I often find myself arguing with the proponents of GF on SB. Usually this is because they are dismissing Suunto (and other) computers for having a proprietory algorithm. They claim that, somehow, because the is Fortean code for GF it is more predictable and so better in every way. In particular that widely available planning software will give identical plans. Then they will quote a load of sources (the diverite graph being a particular favourite) which are wrong, or at least oversimplified, and go on to demonstrate their own lack of understanding of the small details.
My approach to understanding this was to ask a man who really knows several questions (over a long time) and to write an implementation I could test against commercial ones. There is no where to hide from the numbers, although I do have to go back to the code to check on how stuff works sometimes. Also there are things which a reasonable implementation might choose which are not covered in these discussions.
“Everything should be made as simple as possible, but not simpler.“
People really need to remember that all these algorithms are just models. They are not the real thing, they only try so hard to model stuff and have significant limits.
It is a shame that attempts to improve on these models are met with resistance and distain.
My approach to understanding this was to ask a man who really knows several questions (over a long time) and to write an implementation I could test against commercial ones. There is no where to hide from the numbers, although I do have to go back to the code to check on how stuff works sometimes. Also there are things which a reasonable implementation might choose which are not covered in these discussions.
“Everything should be made as simple as possible, but not simpler.“
People really need to remember that all these algorithms are just models. They are not the real thing, they only try so hard to model stuff and have significant limits.
It is a shame that attempts to improve on these models are met with resistance and distain.