NavList:

Frank:

I've only skimmed over the article a little bit so far (but it seems to spend the first two pages complaining more about its perceived audience than the algorithm itself?), but I do know that there were a few responses published in the journal.  In scooping up papers on refraction algorithms a while back I found Bennett's "A Plea for Simplicity," which mentions a response written by HMNAO's Yallop.

I think it was Yallop who pointed out that floating point errors can result in trying to determine an angle with a cosine greater than 1 or less than -1, which required a check that Pepperday complains about existing.  And I think I remember someone (Yallop again?) mentioning that different trigonometric functions on a calculator return results in different quadrants, where one may return a result in the range of 0° through 360° (more appropriate for hour angles), and a different button the same calculator would give a result between -180° and +180° (more appropriate for longitude).

Personally, I think the instructions should have kept a fifth decimal place as a guarding figure (it seems the algorithms were written for systems with limited storage and the user would have had to "store" some intermediate results with pencil and paper).  But more than a few NA users seem to believe the sight reduction tables in the back of the book can give azimuth to 0.1° precision, so it's probably best to leavea as-is.