I think that Brad and crew are off down a well beaten path that leads
to a swamp, in which people have been stuck for years.
The assumption that they make, also made in the ADA effort, was that
program language features per se have a big impact on usability and
productivity. In fact, as the ADA failure showed, other factors, such
as user community size, ease of porting and/or vendor support, have a
much bigger impact.
This leads me to the follwing: Counter claim: The most widely used
programming language is de facto the most useable!
Because...
It is the language for which it is easiest to find reuseable materials
such as programs to copy and libraries to use. This reduces programming
time far more than any language features.
It is the language for which it is easiest to find advice and get help.
This reduces the programmer's overhead in accomplishing the task.
It is the language for which tools are likely to be robust and widely
available. In turn, this reduces programmer effort in coping with
inadequate tools.
Historically, coming up with new programming languages and constructs
has had little impact on productivity. The current Java backslide,
after the initial gigantic claims, is yet another example.
Ruven Brooks
