Parallel computing expertise needlessly scrapped by ANU Computer Science department

On May 18, 2010 I was severely brain-injured in a collision which occurred when I was cycling to my job as a Patient Support Officer (cleaner) at the Royal Brisbane and Women's Hospital. This was the tragic end to my absence from my real career choice and qualifications in computer science, forced on me because of misinformation given to me by my last employers in that field. At the time of the accident I was seriously attempting to return to my career, having learned, to my utter disgust, that the project from which I was sacked after two years, which I was told was a complete failure, had not failed at all. I may have forgotten a lot due to my brain injury, but I have not forgotten this. The injustice is seared in my memory and I am now publishing the documents that record the truth of my pre-injury predicament. They may also help others in this cut-throat industry. This will be the first in a number of contemplated articles on my experience of brain injury. Sheila Newman has also written a book about it which will probably be published late this year or the next.

How a computer scientist became a hospital cleaner and then a hospital patient

I was overqualified for the role of cleaner because I had an Honours Degree in Computer Science from the University of Southern Queensland and had worked for two years to implement the Distributed Java Virtual Machine in the Computer Science Faculty of the Australian National University (ANU) running parallel on more than 80 computers. (Much of the content of the ANU page that describes the details of this project is missing as of 10 January 2014. The error message states: "Fatal error: Call to undefined function: mysql_select_db() in /data0/web/data/feit/functions/connect.php3 on line 14"). At the end of two years on the ANU project, my supervisors claimed that my work on this was substandard and I was forced to accept that judgment at the time.

I have since learned that my work must have been of a very high standard, because, only 3 months after I finished at the ANU, a Canadian Masters student commenced to use the DJVM as a platform on which to build her Masters degree software project. Had there been the slightest bug, let alone any serious flaw in the DJVM, it would have been much harder, if not impossible for her to complete the project. However not knowing that, I saw no alternative but to seek low skilled work. That is why I ended up working as a cleaner. And that is how I came to suffer a severe brain injury.

Thanks to my RBWH supervisor, Jan Dring, for her kindness

Although my position as a cleaner at the Royal Brisbane and Women's hospital was not what I had studied to do, to be fair to the RB&WH, working conditions at the time were good. My supervisor, Janine Dring, now retired or retrenched, encouraged me to apply for clerical work. She was also the person who identified me when I was brought unconscious to Accident and Emergency after the collision. I have been told that she visited me on the ward and one evening, while I was still unable to form memories, she obtained permission and then took me down to spend time with my work colleagues. I spent over two months in the Neurosurgery ward in Brisbane and then a further six weeks in Mater Private, before coming down to Melbourne because my carer was advised that rehabilitation and services were better there than in Queensland.

Below is a letter that I sent to my former supervisor in 2008, when I realised how misled I had been about my abilities.

Complaint to Dr John Zigman of my treatment by ANU Computer Science staff 2002-2004

Subject: Complaint of my treatment by ANU Computer Science staff 2002-2004
Date: Mon, 8 Dec 2008 16:56:53 +1000
From: James Sinnamon
[Email sent to: Dr John Zigman, Dr Steven Blackburn, Peter Strazdins, Dr Ramesh Sankaranarayana.]

Dear Dr John Zigman,

This is just to let you know:

I have not worked in my chosen vocation of computer science since my involvement in the Distributed Java Virtual Machine Project (http://djvm.anu.edu.au) abruptly ended in February 2004.

When I last spoke to you, as I recall, in May 2004, and asked if you reconsider your decision to not offer me a PhD scholarship to work on the continuation of the DJVM project in Arc Project DP0449670, I at least gained the impression that you would do anything you would have been practicably able to to help me out of my dire professional circumstances.

I would never have dreamed that you would have failed to take the small amount of effort that would have been entailed in passing on to me knowledge that would have almost certainly have allowed me to confidently approach prospective IT employers for work.

That knowledge was that the DJVM I had helped to implement had been chosen by a Canadian student (http://webhome.csc.uvic.ca/~jbaldwin/) as a platform on which to build her Masters project (http://aosd.net/workshops/late/2006/later/submissions/baldwin.pdf) and that the DJVM had met all of her requirements.

I have been judged for not having more vigorously sought employment since February 2004, but I wonder if those people making those judgements could understand the effect of having the hard work of two years of my life deemed, implicitly or explicitly, as worthless, by every one of those I had worked closely with during that time.

