In the last thrilling (?) installment we left you with the cliffhanger that "there was still one more skirmish to be fought in the battle ..." and indeed there was.
As I noted in part 3, many of the external review team members thought that we had "sold out" when we made the switch to the hybrid design. We knew they weren't happy, but we didn't realize just quite how unhappy they were. I honestly can't remember the full chronology, and much of the correspondence was in the form of private letters to both Toronto and Rochester management. I had hoped to find the details of the more public aspects of the affair on the Web, but in this instance the wayback machine has let me down. So I'll just have to use my own memories of the events.
Some members of the review group, backed by a number of other notables from the News3X/400 magazine and other publications, published an open letter to IBM in which they criticized our decision and pointed out that even IBM's own compiler people felt that free-form was the right way to go. Omitted from the letter of course was any mention of the basis for our decision. Our discussions with hundreds, if not thousands, of the wider RPG public left us with the very real fear that if we did so, nobody would use the new compiler. The letter urged IBM to reconsider the decision. Of course the brown stuff hit the fan and we were back into management review hell.
Then we had a spot of luck. The leader of the anti-hybrid group started publishing the results of a survey that he and others had been taking at user group meetings and similar events. The results of that survey appeared to overwhelmingly back his claim that if presented with the "facts" RPGers would choose free-form. It might seem that this was a bad thing for our position, but in his desire to get the answers he was looking for, the gentleman in question had produced such a heavily biased survey that it was really hard to take it seriously. The "correct" answers were all but spelled out for you. By the way, if you are not familiar with how surveys can be biased by the way questions are asked, or indeed been by the sequence in which they are asked, there is an excellent piece by Pew Research.
Our management took a look at the survey and decided that the questions were too biased for the data gathered to have any degree of reliability. After thoroughly grilling us on the foundations for our choices, and discussing the matter with a number of customers, ISVs, etc. they decided that we should go ahead as planned. We breathed a sigh of relief and went back to work preparing for the birth of our new baby.
The following months were crazy as we ramped up for the GA of V3R1 in November 1994. The other members of the team and I were frequently on the road, presenting to user groups, ISVs, business partners and many others all over the world. It was a crazy and exhilarating time and it very quickly became obvious that we had made the right choice and that the vast majority of RPGers were thrilled by the new language. But we had built a new compiler so that we could move the language forward, so while introducing the new features we were also querying RPGers as to what they wanted to see next. The list was huge!
So what was the No. 1 request? It might surprise you to hear that it was parameters for subroutines. This wasn't news to us as it had been on our wishlist for the initial release but we were surprised at the level of support. But, we hear you say--you never added that support, how come? For the answer to that question you will have to wait for the next installment of this saga.
Note: Comments are moderated and will not appear until approved