Closing a phase. Good bye GSoC

        When i began with these i was really ignorant about types, now i can say, i know a little bit of this huge world.

         I lerned about many many things, like graphs alogorithms, tree algorithms, reflection, meta programming, loggin , prediction, tarot, etc.

        As fruit of these days of hard work, i get for the comunity three projects:

    1. Concrete Type Inference
    2. Kwisatz Haderach
    3. Paul Le poulpe

       An inferrer, a behavior analyzer and a logger.

       Making a balance, i’m happy. I want to get more funcitionality to this point, over all i wanted to get inference over reflection and messages like become: but well, i’m not that good i supose, i needed and i need to lern and growup to make each step. But, really i cannot complain, reaching the end of this window of coding, i have added:

        Support: 

    • Subresult comparing mechanism
    • Trait extension to useit 
    • Code patterns (like conditional contexts and multiple dimension analysis)
    • Error management
    • MethodContext and thisContext support
    • Pharo 2.0 ( the core project was not really ported, just a bit)  
    • 40 new primitives 
    • fixes and generalization over all the 46 old supported primitives 
    • Semi generalization matching mechanism for primitives 
    • Named primitives support
    • Symbol as method sender ( a bit of reflection )
    • Collection common usage

     
     A suit of 992 Tests with a huge number 568 green tests ( 165 really-hard more tests from the last post 🙂 )

     An easy-to query and well tested call graph, based on the visitor pattern, builded from an integration of the inferencer-stack  management with the announcement framework.

    A highly configurable logger based on the Log4j api that supports too the logging builded in the image.

     Now, maybe you are asking your self: this is the end? The answer is  “no” Why on earth i should stop programming this? :).
     More than this, i’m about to begin a journey really important for my self. Maybe to france, maybe to spain. At least the next two years i’ll be working in this project by my self. I believe that we can build a better future, i believe in pharo and smalltalk, and i want to make more people to believe in this beatiful community and in this way to make software, but, we still needing better tools. Tools based in a dependable type information.

    
      Where is the code?

          Type inference v1.0 & Kwisatz Haderach v1.0

          Paul le poulpe v1.0
  

              

About santiagobragagnolo

Robotics obsessive, computer languages and cloud computing enthusiastic. Metadimensions walker.
This entry was posted in Call graph, Collections, Concrete Type Inference, Conditional Contexts, GSoC, Kwisatz Haderach, Mechanisms, Paul le poulpe, Pharo, Primitives and tagged . Bookmark the permalink.

Leave a comment