Influencing Nerds

Flexible Expectations

14 February, 2019

A big part of my role at Red Hat these days is supporting events that connect researchers on topics we care about. One great example of this is the first-ever Microarchitecture Workshop, a one-day mini-conference on hardware security vulnerabilities that are exploitable via what is called “side-channel attacks.” The most celebrated of these attacks to come about recently was the Meltdown and Spectre vulnerabilities, explained nicely here by my colleague Jon Masters.

The Microarchitecture Workshop, which happens next week at Boston University and is shaping up to be a great success, is a perfect example of something that happens to me all the time: I start off pushing to make a thing happen with a vision of what that will be, but because I refuse to pay close attention to the details (I am a Connection Tilt, after all), the thing morphs into something quite different from what I was expecting. For a manager, this level of uncertainty can be quite terrifying, and I often promise myself things like “Next time I’ll make sure to get the details right on this thing!” But of course I never do.

Fortunately for me, it normally doesn’t matter whether I dig into the details or not, because what I do actually pay attention to is the people I hand things over to. In this case, I found a researcher at BU, Ari Trachtenberg, who was willing to be involved although he had a quite different vision for what the event would turn into. The result? A smaller, more focused gathering than I had imagined, but I think one that will turn out to be more effective than the larger symposium I had originally envisioned. I was also very fortunate to have help from the very capable Jen Stacy at BU, as well as my new team member Sarah Coghlan. (Sarah keeps telling me how much she loves her job, to which I say “Guess I’d better give you more work then.” So far this is working out beautifully.)

I’ll report back here on how the event actually went, but it looks like it’s going to be a really smart collection of people. Gatherings like this sometimes prove to be turning points in the path toward solving real research problems — maybe this one will fall into that category. Good thing I didn’t try to micromanage it.

(Like that would ever happen…)

Open Source, Management, and Volunteerism

5 February, 2019

My grandfather was a Petty Officer in the Navy in the Pacific in World War II. He never made it to Chief Petty Officer, to his great annoyance, but he learned not long after he arrived in the Marshall Islands that there were plenty of ways to get things done without having that rank. The story goes that in those days only Chiefs were issued toolboxes on the base, so if somebody carrying a toolbox asked you to come along and help with a project, you did. Grandpop discovered this and, not being one to be bound by Navy regulations, would just pick up a toolbox when he wanted to raise a crew for some task, and people would come along and help. Conversely, if he didn’t have a toolbox, no one was interested.

One of the things I like about working in open source, and at Red Hat specifically, is that you don’t need to be carrying a toolbox to get people interested in working with you on a task. All you really need is a good idea and some knowledge of who the people are who might want to help. I’m not sure why this should be so — maybe it is that so much of our work is in open source communities that are powered by volunteer effort, or maybe we just all believe we know enough to decide for ourselves whether we want to help with something. Either way, many of the best things we have done at Red Hat have come not from top-down decisions but from “coalitions of the inspired” getting together and pursuing something that looks worthwhile.

There are some drawbacks to this culture. It tends to put a strain on people who are willing to volunteer, because not everyone is — when I was in Brno, we had a lot of manager committees to look after the site, and it was always the same 20 managers who showed up to help (out of 100). It also makes it hard to know whether people are doing what they’re really supposed to be, or working on side projects they find interesting.

I am fairly sure, however, that I’d rather have it this way than the way I imagine things work at other large companies — doing just what’s required, and lots of politics. Plus I don’t see how I could carry a toolbox to all the meetings I go to.

Working, Writing, Thinking

14 January, 2019

I am old enough to remember when work was not about sending and receiving email. When Kim and I founded The Harp Column in 1993, our business was done on the telephone and by writing letters. In fact one of the first things we had to do was design a logo so we could get letterhead printed. Although email certainly existed — I think I had my first email address in 1992, maybe — it was much more of a novelty and there was no expectation that everyone had an address.

This meant that the telephone was all-important and that making and receiving calls was a major part of the day. In fact many people bemoaned the decline of writing skills brought on by the ease of picking up the phone to communicate. The telephone was, in fact, tyrannical. If you were serious about your business you did not want to miss a call, and we went to great lengths to make sure we paid the phone company for voicemail service and processed it carefully. My phone greeting — “Good (morning|
afternoon) The Harp Column, may I help you?” — still rolls off my tongue 25 years later.

So, I’m glad that we’ve returned to writing as our primary means of business communication. Despite the overwhelming volume of mail we all get, I would still rather craft a sentence or two than have a phone conversation. What surprises me is how much writing actually dominates my work. All my most productive time is spent writing or editing other people’s writing, and in fact in the better meetings I’m in I am writing too (if you can describe my crappy note-taking as writing). It’s strange to think that the simple act of putting words on paper or a screen should be considered work, but it is, and if you’re a manager it’s the most important kind.

At some point in graduate school I latched onto the idea that in fact thought does not exist without language, and therefore thoughts as discrete things do not exist until they are written down, and in fact thoughts change and improve as they are being written. (This is why the term “wordsmithing” drives me crazy — the implication that written thoughts can be clarified later by someone who happens to be “good with words” is an indication of laziness on the part of the original author. What it really says to me is “We haven’t really worked out what we mean to say here so we’re going to punt it to some poor hack who wasn’t in the original conversation.”)

So writing is in fact thinking, and if your work is thinking — i.e. if you are a “knowledge worker” — then your work is writing, and the better your writing is, the better your work is.

Now, remind me again why we’re not teaching English any more?