A Possible Future of Sclp

The following things still need to be done:
  1. Medium term goals

    1. Support for the RIPES simulator as another target.
    2. Enhancing sclp to support structures.
    3. Enhancing sclp to support classes.
    4. Integrating interpretation for sclp.

  2. Long term goal. Generating sclp phases from specification.
There is an exciting work by Michelle Thalakottur and her team where they tried to formalize the specification of IRs and translation rules for mapping one IR into the next IR. This opens up the possibility of generating sclp phases from specification as well as generating interpreters for each IR.

This set of slides describes my vision of sclp.