## Abstract

Proofs are programs, so to get a program it suffices to get a proof. What kind of proofs do we need/use for declarative programming? I shall look at the two standard paradigms: logical and functional programming, and show why they are different from a logical point of view. Then I shall look at some possible alternatives between these two and also beyond them.The effect of domain specific knowledge on the proof (and therefore the program) will be considered. A number of ways in which the fact that we have domain specific knowledge can be employed in designing new programming languages will be discussed and examples will be given. Finally the \ideal" solution will be considered and I shall show a) why it is not practical and b) how it can be made practical.

Original language | English |
---|---|

Number of pages | 1 |

Journal | Electronic Notes in Theoretical Computer Science |

Volume | 61 |

DOIs | |

Publication status | Published - 1 Jan 2002 |