I first worked with DJS on graphics for the original ARM chipset in Acorn AR&D, and received some of the best written engineering documentation - in the form of what I would later treasure as Seal-o-Grams (all ascii text, no fancy pictures).

Later, when I returned to join the team at Advanced RISC Machines in 1991, I discovered that about 1/6th of the company was a Dave and  I enjoyed coaching David in designing hardware (in ASim+SIDLE thanks LDS) in exchange for numerical and mathematical support. As DJS grew to take on the ARM Architect role I recall the rigour he put into reviewing early drafts of the burgeoning Architecture Reference Manual - and especially one detailed markup of the near-camera ready version where on each and every page he had ringed a "spot" that the photocopier must have added!

I did have the challenge of managing David for a few years, later on - fantastic on technical quality and detail, more challenging on the timescales and planning - but then with the extra hours David so often worked he made such significant contributions. David's patenting record was legendary in the early decades and the instruction architecture focus was really valuable to Arm.

The last major technical work we did together was jointly contribute to the nightmare known as "Unaligned and Mixed-Endian Access" that we had to face up to as a company in version-6 of the ARM Architecture. The Seal-o-Grams could really have benefitted from being written in red or green fonts where the reader could then chose one pair of filtered coloured glasses and only see the Little-Endian (or the Big-Endian) bigot's view they cared about - rather than the monochrome tables of byte swizzles and sign-extension complexities we needed as foundations... [And the end result was a spec that was a compromise that everybody was equally unhappy with, but reflected perceived balance for the LE and BE partners Arm had at the time!]

I cannot believe it was 11 years ago that I was asked to take on succession planning for David when he announced his desire to take early retirement and get back to the maths research that I know he had put on hold back in Acorn. I had always assumed I would get the chance to meet up and catch up again - and now find I should have made time earlier. A great work colleague who you could trust to do things properly - in his own timescales.