First results from NASL2 are satisfactory.
Clear syntax errors allow quick script development.
Some bugs remained in the new features (e.g. arrays), but all NASL1 script run well.
Although this was not the main goal, the interpretor is MUCH quicker.
Nothing is perfect: double parsing (for description & ID) is too slow => workaround (cache)
Warnings are too verbose: disabled by default, but can be enabled.
Who uses the tracer?