0.8.9 Strange behaviour of snapshot filenaming

I did some experiments to understand the snapshot filename algorithm because I got a load of "ERROR 18" snapshot failures.

I used a DivMMC Future with a rubber key 48K Spectrum.

My conclusions are:

1. Upon power-up, the initial snapshot filename is "SNAP0000.SNA"
2. Pressing the "S" key, attempts to write a snapshot filename, if the file exists then "ERROR 18" is shown. The snapshot filename serial number is incremented.

Side-effects of this algorithm:

1. If the current working directory is changed then saving the snapshot file does not track the pre-existing snapshot files in that directory. It is OK, but the time relationship is lost meaning that the snapshot filename with the highest snapshot serial number may not be the newest snapshot file. For example, if a directory has gaps in the serial number sequence, then pressing "S" can create a snapshot file within the gap with a lower serial number than the highest serial number snapshot filename.

2. If the current working directory contains a large number of snapshot filenames with no serial number gaps (contiguous), then "S" has to pressed many times with "ERROR 18" until the internal snapshot filename is 1 higher than the highest pre-existing snapshot filename which then allows successful saving of the snapshot file.

Improvements:

1. Perhaps "ERROR 18" can be more verbose and indicate the failed snapshot filename ?

2. Perhaps the current working directory can be scanned to identify the highest snapshot serial number and add 1 to the number and use that as the snapshot filename ?

3. Allow snapshot files to be deleted using the NMI file browser.

4. Allow snapshot filenames to be renamed using the NMI file browser.

Is the source code available ? eg. is it closed or open source ?

Thanks.