[cdp] Does extend-zigzag advance?

[Peter P.]

Hi list,

I hope everyone is healthy and fine!

extend-zigzag is one of my favourite sound transformations in the time
domain. 
It has two modes
1: random zigzags: starts at file start, ends at file end.
2: zigzagging follows times supplied by user.

See its html doc here
http://www.ensemble-software.net/CDPDocs/html/cgroextd.htm#ZIGZAG

I am wondering if it could advance gradually through the entire
soundfile in mode 1 while zig-zagging?

It seems that the probability for a forwards- and a backwards read are
currently the same, hence it will ever more or less zigzag around the
same region. I assume it would need a separate parameter to make
advances slightly more likely than reverse reads then?

In mode 2 it is said to read zigzag start and endpoints from a text
file. I can't seem to understand how this works from the docs. Would I
specify times like
1	3
2	4
7	8
as intervals? How long would it zigzag within each of the intervals
then? What does "zigsteps moving in the same (time-)direction will be
concatenated." from the docs mean? What would happen otherwise?

Does another CDP sound transformation achieve the desired effect?

I am excited to hear any answers or ideas,
cheers, P