·      Rabin Ezra 1967-2005


When HPCC was set up, then called PPC-UK, Rabin Ezra was by far the youngest of our founder members – he was still at school. He died of pneumonia, in Malta, on 27 June 2005, a few weeks before his 38th birthday, and will be greatly missed by the club.

Rabin went to a conference in Malta in general good health, though feeling a little unwell the day before leaving. Once there, he became seriously ill, was unable to fly back to London, and died in hospital.

He had already been an enthusiast of computers and computing when he joined our club, and had been one of our most enthusiastic members ever since. Some members will remember his hardware and software hacking of HP-41 calculators, and later how he drilled open an HP28 for the club’s “Research and Development” project. His articles on HP-41 m-code in Volume 6 of Datafile are still a fascinating source of information for HP-41 hackers. Rabin himself described his feats in the article he wrote about HPCC’s early days for the book RCL 20. That article is reprinted in this issue of Datafile. As Rabin wrote there, his main HP calculator lately was the HP-16C. When his colleagues came to tidy his desk, there in the middle was his HP-16C, and two more were at home.

Rabin progressed from school, to university, then to a PhD in computer graphics. While working on his PhD at Queen Mary College in London, he also lectured and supervised students for a term. He then moved on, to an impressive career, in computer graphics with Criterion, part of the Canon group, and recently with Sony, as a senior engineer in the PlayStation 3 development team; he had been told just before leaving for the conference in Malta that he would be promoted on his return. His colleagues reported how he would listen quietly to complex discussions, then make unexpected and clever suggestions; much as he would do at HPCC meetings.

Whether at his parents’ home, at work, or at HPCC meetings, Rabin would be a source of calm and help. Queen Mary College, where he studied, intends to set up a scholarship fund in his memory, continuing his help to those who need it. Details will be given in a future Datafile, for those who would like to contribute to this fund in his memory.

HPCC members, especially those who come regularly to the monthly meetings in London, will feel his absence very much. We shall miss his serious advice and his amusing suggestions, at meetings, at club conferences, and at the club’s Christmas  dinners. We extend our most sincere condolences to Rabin’s Father and Mother and his relatives. He will be remembered with affection, and greatly missed.


Wlodek Mier-Jedrzejowicz


The early years

Rabin Ezra (19)

It is now difficult to remember that, at the beginning of hpcc’s life, the choice of programmable calculator or desktop machine was much harder to make than it is now. The issue of pcw which carried a cover story on the Sinclair zx80 also had a review of the hp-41c. To my 13-year-old eyes, they seemed in some sense similar in the possibilities they offered, with the latter having the advantage that I would be able to carry it around, and it would remember my programs when I turned it off. Today, the gulf in computational power between the laptop on which I type this, and the 48sx on the table, make their respective roles clear.

Somehow, I managed to convince my sceptical parents that they should buy one for me. As part of hp’s support activities, it produced a newsletter called hp Key Notes containing programs and news. In Vol. 6 No. 2, a letter from David Burch announced the formation of ppc-gb. (Its name changed to ppc-uk by the actual start of the club.) Having already discovered the wonders of “Synthetics” thanks to the green Wickes book, I was keen to see what else people had discovered and so joined.

The ?rst major event was the u.k. conference held at a remarkably grand location (or so it seemed to me, as a 15-year-old). Rather than using a college venue, the conference was organised at the Great Northern Hotel in Kings Cross in London. For me, it provided an opportunity to see hardware which had previously only been pictures in catalogues. The various peripherals were probably what made the 41 a computer system rather than a calculator, even if I couldn’t afford them at that time.

The conference also saw the introduction of the hp-75c. I remember being stunned by the slide show of one in bits, as it hadn’t occurred to me that someone might take such an expensive piece of equipment to pieces simply out of curiosity. It also saw a debate as to whether the 75c, perhaps the first ultra-light laptop computer, should be supported by the club. The consensus that eventually emerged was that we should support the hp aspects of the machine, but not print basic programs which might work on any machine.

Sadly, I missed out on perhaps the most important part of any conference; being rather younger than the other attendees I had not opted to stay at the hotel, but rather to commute. I therefore didn’t stay for the late-night informal sessions, which I only discovered had happened when Vol. 1, No. 3 dropped on the mat. With issue 3, the journal gained a card cover and photos of the sessions, as well as the more normal part of the conference. I even managed to sneak into a picture, though nobody knew who I was, so the list of people under the picture labelled me “?”.

The 41, and rpn machines in general, remained the main “focal” point of the club across the ?rst few years of its existence. Even with the 75 and hp-71b, there remained a feeling that the club’s constituency was predominantly rpn machines. As well as conventional ways of programming them, the club also provided a forum for those of us interested in m-code programming.

It was in connection with this that the club attempted its first hardware project. As I had some electronics skills, I became involved. To write machine code for the 41, a box capable of simulating a plug-in rom was required as the memory spaces used for user data and programs were physically distinct from that in which m-code executed. The 41’s processor had a serial bus, so interfacing more conventional components to it required a set of shift registers and some assorted decode logic. As an additional oddity, the words were 10 bits wide, so some packing logic was required if space in conventional 8 bit memory parts was not to be wasted. Eventually the club chose to simply provide a service for members by sourcing eramco units as a full design could not be justified.

The release of the hp-28c marked the close of the early years of hpcc. With its release, hp moved from expandable, interfaceable machines to much more targeted products. Though later products could take memory cards, they assumed that for i/o, the owner would have a pc to connect to. As well as being a closed box, the 28 marked a major change in the usage model; out went the type-limited four-level stack, and in came a completely polymorphic, infinitely deep one. The programming language also changed.

hpcc did pull a 28c apart to check on the feasibility of adding more memory, as it was based on the same processor as the 71b. One quiet evening at Imperial College, I took a drill to the heat sealed case to open it. The machine appeared, reassembled, on the following journal cover with the keys rearranged and still functioning, though lacking a little in the solidity “stakes”.

(These days, the hp I use most is a 16c, without which I would be lost.)