Back in 1990, I was pushing Assembler H to its limits and was invited to
submit requirements for what became HLASM, being planned by John Ehrman.
I was immediately impressed with his approach, which combined a strong
vision of future possibilities with a very practical understanding of
the real life requirements of compatibility, reliability and stability.

I still have a copy of the HLASM requirements and suggestions document
that John created and maintained over many years, including quite a
number of my own suggestions. Many of them were addressed long ago now,
but we still have a huge collection of good ideas to follow up if we
can find the resources!

John was also involved in other areas such as working with the binder
team on the GOFF (Generalised Object File Format) specification, which
made possible a lot of new function to support Program Objects.  (I
later heard from John that I had apparently influenced some of this work
myself when I visited the Santa Teresa lab in about 1988 and discussed
possible orthogonal extensions to linkage editor function with Dick Lee
of the binder team).

I also worked with John more recently (via email and online chat) from
2014 until he retired in 2016, when we were both helping out with
support of various old products such as VS Fortran (another of John's
areas of expertise). I always appreciated his kind and wise help, and
was happy to take on some additional support work to give him more time
to work on his Assembler book.

(I think that John had also hoped to create a similarly authoritative
volume on writing macros once he had finished the first book, but
although that never materialised, at least we have his excellent SHARE
presentations on the subject).

I'm sad that I never actually got to meet him in person (apart from
perhaps a brief encounter at a SEAS conference in Edinburgh in 1987, of
which neither of us had any clear memory), especially as both of us had
outside interests in theoretical physics and classical music.  John and
Tineke had hoped to visit the UK for a vacation shortly after he
retired, but his increasing need for medical treatment unfortunately
ruled that out at the time.

Many others will have encountered him at SHARE conferences or education
sessions, where he clarified the mysteries of HLASM and the binder
(among other topics) and helped many on their way to become better
assembler programmers.

As current HLASM team leader I am reminded every day of John's influence
on the product, although he did set the development team some tricky
challenges which we are still trying to get right to this day, such as
removing some awkward restrictions on built-in functions and avoiding
ambiguities between SETA and SETB expressions, which we were still
discussing by email less than a couple of months ago!

He made a big difference.  I will miss him.

Jonathan Scott
HLASM team, IBM Hursley, UK

Reply via email to