I'm a little slow on the uptake picking up beta 2, so I just noticed this issue.
There is a note in the changelog for this release about floating point precision
parsing command line options being reverted back to v3.6 semantics. I note that
this is causing things to break in my v3.7 and higher scenes (the only version I
have used recently).
The code in question:
#declare camera_location=<0, 0, 0>;
On the command line (windows) I use: Declare=test_case=1.1 and the case is not
picked up in my switch. Printing the value of test_case out with str() and a
precision of 7 or higher, I see trailing non-zero values that are causing this.
I have a work-around to truncate the value to 3 decimal places, as I never use
more than that to designate a test case number, but thought I'd pass this along
as it can be difficult to debug in a scene.
-- Chris Rath
Post a reply to this message