Title: The Early Days of C++ in UK C User Groups

Author: Bob Schmidt

Date: 06 July 2019 01:03:07 +01:00 or Sat, 06 July 2019 01:03:07 +01:00

Summary: Francis Glassborow looks back on the formation of the ACCU.

Body: 

From about 1990, I had started to include material on C++ in C Vu. I was also sneaking the ‘++’ into the consciousness of the membership. For a time I slipped a ‘++’ inside the ‘C’ of C Vu (an ancient battle, long lost but there was meant to be a space between the ‘C’ and ‘Vu’).

I had excellent contacts with most of the main producers of compilers and in particular with Borland. In autumn 1992, my Borland contacts alerted me to a proposal from a group of enthusiasts to create a Borland C++ User Group. That seemed a little silly to me so I contacted them and suggested that they might like to be a sub-group of CUG(UK) as it then was. After quite a bit of negotiation, they decided to give the idea a try.

Unlike the founders of CUG(UK) whose newsletter C Vu was originally almost entirely reprints from the US, the Borland C++ enthusiasts had spent many months putting together the first issue of a magazine that they proposed to call Overload. One problem was that they had put all their planning into the first issue and had not given much thought to subsequent issues.

We launched Overload in April 1993 and the editorial control was firmly in the hands of the nascent Borland C++ User group. It soon became clear that the limitation to a single tool provider was not what people wanted. They wanted a robust publication on C++ even if the Borland compiler was the one they were using. At that time, the Borland Compiler was ahead of the Microsoft one in that it tried to track the developments coming out of WG21. There were quite a few other contenders. Most have withered away.

As the first year progressed, the small group of enthusiasts began to lose interest in the actual production of a magazine and it became increasingly reliant on the more general membership, at least those who wanted to read about C++ and were happy to pay a higher membership fee to include that publication.

Coincidentally another C++ user group was testing the waters in 1992–93. That was the European C++ User Group. Their main motivation was to run economically priced conferences for C++ users. I submitted a speaking proposal for their second conference which was scheduled to precede the WG21 meeting that was in July 1993 in Munich. I have forgotten exactly what my talk was on and it is probably best forgotten. The content was pretty naïve and it was my first venture as a conference speaker. However, attending the event was life changing. I was well entertained and the evening spent with Frank Buschmann (our host) Bjarne Stroustrup and Alan Bellingham in the English Beer Garden was memorable not least for Bjarne insisting that he bought a round after we had all imbibed 3 Steins of excellent German beer. We did eventually manage to get back to the hotel thanks to the excellent Munich public transport system.

The next day, three of us (I think Alan was planning to return to England – by car – and so missed out) together with Josée Lajoie (a wonderful Canadian academic and, in those days, chair of one of the three WG21 core work groups) went off to the Kloster Andechs Biergarten located within the Andechs monastery about 40km south west of central Munich. It was a lovely summer day and we had a relaxed and enjoyable time there. Josée was driving but we only drank in moderation. It was there that Bjarne, on learning that I was reviewing a 1500-page book on Borland C++ opined that you could not write a book on C++ over a thousand pages. The type face was larger than that in his book so maybe it was not an entirely fair comparison. I reminded him of that claim when the 4th edition of The C++ Programming Language substantially exceeded 1000 pages. As C++ has grown so has the page count for that book.

Sadly, EC++UG struggled and the Munich conference was its last. I offered to take over its membership, allowing them a certain number of free issues of C Vu and, I think, Overload. The membership records were chaotic so I am not surprised that it struggled. However, the result of injecting a nucleus of a hundred-plus hardened C++ enthusiasts/professionals was that Overload now had a solid readership.

In absorbing EC++UG, it had always been my intent to honour the EC++UG conference plan. It was to take a while to achieve but when WG21 was due met in London (well, it is listed as Cambridge but I am sure it was actually in London though hosted by Cambridge Analytica) in July 1997, I arranged that we would have a conference in Oxford. The event was held in Oxford Town Hall on July 18/19. July 18 (Friday) was a multi-track event. I do not recall much of the content. Saturday was a single track with Bjarne Stroustrup, Tom Plum, Bill Plauger (I think) and Dan Saks giving an excellent set of talks well pitched at the audience.

One very important decision right from the start was that ACCU (as it had now become) would provide the speakers but the administration/organisation had to be done by professionals who would keep the money. We went through several organisers over the early years. There was always a look of disbelief on the faces of prospective organisers when we claimed we could provide top speakers from around the world.

Some would be delighted to do the organising and charge us for it. That very definitely was not what we wanted. We did not want conference moneys flowing through our books with the result that we would become liable for VAT on everything we did. Volunteer treasurers do not want the complexity and legal liability that would entail.

The Borland C++ User Group and the European C++ User Group are distant ghosts in our past. Nonetheless they live on in ACCU and the ACCU Conference as we have successfully absorbed them and, I hope, fulfilled what each set out to do.

Francis Glassborow Since retiring from teaching, Francis has edited C Vu, founded the ACCU conference and represented BSI at the C and C++ ISO committees. He is the author of two books: You Can Do It! and You Can Program in C++.

Notes: 

More fields may be available via dynamicdata ..