NOTES.txt

Where we are:

a) modify schema so they conform to DID-database and use base instead of ndi_document
b) modify ndi.document so it operates with base instead of ndi_document
c) test the code to make sure all test cases pass


What is next:

a) make a new version of the dumbjsondb implementation that
  1) stores the binary files somewhere else and handles the file copying on loading
  2) allows the name of the binary file to be specified
b) modify use of writing to binary files so the binary files are made before
   the documents are added to the database
c) test the code to make sure all test cases pass

Then, finally:

a) modify ndi.database implementations so it uses did did sql
b) test the code to make sure all test cases pass

Some notes:

a)

ndi.app.spikeextractor: 
	line 186 *
	line 188 *
	check lines 437, 458  * (reading)

ndi.app.spikesorter: 
	line 231 *

c) update ndi.document so merges file names



To update other stuff:



grep "ndi_document\." */*/*.m

also grep "\'ndi_document_"