I formed the distinct impression from you that the DJVM I had tested on 72 nodes, packaged and documented was to be shelved in favour of another DJVM to be built using BCEL and it was not until years later that I was to learn that my DJVM had, indeed, been used, after all, on ARC project DP0449670.

If, after working late into the night every night often for weeks on end facing heartbreaking bug after heartbreaking bug (never as far as I recollect having been introduced by myself), I could not produce a product of worth to anyone else, then what possible chance did I stand of being able to produce anything of value to any employer by working normal hours?

As a consequence, I have not been able, for the last four and a half years, to even bring myself to write my resume, let alone a job application. Every attempt I have made to do so has only driven me into despair and depression.

The simple knowledge that the DJVM I had helped to implement had been successfully put to use would have changed everything for me. How that could not have occurred to you is beyond me.

I can only come up with three conceivable reasons why you would have withheld that knowledge from me:

1. That my fate and well-being mattered so little to you that it never again entered your head again after I last spoke to you around May 2004
2. Malicious intent.

3. Acknowledgement of my contribution to the DJVM project would have somehow been an embarrassment to you and would have threatened your own reputation.

4. Some combination of the above.

Instead, as a result, I am working today as a cleaner without any realistic prospects of even being able to gain employment in a white collar occupation let alone in IT.

When I went to Canberra in early 2002 I had hoped, if nothing else, to get from my employment simple unambiguous achievement that I could subsequently use to impress prospective employers.

I badly needed a crowning achievement in my CV because, prior to that, my IT career had been very patchy, I would argue, largely for reasons beyond my control.

As it turned out my implementation was indeed exactly that, but neither you nor anyone else bothered to tell me and I was until April 2008 over four years later that I was to learn this.

Because I was not told this, it looked to me, and to any prospective employer, like shelfware which, for all anyone knew, may not have even worked at all.

As I had been denied the chance to actually use the application I had poured my heart and soul into for two years I had no way of knowing if it would have met the requirements of any other user and prospective employers would have only had my word to go on that it had been meticulously tested on up to 72 nodes using novel debugging techniques I had devised myself.

If I had not worked so hard to do that, on occasions against your own counsel, it is most likely that it would have not met the requirements for the Masters Project to implement Aspect Orientation.

Even if it was true that I was seriously wanting as a prospective PhD student as you, Dr Peter Strazdins, Dr Steven Blackburn, Dr Ramesh Sankaranarayana all implied, then, surely, at least I was entitled to some honest feedback well before October 2003, less than two months before my contract was originally due to expire?

In fact, I distinctly recall being told by you only a matter of days before the news was broken to me that I was not to be part of the DJVM project that you considered me an indispensable part of the DJVM team, but that all changed, apparently after the others told you they didn't think I was up to it.

I was subsequently told in January 2004 that only after that did you bother to think more deeply about my performance and come to arrive at the conclusion that I had, indeed, demonstrated my unsuitability months before that as a result of my own contribution to the ARC research grant application having been judged unsatisfactory and as a result of the way I had worked on the first version of the DJVM in 2002.

Whatever the validity or otherwise of those claimed concerns about me, to have not, at some more timely stage, made the effort to have sorted out in your own mind their implications for my own future with the DJVM project and make them known to me, I consider a gross failure in your duty of care to me as my supervisor.

From all the evidence I can see on the DJVM web site, little has been achieved in well over four years since I was unceremoniously dumped from the project.

In fact, given the seeming lack of progress, it seems to me highly probable that you, Dr Peter Strazdins, Dr Ramesh Sankaranarayana and Dr Steven Blackburn have been hasty and wrong in your judgement of me, made behind my back in 2003.

Perhaps now it may be acknowledged that the supposedly excessive time I spent was, indeed, what was necessary to solve the problems I was faced with.

Wherever the truth lies, I was entitled to much more timely feedback so that either I could have been given the opportunity to lift my game or, if that was not possible, to have begun to make realistic preparations for an alternative subsequent career.

