jbelanger wrote:I'm sorry but that's more a misunderstanding on your part than a WTF. Well, you may want to say WTF but there is no accumulated error of 0.4 degrees.
I'm glad you tagged that on the end! :-) It was a double edged WTF, either it's in accurate ign wise, or it's inaccurate code comprehension wise, one of these things is true, both as bad as the other.
It's only that each of those angular values are accurate to about +- 0.05 degree. So the effective error is very small and James comment is totally valid if somewhat succinct.
How is it that they are accurate to a level that is different to how they are defined? That seems fundamentally broken.
What is the angular precision in FreeEMS for each edge of every tooth?
Currently, or in future when finished and after 4 years of development from a working base and on version 3.1.1? :-)
Currently 1 degree, but it is not possible for them to add up to a different value, and it's accurate to that level throughout, from the code:
Code: Select all
/// @todo TODO From 0 - totalEventAngleRange degrees, scale: x10, x60 or x90? 1x is NOT enough. Currently 1x (1 deg resolution) review all related code for potential overflow and put checks in place before adjusting scaling.
I guess anywhere between 1x and 90x is possible, 0.05 implies 20x scaling, which isn't a bad option, I guess :-)
You may remember my old saying "make it work, make it right, make it fast" we're still on step one for a few more weeks :-)
The other thing that really made me WTF even more than these two comments is that they are STILL having significant inaccuracies in their PID stuff, Abe was on here whinging about this, what, 2 years ago? And it's still got people complaining AND still got people fixing it
for them... It shows a total lack of QC and haphazard dev style IMO. Much like their release semantics and various other aspects.
Fred.