Rlab2 is an evolution of Rlab (or Rlab1). In case you are not familar with Rlab1 it is a very high level language for scientific/engineering applications. Rlab is _not_ a clone of the Matlab language. Although it offers similar mathematical operations on numeric objects (+,-,*,.*,/,./,^,.^,',.', etc...) The languaage syntax and semantics have been "improved" to offer a more consistent and powerful programming environment (IMHO). In addition to global and locally scoped variables, there is a file-static scoped variable (this includes functions). Multiple functions, and mixtures of statements and functions may be combined in a single file. A N-dimensional associative array structure has been also been added.
Rlab2 is mostly upwardly compatible with Rlab1. Noticeable changes are:
Sparse matrix storage is now available. This is not yet complete in the beta version, ECD is soon.
Nearly complete reference manual in HTML and Postscript formats.
Less noticeable changes are:
Overhaul of the interpreter - class interface. This makes adding a new data type conceivable
Use of Boehm's generational garbage collector. Performance has improved because of this, and the probability of encountering a memory leak is _very_ small.
Internally, object management has been improved allowing for more efficient use/re-use of multiple instance of the same object.
This version has been in testing for over 6 months now. But, I still consider it beta as the sparse matrix class and the reference manual are not yet complete. At present it is available at: