[Bioc-sig-seq] trimLRPatters: inconstitency re max.[LR]mismatch

[Harris A. Jaffee]

On Feb 17, 2010, at 1:56 PM, Harris A. Jaffee wrote:
> So, I would change the man page (and its supporting code) to  
> something like
>
> 	A single numeric value in '[0, 1)' is taken as a "rate" ...

Sooner or later, someone will realize that there's also a  
discontinuity at 1,
but I want to say that nobody needs such a rate in the first place.   
Thus, we
might limit the legal rates to [0, .5], or whatever.  We could also  
disallow
single positive integer values of the max.L/R-mismatch parameters,  
which don't
have a very productive implementation, IMO, and I don't know if there  
is one.