Hi Edgar,
No we did not loose faith. We think it is good to separate execution/deployment mode from development mode. (isn't this the same reason why we have virtual machines in the first place?). In execution mode you have different requirements (e.g. scalability, speed, supporting hosting providers) compared to being in development mode (e.g. live debugging, instant feedback, reflection).
This graduation project is not about going away from Smalltalk or going to java, python or whatever other technology with an identifying name. It is about investigating separation, and we believe it is a good approach to pick an execution platform which is especially not-smalltalk to help us prove/investigate if our "business descriptions" are indeed as implementation independent as we hope they are.
Of one thing we are sure though: In many many many ways Smalltalk is absolutely superior (even after all those years it is around!) but is not the end, nor is Python, nor is Java.
greetings, Mathieu van Echtelt
On Fri, Nov 28, 2008 at 3:02 PM, Edgar J. De Cleene edgardec2001@yahoo.com.ar wrote:
On 28/11/2008, at 09:54, Mathieu van Echtelt wrote:
A student could research if and how our current business domain model
descriptions can be executed on an alternative execution platform
(e.g. Java, Python). This research might conclude other platforms are
not easily supported, because certain constructions in our current
descriptions are to Smalltalk specific. In such a case, it should be
made clear what the consequences are to remove the Smalltalk
dependencies from our current business domain descriptions.
So you lost faith. Believing Java or Python solves some and Smalltalk not put you almost out of Our Magic Kingdom :=) I not saying you can't got more money with your choice... Edgar
-- AG5 - Product manager
squeak-dev@lists.squeakfoundation.org