Sunday, 26 December 2021

A First Visit to Nice

In January 1980 I attended my first ever standards meeting. Many more were to follow. This was in the very early days of the OSI (Open Systems Interconnection) effort, and a group called ECMA was working on the definition of the transport layer protocol for it. My day job at DEC (Digital Equipment) was leading the team building the X.25 network protocol for the Vax, so there was a connection, but I’m not sure why I was asked to go. It seemed interesting, and I was happy enough to go along.


The group was based in Geneva, but their tradition was that alternate meetings were held there, and at locations hosted by the member companies. In this case IBM, whose primary centre for network software was nearby at La Gaude in the hills above the valley of the River Var, was the host.


I’d already travelled quite a bit for DEC, mostly to the Boston area which was home to the company and nearly all of engineering. I’d also been a few times to see the group in Annecy, near Geneva, where some of the network hardware was designed. I’d never been to Nice. Like most Brits, I knew its reputation as a super-chic destination for the wealthy, with their yachts and private jets, so I was happy enough to see it for myself.


I don’t remember how I flew there. Back then there was just one direct flight per day, so quite likely I had a connection in Paris. Things have changed since then: prior to Covid, I counted over a dozen direct flights every day between Nice and all of London’s airports. I took a taxi along the Promenade des Anglais to the meeting location, the Hotel La Perouse on the waterfront and about as central as possible.


It’s called the Promenade des Anglais because it was the British who turned Nice into a holiday destination. The city is very old, going back to Roman times and with a very complex history. Until 1860 it was part of the Duchy of Savoy. The official language was Italian. Italy, France and the Duchy had been fighting over territory in the region for centuries, but in 1860 France played a big role in the unification of Italy against the wishes of the Austro-Hungarian empire. Out of gratitude, the new country of Italy handed over Savoy. The southern part became the French department of Alpes-Maritimes, with Nice as its main city.


The British started coming to Nice, as a winter resort to get away from endless grey skies and rain, in the mid-19th century. They sponsored the construction of the original promenade - until then the shore was just a shingle beach. Even in 1900, Nice was a winter destination. Hotels and restaurants closed in the summer, because it was too hot.


At street level, the hotel had just a reception area and our meeting room. There was a lift to the first floor, which gave access to a terrace and to a second lift which served the rooms. I was a bit puzzled by this arrangement, though it all became clear later. The terrace was delightful, surrounded by mature lemon trees. Every morning we ate a typical light “continental” breakfast there, delicious bread and croissants with coffee.


The view from my room was just staggering. It covered the whole length of the Promenade, right out to the airport 5 km away, and the old centre of Nice, with the hills and mountains behind and the bright blue Mediterranean to one side. It really seemed like I had gone to heaven.


The meeting started after lunch. The group was called Technical Committee 24, and was responsible for standards for the OSI lower layers. Another group, TC23, dealt with the upper layers, but they met separately and I never encountered them. The chairman was an old-school Englishman, actually from Gibraltar, called Clive something, who worked for one of the non-IBM mainframe companies, maybe Burroughs. The group mostly split in two, and the transport and network layer working group was run by an Italian from Olivetti called Guerrino da Luca. Later he became the CEO of Logitech, the company that makes keyboards, mice and other peripheral stuff for computers, and it looks as though he is still Chairman of the Board there.


At this meeting I was a bit lost. I didn’t really know why I was there, so I just sat and listened. Work was just finishing up on the first version of the ECMA transport protocol, which was soon published as ECMA Standard 72. The work had been led by Bull, back then the primary computer vendor in France. It was already way too complicated, with multiple classes depending on which features you wanted to use. Later it turned into a complete dogs’ breakfast as it was “aligned” with similar - but different - work under way at CCITT (now ITU-T), the phone companies’ standards body. Later still, it became my job to make it even more complicated by adding support for Internet-style networks in what became Class 4 (TP4 for short). But that came afterwards.


When ISO later published the official international standard for the transport layer, as ISO 8073, it was identical to the ECMA work. In contrast, the network layer activity at ECMA led absolutely nowhere. It was led by two characters who loved theoretical argument and seemed to know nothing at all about actual protocols or their design. The noisiest of them was from ICL, Dave Ackermann. He delighted in pointless discussions about abstruse theory, though I suspect he rarely knew what he was talking about. His partner in crime was a genuinely brilliant Dutch guy, Oscar Rikkert de Koe. (His last name means “Richard the Cow”, a hangover from when Napoleon forced the Dutch to adopt consistent family names. They thought it was a big joke and made up silly names, little suspecting they would be stuck with them two centuries later). They have both passed on now, so there is no risk of offending them.


