Bo Simonsen wrote in a message to Roy J. Tellason:

 BS>> Well right now I don't have the understanding which is required for
 BS>> programming B-Trees which a V7 nodelist is using..

 RJT> I can't quite remember the guy's name,  at this point,  but there 
 RJT> was a guy in C_ECHO way back when (Dave something...?) who did a 
 RJT> bunch of work on that stuff in the first place.  

 BS> The Bonk-man?

No,  the guy who originally came up with the v7 nodelist stuff.  Maybe Bob
Stout would know?

 RJT> I *think* that this code was an outgrowth of the Mix Database 
 RJT> Toolchest code,  which does pretty similar stuff,  but uses one 
 RJT> index file for all indexes rather than the separate ones that the 
 RJT> nodelist setup uses.

 BS> Okay.

 RJT> While this package isn't to the best of my knowledge freely 
 RJT> available,  it _is_ very inexpensive (I think I paid $20 for it, 
 RJT> and maybe a bit more -- another $10? for the source) and the book 
 RJT> that comes with it is worthwhile for the price,  explaining that 
 RJT> stuff in more detail than you probably wanted to know...

 BS> I've a book called "Data Abstraction and Problem Solveing in C++"
 BS> it describes very clear B-trees and so on..

Yeah,  but I think this may have some specific application to what you're
trying to deal with.

---