2020 In Review

It’s pretty obvious what to begin with: We’re in the middle of a pandemic. You might have heard of it by now, and it’s got no real end in sight even though vaccines were approved in record time and are starting to roll out over most of next year. Then a bunch of other stuff happened, and it was all largely unimportant in comparison.

It was a year where hardly anything turned out as expected (best laid plans and all that), and given the nearly year-long confinement and isolation there is a natural tendency to focus on the stuff I missed rather than go on about what I did–but I like to follow the same pattern I did last year and the year before that, so here goes:

I have to say this was likely the best working year I’ve ever had at Microsoft:

• most senior team
• by far the best line managers I ever had
• highest (internal) visibility
• largest and most challenging customer projects
• and, with Microsoft’s renewed focus on the telco industry, also by far the best use of my skills over the past five years.

But, alas, it was simultaneously one of the worst.

Leaving the Portuguese subsidiary (and the relatively small market, with all that comes with it) was the best decision I ever made, if only because I am again fully immersed in the kind of international/culturally diverse work culture that I missed since leaving Vodafone.

And this might just be bias due to my current assignments, but after five years I can finally say I really like working at Microsoft (with a few caveats regarding tech I won’t go into).

I am (sort of) back in the telco industry (i.e., currently working almost exclusively with telco clients, although there is a fair bit of variety and I’m sure to end up in the financial sector as well), taking on the sort of groundbreaking mega-projects that were the sort of thing I loved doing at Vodafone, and bouncing between technology, strategy and business.

I can’t really begin to explain the affordances in terms of scope and perspective this kind of work brings, but it hasn’t been all roses.

Part of the worst was obviously the pandemic-driven shift to 12-hour days of 30-minute meetings, and I am completely fed up with not being able to have a stable, regular, family-friendly schedule.

But that was not all of it. I am now past my third re-org in this calendar year, a lot of work was grueling, uneven and extremely depressing (E_TOO_MUCH_EXCEL), and I spent most of the year juggling too many projects, teams, business issues, high-level outcomes, and, in general, far too many different contexts to feel like I was making significant headway into all of it at any given moment in time.

It’s the firehose that keeps on giving, and I miss focusing on one thing during a single day.

For starters, I learned that I can keep working (albeit at great personal cost) amidst the whole mess, and that I can bring clarity to some pretty gnarly business discussions by drawing upon years of experience, a renewed amount of soft skills and… a very large, very far reaching (but metaphorical) stick comprised of a good understanding of both technological status quo and hands-on experience on all sorts of stuff.

Core cloud concepts like Infrastructure automation, application modernization, and deep analytics stuff are still largely outside the realm of traditional IT in many places, and many companies are eager to catch up. So looking back, last year’s notes on both Kubernetes and machine learning were a bit more than prophetic, to say the least.

I also gathered a lot of evidence that multi-tasking is fine and good, but completely overrated–when taken to the extreme, context-switching will grind you down.

The upshot of this all has been that I am currently torn between doing some pretty unique things and having great impact where I am at now, and wanting a smaller, nicer context that would allow me to get to know people better, do proper technical deep dives into actual engineering problems (IT architectures are boring, no matter how much modernization you throw at them) and build stuff.

Above all, I’m still looking for a mission of sorts. I need to do work that I know matters instead of wading through paperwork and various procedural hoops that completely take away my creative focus.

So even while the business strategist in me is quite enjoying the ride, the engineer in me is slowly dying and feeling inexorably drawn back into simpler, “smaller” (in scope if not in scale) things that don’t require me to forego family life, technical depth, and my general well-being.

I’m now at the point where I need to code just to stay sane and reassure myself I’m not turning into a PowerPoint jockey, and that sometimes means spending even more hours in front of a computer.

I like to do the weird, totally out there stuff that does not fit into any pattern anyone’s done before, and traditional IT is all about managing risks (which I seem to be good enough at, given my continued survival) and following tried-and-true approaches (which bore me to tears).

But if there’s something I’ve learned over my nearly 30 years of work (if you include freelancing and college startups since the early 90s), is that personal satisfaction and innovation has never come from playing it safe.

My new office has been my “natural habitat” for a year now, and the timing couldn’t have been better.

Standing has been good for my health and motivation (so that is definitely staying), but I need to make things physically tidier and make more room for other pursuits.

I haven’t upgraded my main setup yet (still using my battered old iMac as a glorified Remote Desktop console, with an external monitor to each side) and the new Surface Book 3 has been more than good enough for work, but I’m planning to do some changes next year based on what I’ve learned over the past few months.

