README

	Dylp: An implementation of the dynamic simplex algorithm

  Dylp is a linear programming solver, based on the dynamic simplex algorithm.
  Roughly, this means that dylp attempts to maintain an active constraint
  system which consists of only those constraints and variables that are
  relevant at the current extreme point. The bet is that this will be a
  useful property in situations where the same constraint system is
  repeatedly modified and reoptimised (e.g., in a branch-cut-price MIP
  code).  You can also tell dylp to operate with the full constraint system,
  which is useful for obtaining an initial solution.

  Dylp is intended as a development code. It can produce a range of
  statistics, and it can tell you more about how it's solving your problem
  than most people will ever want to know. It also includes extensive
  cross checks to catch errors during program development.

  Generally speaking, if your computing environment supports IEEE floating
  point then it should be possible to build dylp with very little effort.

  ============================================================================

  Maintainer: Lou Hafer (retired)      lou@cs.sfu.ca

  See README.md for additional comments.