@alon24 this is Travis-CI.
2 versions one with float enabled, and the other with float disabled only using integers.
Floating point uses more memory, not allowing code that worked on firmware builds without float enabled to load on float enabled firmware without memory issues.