Re: ITP: libavl -- AVL tree manipulation library

[Ben Pfaff <pfaffben@msu.edu>]

Wessel Dankers <wsl@fruit.eu.org> writes:

> To summarize the differences (as I understand them):
> 
> The GNU avl library:
> 
> - is reasonably well documented;

True for 1.4.1.  (libavl 2.0 is already documented better than
almost any other program you will ever find, and this will only
be improving.)

> - has four (slightly) different algorithms, including one
>   that isn't really AVL;

libavl 2.0 will have at least 9 core algorithms, only 3 of which
are AVL.  I'm probably going to rename it to something more like
"libsearch" or "libtable", but haven't come up with a name I like
yet.

> - only supports an ordered set interface;
> - has a small node size;
> - is very fast.
> 
> My little avl-tree library:
> 
> - is not very verbosely documented (most of it is in avl.h);
> - contains only one, albeit slightly augmented, (AVL) algorithm;
> - has an interface which allows one to use the tree for a variety
>   of purposes;
> - uses a somewhat larger node-struct;
> - is very small code-wise.
> 
> Since the most visible difference seems to be interface-related I guess
> making it a seperate package could be considered justifiable. But that
> is really a matter for Debian maintainers, of course.
> 
> If renaming is necessary I would suggest `avltree', since that resembles
> the name of the original library.

I am happy to see your library packaged under any name you like,
since I'm planning to change GNU libavl's name eventually anyhow.
I'd rather that GNU libavl not be packaged at all, at least at
this time.

Thanks,

Ben.
-- 
"Welcome to the Slippery Slope. Here is your handbasket.
 Say, can you work 70 hours this week?"
--Ron Mansolino