Hi,

> some professor of mine who was in love with b-trees kept on talking
> about b-trieve, and how terrrible (?) it was that they now belong to
> pervasive or something like that.

Well, a B-tree is a classical data structure, while B-trieve is a
library for managing structured indexed files (ISAM). This library
(proprietary software) was very popular on the golden Netware years. The
company was renamed Pervasive and built a relacional, sql-based,
database server on top of the Btrieve libraries.

Believe me, Btrieve was a pain to develop for. The API was very
low-level, and it had to be that way as it was meant to be
cross-language but had no object model to use (the old DOS days). It was
very fast and reliable, but I won't miss it. :-)


> it seemed as if b-trieve was somehow capable of storing data internally
> as trees or something like that, but i haven't heard anything about it
> for a while, so i guess it got a bit obsolete...

Most databases availabe today (and during the latest decades also) use
B-trees as a way for organizing indexes, and the simpler ones (like
Btrieve itself, C-ISAM and others) used B-trees for storing the data
also. When you create an index-organized table on some databases (like
Oracle) you are telling the db engine to store the table as a B-tree.
The B-tree is a kind of balaced tree (like plain binary trees) optimized
for disk storage.


[]s, Fernando Lozano

Reply via email to