widman@cs.utexas.edu (Lawrence Widman) (12/04/89)
This is posted for a friend who does not have news access, but can receive internet mail. Please direct replies to freedman@euclid.math.temple.edu. "I have a question about indexing in the dBase and clone series of databases for the PC. I remember reading about some sort of variant of b-trees (b+-trees, maybe) in a dBase III+ implementation book. I also know that clones such as Clipper and FoxBase+ have their own indexing methods which are supposedly one reason why they are faster than dBase (in addition to preparsing/compiling, cacheing, ...). Can anyone give my information about what algorithms dBase III+ (and the clones, perhaps) use to do indexing? "If there is any interest in this topic, I will post a summary of the replies to the network." Thank you in advance, Avi Freedman (freedman@euclid.math.temple.edu)
awd@dbase.UUCP (Alastair Dallas) (12/06/89)
In article <7284@cs.utexas.edu>, widman@cs.utexas.edu (Lawrence Widman) writes: > This is posted for a friend who does not have news access, but can receive > internet mail. Please direct replies to freedman@euclid.math.temple.edu. Sorry, but it's a heck of a lot easier for me to post than to email under our current setup. > Can anyone give my information > about what algorithms dBase III+ (and the clones, perhaps) use to do indexing? The answer is: no, anyone (like me) who knows the algorithms that dBASE III uses can't divulge them. However, if you were to read "The Art of Computer Programming, Vol. 3" by Knuth (Addison-Wesley, 1973!) pp. 473-479, with particular attention to the bottom of page 478, you'd have a fair idea of what's what. It's not like we're talking atomic secrets here :-). Hope it helps. /alastair/