>>The conclusion is that mdot does not compensate for lack of common sense.

Compare to PHP that can execute text non-IL code like VFP, but uses a compulsory resultset prefix for database fields. Is that harder or easier than mdot? Either way it reduces possibility of error which is why some people insist on resultset prefixes (myresultset.myfield) in VFP as well. But IMHO VFP's unadorned variables or field names when a resultset is selected, along with browse and resultset munging from the command line, is highly attuned to human-readable data. It works fine in the hands of somebody who knows the rules, even if their way of skinning a fox is different from mine. ;-)

>>I thought that it was a particularly interesting topic for minifiers, obfuscators, and such things for which disambiguation is extremely important.

Agreed. Just in case, VFP Compiler has compiler directives for variables or even chunks of code that are best not obfuscated. IME almost all code can be obfuscated and it's easy to add a compiler directive for any variable you don't want messed with. In the early days we did that a lot but in 2017 the compiler is so accomplished that we expect most apps to compile without issue.