jpff@maths.bath.ac.uk wrote:
> b has the effect of chnaging the interpretation of all textually later
> note onset times by adding the b argument. 

Thanks for the clarification. This makes perfect sense to me. I think it
is actually a much clearer explanation of what effect b has than the
current (3.57) manual entry. I would like to suggest the following
manual entry, based on jpff's above-quoted message and some further
experimentation:


b Statement

  
  b  p1



Description

This statement resets the clock. 

P Fields

  p1    Specifies how the clock is to be set.

Special Considerations

p1 gives a value to add to the start times of all textually (not
temporally!) subsequent notes in a score section. This "clock" value is
reset each time a b statement is given; there is no cumulative effect.
If p1 is positive, the affected notes will occur later; if p1 is
negative, they will occur earlier. If p1 = 0, the clock is returned to
its original position and subsequent events will occur at their given
p2.

The b statement can be used to change the relative start times of any
number of events within a section without having to change each note's
p2. The advantage over using separate sections is that the notes
affected can thus be made to overlap or even change their order, whereas
sections are always strictly sequential.


Example

i1      0       2
i1	10	888		
b 5				; set the clock "forward"
i2      1       1       440	; start time = 6
i2	2	1	480	; start time = 7
b -1				; set the clock back
i3      3       2       3.1415	; start time = 2
i3	5.5	1	1.1111	; start time = 4.5
b 0				; reset clock to normal