[Rivet] rev. 890 broke CDF_2008_DRELLYAN

[Andy Buckley]

Hendrik Hoeth wrote:
> Hi,
> 
> I received updated numbers and cuts for the CDF underlying event
> analysis in Drell-Yan events. When I changed the analysis I realized
> that the CDF_2008_DRELLYAN is broken since revision r890.

How broken? As in "it segfaults", or "the plots are bad"?

> Andy, do you have any idea what could be going on here? It seems to be
> related to the changes you made to the math classes.

The math class changes were:

 * Using Rivet namespace for math util functions (this was a bigger
change than it sounds - almost a day's worth of rewriting)
 * Better range checking for angular ranges
 * A few other things that should have no user effect

Which functions are being used?

Andy