As usual with standards groups, we would generally go out for dinner together in the evening, followed a beer or three in the hotel bar. Ackermann loved this. He would always start technical arguments - friendly ones - by stating something obviously counter-factual and waiting for a reaction. Then he would distort the discussion so others were always somehow on the defensive. For my first few meetings I played into his hands and found this annoying. Finally, I realized that the way to disarm him completely was to agree with him. Whatever he said, no matter how outrageous or silly, I would just say “You know, Dave, you’re probably right.” He just sat there, utterly deflated, trying to find some new approach that I would disarm in the same way. It was just too easy.


A highlight of all the hosted meetings, away from Geneva, was a dinner organized by the host. The lead delegate from IBM explained to us that we were going to a restaurant famous for its version of a famous local dish, and we should be sure to appreciate it since we would never get anything as good. A bus took us there, a few kilometers down the coast. The meal started with fish soup. It was delicious, but we were all conscious that we needed to leave room for this famous local delicacy, and ate sparingly. Everyone was surprised when the next dish turned out to be dessert. Of course the soup was the famous local dish, bouillabaisse. We all felt a bit silly, not to mention hungry.


Looking back, I wonder why most of the participants were there. There was no technical discussion at all, and very little of any other kind. When later I introduced TP4, nobody ever questioned anything I wrote. Yet the same regulars turned up at every meeting. One I remember well was Peter von Studnitz, who worked for Phillips. He was very distinguished, befitting the “von” in his name, though he explained to me that this was a reference to the family estate now lost to East Germany. Still, he wore a magnificent cloak to our meetings, instead of a coat, and walked very smartly just like a feudal baron. During the meetings he smoked an especially foul pipe, that filled the room with blue smoke. Back then smoking at work was considered perfectly normal, it was non-smokers like myself who were considered odd.


Another participant sticks in my mind because he was a genuine kleptomaniac. One evening a big group of us went for dinner at a typical local restaurant, drinking numerous bottles of their house wine. Their method of making the bill was very simple: they simply counted the empty bottles on the table. Our kleptomaniac colleague had hidden several of them under the table, so we didn’t pay for them. There was absolutely nothing to gain from this - we were all on an expense-paid business trip. At the end of the meeting, he took as many lemons as he could from the trees on the hote terrace and filled his suitcase with them, for the simple pleasure of stealing. He never showed up to another meeting - I think his company must have been embarrassed by him.


Editing the documents was all done by hand. Even very primitive laptop computers were still a decade away. I remember da Luca writing new text in all the available blank space, ending with a circular tour round the margins. Once I realized how it was done I equipped myself with an “editing kit” in a little pencil case. It contained pens and pencils, an eraser, scissors, white-out, sticky tape, glue and a tiny stapler. TP4 was all put together - literally - with this kit.


The hotel is built on the Colline du Chateau, the original centre of Nice which dominates both the old town on one side and the harbour on the other. One lunchtime I went for a walk and discovered the paths leading to the ruins of the chateau, with wonderful views in every direction. Now that we live in Nice and are slowly discovering its interesting bits, we repeated this walk. The hotel’s name comes from the famous French explorer, Jean-Fran├žois Comte de la Perouse. In the late 18th century he explored much of the Pacific and various things out there are named after him, for example the La Perouse Straits separating Sakhalin from Hokkaido. There is absolutely no connection between him and Nice though.


The meeting was held in a good-sized room at street level, behind the hotel reception. That meant taking the lift down from the terrace area - the “first floor”. One morning I got tired of waiting for it, and decided to walk down the adjacent staircase. It was far longer than I expected, going down and down and down. That was when I understood the layout of the hotel. The “first floor” is actually at the level of the roofs of the adjacent buildings, roughly at fourth floor level. The terrace is about 100 feet above the streets and the rooms even higher - which explains the spectacular view from my balcony. In the picture, you can see the yellow ochre hotel building snuggled up to the tower, its roof level with the top of it. The entrance is just to the right of centre.


Later that year da Luca moved on to greater things within Olivetti, leaving the position of transport group chair open. To my amazement Clive, the overall chairman, asked me to take on the role. It made sense, since I was by then the only one doing any actual technical work in the group. That meant that I had to start representing TC24 at the international meetings held by ISO - which is another story.


Today, in 2021, we went for a walk around this part of Nice. On impulse I went into the hotel reception, and we took the lift up to the “first floor” and enjoyed a coffee on the terrace. It was exactly as I remembered it, still with the lemon trees, and extremely pleasant. The receptionist flatly denied that there had ever been a meeting room at street level, although I know there was. But, soberingly, she hadn’t been born when we had our meeting there in 1980, and nor had any of the other staff we encountered.