Re: Standard source file names?

On Wed, Apr 1, 2015 at 5:04 PM, Buck Calabro <kc2hiz@xxxxxxxxx> wrote:

On 4/1/2015 4:05 PM, John Yeung wrote:

IBM has the power to make this experiment happen.
They could even do it without losing any money (by increasing the
price for the compilers to compensate).

That experiment was already performed and the results were abysmal
adoption rates.

I know that. I have already said (previously, not in this thread)
that today is a very different environment.

There have been previous attempts to build electric cars. By most
commercial standards, these attempts were abysmal. I am glad Tesla
did not go by your logic.

I don't have the exact time frames to hand, but there
was a multi-year period when IBM bundled all the compilers and all the
development tooling into one price. If you bought a compiler you got
all the other compilers, SEU, PDM and WDSC all together.

One thing I still am not sure I ever found a concrete answer to: Back
then, did they actually LOSE MONEY compared to what they would have
gotten had WDSC been a separate-cost item?

If the answer is yes, then, well, I guess I really can't make much
further argument about the WDSC bundling experiment. If the answer is
"no" or "there's no way to know" then for me, the question is:

OK, so then what's wrong with a low adoption rate?

If you (or IBM or whoever) cares about adoption rate, and you're NOT
losing money, then why NOT just put it out there?

My recollection is that the complaint back in the days of free (bundled)
WDSC was that the product wasn't good enough to use. I will say this: I
myself used WDSC to work on many hundreds of thousands of lines of code,
and I vastly preferred even Code/400 to SEU.

And there were some people that swore by GM's EV1. (I believe there
actually still are.)

Languages with the features of Java were available since the 1960s.
Objects, garbage collection, and even stuff that's more advanced than
anything Java has today. (The big one is Lisp's macros.) Sit and
absorb this for a second. Languages MORE POWEFUL than Java. In the
1960s.

But guess what. The environment in the 1960s wasn't right. I used
Lisp in the early 1990s, and the environment still wasn't really right
(at least, the Lisp I had back then was dog slow). And yes, some
people did prefer using Lisp (or other "slow" languages), even back in
the '90s, or the '60s.

There are so many reasons I believe the bundling experiment would go
much better today.

John Y.