I am responding to some confusion that has arisen with regard 
to the "Lose the 0.35!" article in Dec 12th issue of EDN.  
The confusion arises from the statement that Tr=0.35/BW is a 
rule of thumb, implying that it is not based in mathematics.  
Dave Thomson,among others,pointed out that...
 
       BW=0.35/Tr 
 
is based upon sound mathematics for a single pole response. 
I did not mean to imply otherwise.  However, oscilloscopes 
rarely have single-pole responses (except when in bandwidth 
limit mode).  I was referring to the relationship for a gaussian 
response.  Howard Johnson states  in his book "High Speed Digital 
Design" that this relationship is... 
 
       Tr=0.338/BW
 
This is the "rule" I am referring to, and we typically round 
this relationship up to 0.35.  It is unfortunate that 0.35 also 
applies for a single-pole filter, as this has led to the confusion.
 
I hope this helps clarifies this issue.
 
Best Regards,
Dennis