in the article) to existing tools. It's still a big work in progress though, and is barely useful at the moment.
Why does it take teams of developers to create and manage applications which do really simple tasks in the grand scheme of things (and I realize the amount of complexity in building applications is staggering, but large portions of it could be better automated). Where are the auto generated GUIs, where is the ability to ship execution control to arbitrary devices, where is a hypermedia layer with independent view and presentation code?
I'm approaching this from a different angle than the light table guys appear to be (I agree with everything they are saying). My angle is an attempt to build a cross platform module system (where platform includes runtime and programming language as well operating system and architecture): https://github.com/OffByOneStudios/massive-dangerzone
My argument is: before we can build the next generation of useful tools, we need a framework for managing both generations. massive-dangerzone is an attempt at bringing next generation usage (like that described in the article) to existing tools. It's still a big work in progress though, and is barely useful at the moment.