Quoting chris2be8 <chris2be8@...>:

> ggnfs-doc.pdf page 17 says "It should be at least 1". So I assumed I should
> invert the result if less than 1.

The skew being larger than 1 is a good rule of thumb, and is especially
useful becase the rest of the code assumes relations will have an 'a'
coordinate much larger than the 'b' coordinate. If the skew was much
smaller than 1 then the postprocessing code would fail to parse many relations.

However, even when the optimal skew should legitimately be smaller than 1,
it is never *much* smaller, so skew < 1 is not a big problem.

jasonp

------------------------------------------------------
This message was sent using BOO.net's Webmail.
http://www.boo.net/