[elephant-devel] Patch to 0.9.0

[Ian Eslick]

There was a disconnect between the case-insensitive string comparison  
inside BDB and that assumed in the lisp code (i.e. two lisp strings  
were not equal when they were equal in BDB causing the map function  
to misbehave).  I believe I've fixed this.

Are there any assumptions about case sensitivity in the Postmodern  
backend?  I know that the SQL backend relies entirely on the lisp  
comparison function.

I'll add some tests for this case and commit the patch to the 0.9.0  
CVS tree as well as my local 0.9.1 darcs branch for near-term  
integration into the main darcs repository.

Cheers,
Ian

PS - I'm not making my local 0.9.1 branch available not because I  
can't but because it is not in a coherent state.