Longest job20 October 2017
Around now marks the point that my current job has lasted longer than any previous job, and even though these days I am somewhat averse to writing about my professional career, this is one mile-stone I decided not to let pass. This article has been in the making a long time, most of it having been written 3-4 months ago, and unlike previous writings on the subject feels like being written from a position of strength rather then justification. It is also as a result also somewhat long-winded.
Lost interest in writingIn the past I wrote quite often about stuff related to my professional life, which even if not explicitly linked to my then-companies, were based at least partly on stuff I did during my work-day. Good example is my GTK for Win32 guide, and I suspect thr knowledge would have long lost the knowledge had I not turned it into a web article. These days however I've mostly lost interest in writing articles related to my work, which is down to several partly-overlapping reasons:
- My previous companies were all sufficiently below the radar that all I had to do was omit details that would cause it to pop up on a Google search by a client, as only an insider would know the connections. This allowed for quite risqué stories such as when it took several weeks to untangle a build system. With big well-known companies that attract all sorts of trouble-makers who have the resources & time to go data-mining, a different level of caution is required.
- All companies are dysfunctional in one way or another, and the incidents I see these days don't seem worth writing about. I suppose it was luck that my earlier companies, mostly being startup-sized, provided a steady stream of juicy stories. I also feel that non-company-specific stuff like Scrum and Mythical Man Month that nevertheless drew on my experiences working at certain organisations, are topics I've covered enough as-is.
- The projects I work on are mostly open-source, so any views I have on my current work that are appropriate for publishing are already in the public domain. What little new information I could add here is not worth the disclaimers and context explanation (both of which are worthless in practice anyway) I am probably supposed to add.
- These days I am much more concerned about the need to have a clear separation between professional and personal things than I used to be. This is partly due to the more onerous requirements related to conflicts of interest, which are as much about image than wrong-doing.
Previous rolesI'm always a little jumpy about talking publicly about previous companies as they inevitably dwell on the negative aspects of why I left them, but it is often only in hindsight that one can see what they got out of any given job, and it takes a few companies to work out what is personally important. At graduation the VC said in his speech that we would expect to have three careers in our lives so to have had three jobs since then is nothing extravagant, and it means I have a good idea what I would like as a career.
The dim-and-distant pastMy first “proper” job was working in a Microsoft-based start-up, and it was pretty much destined not to work out. First jobs tend to be somewhat rocky experiences, but this one had problems that would have made it challenging even for a seasoned professional; unstable & non-standards-compliant platforms, questionable company practices, and nothing even close to proper product planning. I didn't know it at the time, but the company had a track record of people walking out, and in hindsight I probably should have done the same — at least then I would have avoided my self esteem getting crushed.
Not even sure how I scraped up the motivation to recall much about a company I spent 3-4 months at over a decade ago and never really looked back at since, but even today the buzzwords on old CV scattered all over the place still attract emails from random recruitment agencies. In hindsight if this job was to lead anywhere, it probably would have been corporate DevOps.
InternshipDuring my Ph.D I spent some time working at Toshiba TRL, as they had part-sponsored my studies. At the time I treated it as the chance to chalk up some commercial experience, although in reality it was something of a side-project in Machine Learning that I tried to squeeze a publication out of. On the whole the time there went as well as could have been expected, and the technical side still has relevance to my career, so I still keep it as a line on my CV. I think I considered working at TRL after my Ph.D, but as it happened I got a job before getting round to applying, and I suspect that there would not have been any suitable open roles anyway. At the time the boundaries between TRL and the university department were somewhat blurred, with some saying that TRL is even more ‘academic’ than the former, and by this point I was not really interested in a career in academia.
Bristol post-docLooking back I have a generally favourable view of my first post-doc job, and until my current company is the one I spent the longest in. In hindsight much of the decision to leave was due to being fed up with Bristol, and general feelings that the place was too much a graveyard of memories that was stopping me looking forward. Bristol had been a vibrant city in the mid-2000s, but long story short the place was hit very hard by the credit crunch, and it would be 2014 before it showed any signs of recovering. The job itself I obtained when I really needed it and it turned out to be a direct hit in terms of career interest, the core part of the job being network video programming. After years of the glacial speed at which academia gets things done it was nice that people actually had proper interest in what I was doing, and I was being paid for it, which was a major contrast as to how academia treated me.
There's already plenty written about my other thoughts of this company, in particular my reasons for leaving; it was a living dangerously small neo-startup at the time the UK economy was a complete wreck, which is not a good combination for a proper work-life balance. In such circumstances it is more a case of you work or you life hangs in the balance, and eventually going through several work crunches in short order took its toll. Maybe an unguarded criticism for me “wasting time” coming out to a sub-contractor's site after having recently worked overtime was the turning point, but these things are always sparks to underlying situations.
Time in New ZealandWhile I did have some business interests out in New Zealand, the day-job proved illusive, and in the end I ran out of time. Maybe my commercial experience at the time was that bit too much on the marginal side, maybe I was in the wrong city, maybe I had the wrong connections, or maybe I was destined to fail regardless for other reasons. The one thing I am sure of it was not for lack of effort, but all this is beyond the scope of what I want to write about here. Coming back from New Zealand was partly admitting defeat, but it was also the opportunity for my life to start progressing again. Yes I built up a lot of fond memories, but ultimately my life had been on hold, as I could not switch away from being semi-transient and start to put down proper roots.
Python devopsHaving come back from New Zealand I got what turned out to be a Python DevOps job, the particulars I have once again already written about. It was destined to be a stop-gap, and although it provided good experience with tools (Gerrit, Jira) & work-flows (Agile et al) I did not previously have, in hindsight I doubt I ever had any real liking for the project itself. I finally decided to leave when I concluded that the core technical part of the job, namely the grubbier details of Puppet & MCollective, were of no interest for what I wanted as a career. I also came to the conclusion that how it did things was seriously flawed, with a major headache being the effort required to get code through the byzantine Continuous Integration system. For me a core metric of productivity is the proportion of time spent writing business logic (i.e. implementing new functionality), and I estimated it to be around 10-15% of my working week, whereas at my then-previous company it was typically 80-90%.
Good in terms of having proper development procedures, but compared to my previous job the lack of any real sense of ownership — writing a CV highlighting what I did was painful — meant having no real loyalty for either the company or the project. Pretty much put me off Agile, and looking back some of what I remember is probably inherent in any company that operates this way, with some such memories being eerily similar to what others have written about the negatives of Agile/Scrum. All the daily stand-ups meant no real down-time, and in hindsight I remember plenty of cases of both busy-work and overdoing of information requests. I personally preferred working on bugs, as at least they didn't require any of the time-estimate charades. Ironically shortly before I left I saw an internal email about a C-based project, but was told it was already gone even though I replied same day, which turn out to be the last chance the company had of holding onto me.