The Crushing Weight of Knowing What You’re Doing

“Who are you and why are you here?” –Dave Cutler (DaveC)

Source: 012. I Shipped, Therefore I Am

Steven Sinofsky, once a huge wheel at Microsoft, for a very long time, is writing a series of articles chronicling the halcyon days of the early PC business at Substack. I can’t quite bring myself to subscribe, because most of it is free already. Plus, there aren’t many surprises for me, since I was living it during that time.

When Windows NT was introduced, I was quick to jump on board. I was already experimenting with Linux towards the end of ’94. But then I saw a disc of NT 3.5 (not even 3.51 yet) on someone’s bookshelf. He said he wasn’t using it, so I snapped it up and installed it. For the next 20 years, I would dual boot my PC’s between Windows and Linux. I only used Windows for gaming, but for that use, it was obstinate. I tried every incarnation of wine and Crossover and PlayOnLinux and everything else. Nothing has ever let me run Windows games on Linux well enough to warrant getting rid of a native partition.

The content of the slide above is of no consequence, as is pretty much the case with all presentation slides. What’s interesting to me is the little toolbar on the top, left side. It’s from the early Office XP days, back when Microsoft was new and cool. “Before the dark times. Before the empire.” Seeing it evoked a visceral response. As a computer nerd, those really were interesting and exciting times to live through. From the article, that screencap is from 1992. Competing against giants like IBM, HP, and Sun, Microsoft’s eventual dominance was anything but sure at that time. And that’s what’s prompted me to write this anecdote.

In 1995, my Fortune 250 company didn’t even have an internet connection yet. I was using a phone line, and a modem that I conned my boss into letting me get. It was over this modem that I downloaded all 54 floppy drive images of Slackware Linux, on a computer running Windows 3.11 with Trumpet Winsock, connecting to a free SLIP dialup bank in California.

At first, I was much more into NT than Linux. I skipped Windows 95 entirely. I don’t think I ever had a computer that ran it.

I remember how easy it was to setup a dialup connection in NT. By 1996, I was running a dual Pentium Pro with 384 MB of RAM, SCSI hard drives, and a $2,500 video card to do FEA work. The total cost was about $10,000. A coworker got a SGI Indy to do the same sort of work, to the tune of $80,000. The company still didn’t have an internet connection, so he also got an external modem, and hired a local ISP to come set it up. The guy came and screwed around with the connection for 4 hours. I kind of razzed him, by pointing out that it took me all of 15 minutes to configure the same thing on NT. That’s how smug I was about NT versus Unix at the time.

The best part was still to come.

For the next week, the ISP guy still couldn’t get that Indy on the internet. Every time it would connect, the kernel would segfault, and the machine would crash.

But that’s not the best part.

The ISP guy worked with SGI to patch IRIX to fix the modem driver, and finally got it working. My coworker left it connected to the internet all the time to get his email. Things worked fine for a few weeks.

Then the company got a T1 internet connection, and then connected our facility to the main office via a sonet ring. I was really looking forward to not needing my dialup connection any more. But, the first morning, no one could access the internet. Complaints were made. Investigations were performed. Our internal IT would fix the problem, and then it would come back.

Here comes the best part.

Finally, someone realized that computers inside our facility were getting the wrong gateway address to get to the internet. They realized that they were picking up the IP address of my workmate’s Indy, which was advertising itself as a route to the internet, and since the number of hops from computers in the office to the Indy were less than skipping over to the central office, they were preferring its modem, and the Indy’s phone line would choke from the load.

I recall very clearly that there was a simple checkbox in the dialog for setting up a dialup connection in Windows NT for advertising the connection to the LAN as a route to wherever you were connecting. It was on by default, but when I was running through the process, I quickly realized that this was NOT what I wanted, and un-ticked it.

And I felt pretty smug about being serious about NT at the time.

