New changes in this release:

• Improving vectorisation of 16 bit floating point types gpu_half and gpu_bfloat16.
• Improving performance of CPU debug mode.
• Improving performance of kernel creation.
• Setting thread names to make debugging more comfortable (not available on windows).
• Adding optional width parameter to the work_group_any and work_group_all functions.
• goopax_future callback function can access the return value of the kernel.
• Adding scripts to build external libraries for example programs.