[LEAPSECS] Documentation review as important as code review

[Gerard Ashton]

Poul-Henning Kamp made some inquiries about how quickly Rob could review 
code. I suggest this question misses a few important items.

1. Documentation. If the documentation for any given system claims UTC 
agrees with UT1 or GMT within 0.9 s, there is the danger that some 
programmer who only looks at source code, program documentation, and 
reports of UTC and UT1 at some point in the 2020s will think the program 
is malfunctioning, and attempt to fix it without first getting a more 
complete understanding of the situation.

2. Absence of documentation combined with assumptions. Basically the 
same problem as 1, except instead of explicit documentation, the 
programmer relies on what he remembers from some college course on 
surveying, astronomy, etc., or perhaps a navigation course he took years 
ago from the U.S. Power Squadron or the like.

Gerry Ashton