SNAPSHOT has full support for 32-bit floats

Skip to first unread message

Peter A

Jan 3, 2017, 1:04:30 PM1/3/17
Support for 32-bit matrices has been added and is available on GitHub.  It's auto generated using the 64-bit (double) code.  Unfortunately to get auto generated code working easily I had to change the name of just about every class.  This breaks backwards compatibility in a major way, but for the most part a simple search/replace can be used to update existing code.  E.g. CommonOps -> CommonOps_D64 and CommonOps_D32.  SimpleMatrix supports both 32-bit and 64-bit floats, but you can't mix the two together yet.

Make sure you read the README.  It tells you how to generate the 32-bit code.  If you don't it can't compile.

Since backwards compatibility is being hosed this is a good time for API suggestions.

- Peter

"Now, now my good man, this is no time for making enemies."    — Voltaire (1694-1778), on his deathbed in response to a priest asking that he renounce Satan.
Reply all
Reply to author
0 new messages