I got myself a sizable chunk of music hardware I’ve written about (like the Reface DX, which I love), and there is more music-related stuff I haven’t published yet due to lack of time, and having that gear set up permanently is something I’m keen on doing.

But if I had to highlight one thing I got this year that made a significant difference, that would have to be the Oculus Quest 2.

Yes, VR has always been gimmicky. And yes, every time I use it I am wary of it being tied to my Facebook account, and avoid using any of its social features.

So I’ve hedged my bets on its ecosystem (and adjusted my expectations) precisely because of that, but there is a lot more to this than just playing Beat Saber and Superhot and stepping outside the real world for a few hours now and then.

Having a standalone, fully independent device is where the game is (literally) at, and this is a field I am going to enjoy keeping track of during 2021 because I’m positive they need decent competition, and it’s not going to come from PC headsets¹.

The caveat, of course, is that the ecosystem seems to be somewhat stalled creativity-wise. There are more copycat shoot-em-ups and re-spins of older games (like Myst, which is a laudable exception because it is just so much better this way) than original ideas, but the potential is there, and I think VR (even if only for gaming) has a better chance of becoming relevant than, say, Linux on the desktop².

In a nutshell, I think of it like this: a Quest is, at least for me, a much better investment than a “regular” games console at this point. I just really wish they’d sorted out the family angle.

With a new Synology NAS and nowhere to travel to, a lot of my stuff has migrated to home infrastructure (ironic for one who works primarily in the cloud, I know).

I now have mirrors of all my GitHub repositories in a Gitea instance on it, as well as my own drone CI/CD setup (for building ARM binaries and some ESP32 stuff) and a plethora of other services, most (including my IoT stuff) fronted by simple but effective Node-RED dashboards³.

All of it sort of gravitated “home” during the latter half of the year, and some of it was set up because I just wanted a to test out a specific feature, but then it also became about a mix of keeping my wanting to rely less on public services.

It’s all containerized, all backed up to the cloud, and requires remarkably little maintenance. However, I’m not using Kubernetes at home, since I can’t really run it on the Synology (yet) and am sticking to docker-compose and vanilla Ubuntu for nearly everything.

Although I have been meaning to overhaul my home automation setup for the whole year, it has also just kept ticking and sprouted tiny ESP32 cameras and other HomeKit-compatible goodies without any hassles.

I’ve been hammering away in the background to move to static hosting, because of late I’ve felt that would be the right way to make some things simpler. Given the current state of CDNs, gone are the days when I’d need fancy HTTP optimizations on the server side, so I’d rather focus on running APIs and spending my compute budget on k3s than keeping a VM up “just” to serve randomly-updated HTML.

I love to write, and have tried to publish at least one thing of substance every week, but truth be told that there is no real reason to keep a CPU core spinning 24/7 to do so.

Call it an exercise in extreme penny-pinching, if you will, or just an excuse to refactor pretty much bullet-proof code, but the current blog/wiki engine is now able to render directly to Azure Storage, partly because that’s what I had handy, and partly because I keep pushing Python and asyncio to the limit.

The entire site can still run on a Raspberry Pi 3A without issues, and I’m testing static generation and upload on the same board, so I’m also having some fun in the process. And no, I haven’t had the time to rewrite the whole thing in Rust yet, but I’m sorely tempted to do so.

The production site is still deployed via piku, and the back-end has been moving around a few different cloud providers (we’re back on Google Cloud this month, since I wanted to clean up my Azure and AWS accounts), but I expect to go fully static in a few weeks, depending on whether I just use drone, move everything to the k3s cluster where I’m keeping a few public services or do something a little bit more creative.

None of us has gotten COVID (yet, that we’re aware), but the lack of actual exercise has taken a toll, especially considering that before January I was out of the house at least half of the day and walked everywhere.

I’ve been feeling tired, anxious and depressed more often than not, and I blame it squarely on overwork rather than confinement.

The standing desk improved things a fair bit, but one of the things I’m trying to figure out is how to do more exercise–and it’s not so much about gear (there’s a hardly used elliptical trainer in my office) but about having the time, by which I mean regular hours.

Since I routinely get meetings scheduled on top of private events and family time and hardly push back, that is likely to have to change next year.

Another thing I will strive to find time for are my hobbies. I have kept (literally) investing in music gear and software, but, like reading, music was one activity I just didn’t have much heart for this year. I did manage to put some things up on SoundCloud, but nothing too fancy yet.

Like exercise, music requires time and motivation, and neither have been easy to come by this year.