Apologies to the 720 team, I got a little ahead of myself releasing MIXIT
before its time.

While developing LLF v2, I decided that I needed to replace the .LDA
file format with one that is more effecient (spelled: f-a-s-t-e-r). Enter
the new and improved .VLDA format. Indeed, the output phase of LLF runs
a little better than twice as fast as it does while generating a .HEX
file and the .VLDA file size itself is about 1/3 the size of the .HEX
file. The problem with this is that, as the 720 team so rudely found out,
MIXIT, DIO, IMI, DOWNC and DOWND had to be fixed to allow the .VLDA as an
input and to use ".VLDA" as a default input and output file type.

Anyway, now its done. If you don't explicitly specify a file type on the input
or output of any of the utilities, a default file type of .VLDA will be assumed.
However, on input, all the utilities will change the default input to .LDA if
the .VLDA file cannot be found. This should prevent the problem that the 720
team had and should also keep any existing command files from "breaking".

In case anyone's intrested, the .VLDA format will also encompass the .OBJ format
generated by the new and improved MAC65 (whenever I get that done) and will be
able to be re-input to LLF for additional linking if necessary. The object
output of the BLISS compiler is the same format as the .VLDA so it can be input
directly to LLF without doing the assembly step. Anyone doing system II programs
want to sign up as beta test site? It appears that LLF is going to be neither
faster nor slower than LINKM, but it will afford a lot more flexibility in
dealing with overlays as well as removing the restriction of 6 char filenames
and stuff.

ds