In Dr Peter Strazdins' case, it has since become clear to me that he had decided well before the outcome of the ARC research grant applications were known that I was to be dumped from the project but had never made the effort to inform me. To the contrary, he even asked me to participate in the 24 hour pushbike race that was to be shortly held. I took this to be confirmation that he expected me to remain at the ANU. How anyone would imagine that someone, whose career prospects were as grim as he must have known mine to be, would have wanted to spend weeks training for such a physically gruelling event is beyond me.

On the day that the outcome of the ARC research grant applications were announced in October 2003, you burst into the office to break to me the good news that our application had been successful. That naturally led me to believe that I was to be included on the project and that you wished me to be included on the project.

I asked all present including, as I recall, Dr Steven Blackburn, Dr Peter Strazdins and yourself that a meeting be held ASAP so that we could discuss the future of the DJVM project. no-one else bothered to tell me that I was not to be included. In fact, almost a whole week went by before Dr Peter Strazdins bothered to break the news to me and, even then, it was only after I had gone into his office and had raised the issue.

These are only some of the reasons that I consider your overall treatment of me to be inexcusable. It is hard for me to conceive of a more shabby way for any professional to have been treated by former work colleagues, let alone by colleagues I had considered to be friends.

I intend to make a formal complaint of this through whatever channels are still available to me.

Whilst it is not clear if you can be held to account for what you did as you apparently no longer work for the ANU, I also hold Dr Peter Strazdins, Dr Ramesh Sankaranarayana and Dr Steven Blackburn culpable for what happened to me.

Regardless of what my prospects are and regardless of the eventual outcome, I intend to put this on the public record so that others may be warned not to place their trust in you.

Yours sincerely,
James Sinnamon [...]
Red Hill QLD 4059

Documents indicating DJVM project was successful after all

Below I reproduce two emails recording my correspondence with Jennifer Baldwin, the Canadian student who used the virtual java machine that I successfully produced.

Subject: Re: Distributed Java Virtual Machine and AJVM
Date: Fri, 25 Apr 2008
From: Jennifer Baldwin
To: james.sinnamon [...]

Hi James,

No worries about the delay in replying, takes me awhile as well. But
you never did call me 'Janet'

> So am I right to assume that you encountered no serious bugs in the dJVM?

That's right, we never did find any bugs that weren't introduced by my own
code.

> Do you mean http://djvm.anu.edu.au/download/DJVM-1.1.0.tar.gz

Yes, that's the one. I haven't really looked at it other than noticing
it no longer has patch files so I couldn't use it for the type of
research I was doing anyways.

Jennifer
-------------------------------------------------------

Subject: Distributed Java Virtual Machine and AJVM
Date: Fri, 11 Apr 2008
From: James Sinnamon
To: [Jennifer Baldwin]

Dear Jennifer Baldwin,

I worked as a research programmer on the dJVM for two years from 2002 until
March 2004.

I was gratified to learn that the DJVM had been put to use since my
involvement ended. I have just cursorily viewed your documents and found
the dJVM within http://webhome.cs.uvic.ca/~jbaldwin/AJVM.tar.gz

Can I ask how difficult or easy it was to get the dJVM built and then used?

Whilst I had tested it on up to 72 nodes with, as I recall, relatively trivial
Java applications it failed to run a considerably more complex multi-threaded
Java application I had written for an Honours Year project in 2001.

Interfaces were never dealt with as it seemed to me as if no further
modifications to the RVM were required, but when my applicaiton crashed, I
speculated as to whether this because modifications were necessary for
Interfaces. I wasn't able to pursue this.

Yours sincerely,

James Sinnamon

A cut-throat industry

Given the extremely high rates of unemployment in the Information Technology/Computer Science industry and the lack of industrial protection in that field in Australia, my story must echo the experiences of many computer professionals. Perhaps few will have had their misfortune amplified to the degree that mine was, but my awful experience still serves to highlight the downstream impact that casual arrogance in employers and supervisors in any field may ultimately produce. Surely in a world less characterised by the anonymity of globalisation and constant people movement, and so-called efficiency, it would be much harder for human beings to treat each other so badly, without actually declaring war.

Towards maintaining credentials in the computer industry, I built the political and environmental internet site candobetter.org (which later became candobetter.net). After my injury I was no longer able to maintain the site, but Sheila Newman, my major collaborator on candobetter.net, has gone out of her way to salvage and oversee its management. For me, this was tremendously important. Without that site I would possibly have little connection with my former activities in the social and political sphere.