I stuck with NT as my primary interest until some time around 1998 or so. Then Nat Friedman and Miguel de Icaza released Ximian Desktop for Linux, which made Linux on the desktop really pleasant to use. I wasn’t doing analysis work any more. I had transferred to become the system admin of all the Unix machines in the advanced engineering group, so running Linux was a perfect fit. After that, it was pretty much all Linux, all the time, until switching to Macs just a few years ago.

IT Project “Thermocline”

Source: https://brucefwebster.com/2008/04/15/the-wetware-crisis-the-themocline-of-truth

A thermocline is a distinct temperature barrier between a surface layer of warmer water and the colder, deeper water underneath. It can exist in both lakes and oceans. A thermocline can prevent dissolved oxygen from getting to the lower layer and vital nutrients from getting to the upper layer.

In many large or even medium-sized IT projects, there exists a thermocline of truth, a line drawn across the organizational chart that represents a barrier to accurate information regarding the project’s progress. Those below this level tend to know how well the project is actually going; those above it tend to have a more optimistic (if unrealistic) view.

This is all true, but the article assumes that everyone is acting rationally, in service to the stated goal(s) of the project, and that problems with the timeline are just honest mistakes. Unfortunately, in my 25 years, I’ve witnessed a nauseating amount of political infighting that sought to undermine projects in attempts to build and/or preserve personal power. This behavior employs the two things readily at-hand for ruining estimating: bad-faith technical decisions, and good, old-fashioned feet dragging. So the problem isn’t just people being wrong, there’s also a large component due to people actively sabotaging a project for their own purposes. 

I’ve spent most of my career in Fortune 250’s, but I’ve seen this happen in a couple of very small companies too. As someone with a personality that is honest to a fault, this has caused me a significant amount of distress in my career. More than once, I’ve been the lone voice in the wilderness crying about the forthcoming train wrecks, only to be ignored, and then ultimately blamed for the crash, because I was the only one that people could point to for having said anything about it at all.

The Gervais Principle II: Posturetalk, Powertalk, Babytalk and Gametalk

Finally, Sociopaths and Losers speak rarely to each other at all. One of the functions of the Clueless, recall, is to provide a buffer in what would otherwise be a painfully raw master-slave dynamic in a pure Sociopath-Loser organization. But when they do talk, they actually speak an unadorned language you could call Straight Talk if it were worth naming. It is the ordinary (if rare) utilitarian language of the sane, with no ulterior motives flying around. The mean-what-you-say-and-say-what-you-mean stuff between two people in a fixed, asymmetric power relationship, who don’t want or need to play real or fake power games. This is the unmarked black triangle edge in the diagram.

Source: The Gervais Principle II: Posturetalk, Powertalk, Babytalk and Gametalk

I am re-reading the whole series, because it came up as a response to something I said on HN. Basically, I had reinvented this 3-layer dynamic from first principles, based on my observations of the past few years of my career. Now that someone pointed me back to it, I remember reading it originally, but this was written twelve years ago now.

Anyway, this passage really resonates with me. Every time I’ve gotten face time with a serious power broker in a company, this has been true. No games. No BS. Just straight down to business. I have something to say that will help the organization, and they’re ready to hear it and incorporate it. It never accomplishes the full intention, but I understand that they have a lot more pressures that I can see from my vantage point.

… for Sociopaths, conditions of conflict of interest and moral hazard are not exceptional. They are normal, everyday situations.  To function effectively they must constantly maintain and improve their position in the ecosystem of other Sociopaths, protecting themselves, competing, forming alliances, trading favors and building trust. … They never lower their masks. In fact they are their masks. There is nothing beneath.

Though distant from our worlds, criminal worlds have the one advantage that they do not need to maintain the fiction that the organization is not pathological, so they are revealing to study.

For me, as a non-sociopath, this is a source of continual failing: to recognize that the the people pulling the levers of power in the organization are, in fact, sociopathic, and out for their interests, without regard for anyone else’s feelings or fortunes, not mine, or even necessarily the organization’s. Forgetting this base and simple fact has bitten me in the rear end more times than I can count.