Brain Damaged Engineering

I spent most of today working what should have been a fairly simple project in my shop. I was helping my neighbor change the front rotors and brake pads on a 1990 Honda Accord. Generally speaking, disc brake work is some of the easier work to do on a car. Turns out, the 1990-1997 Accord rotors are attached in the most incomprehensible way. Most cars and trucks have their rotors held onto the hubs by the wheel studs and the calipers. Removal of the rotors consists of 2 or 3 steps. Step 1: remove the wheel. Step 2: remove the two bolts that hold the caliper to the spindle. In some cases, a step 3 is required to remove the spindle nut and the rotor (with bearings) comes off.

In any case, every disc brake assembly that I've ever worked on was quite painless to remove. This includes the super heavy duty rotors on my Ford F-250 4x4. Turns out, the rotors on a 1990 Honda Accord are held onto the car with more bolts and more resistance than those on the 8,800 pound F-250.

So, the steps to remove the rotors are, step 1: remove the wheel. Step 2: remove the two bolts that hold the caliper to the spindle. Step 3: remove the axle hub nut. Step 3: push the axle back out of the spindle as far as possible. Step 4: using a E12 (reverse torx socket) undo 4 bolts on the back of the spindle holding the hub and rotor assembly to the spindle. As if step 4 wasn't hard enough, its at this point that things get really ugly. I could go on with the details, but that really isn't the point.

The point I'm trying to make is what possible good engineering went into designing something like this? It can't even be argued that it is done for the sake of generating business for the dealership. There are no special tools really required (although there is one that helps), it is just an exercise in disassembly of the front end. Generally speaking Honda is known for their well built cars and the quality engineering that goes into them. The rotors are still not separated from my neighbors spindles, but we are making progress, maybe day 2 will see it finished.

On a similar note, I brought my grandmother's 1968 Ford Mustang down to the shop last summer for long term storage. Grandma can't drive anymore so the car needed some attention to keep it from completely biodegrading. Since the car hadn't been started in years, before I started it, I wanted to prime the oil system. Generally speaking, this involves pulling the distributor out of the engine and spinning the oil pump with a drill for a couple minutes to push oil through the system. Older Fords are known for being PITA's to work on but very little could prepare me for what ended up happening to me.

The car has a 289 small block which has the distributor in the front of the engine and it is a giant engine bay in those early Mustangs. It should have been no big deal to do this. I marked the position of the distributor cap, then marked the position of the rotor once the cap was off and pulled the distributor. As I pulled it about half way out, I heard something drop down inside. Not having spent much time working on Ford motors, I wasn't sure what that was, but pulled the distributor out without much further thought. I brought over the drill with the oil primer I had bought from Summit, stuck it down into the galley and proceeded to spin the drill. Pretty quickly the drill slowed down due to the resistance of turning the primed pump.

It was now time to drop the distributor back into the engine. Everything was lined up and even put a bit of assembly lube on the distributor gear to make things go smoothly. Nothing doing. I was used to Chevy motors, so I got the flash light to see how much I'd have to turn the oil drive shaft to get it to line up. Turns out, the shaft is a 1/4" hex drive and it sits loose in the oil pump. In other words, the stupid thing can tilt off center by about 15-20 degrees. The bottom of the distributor does have taper cuts, presumably to slip the drive shaft back in when reinstalled, however, this was of near zero value. Long story short, I spent about 4 hours attempting to get the shaft and distributor to line up. Eventually, I got it done, but ended up being 1 tooth off of the timing gear. Fortunately there was enough slack in all of the plug wires that I was able to rotate it to get the timing back. I firmly believe that the engineer responsible for the decision to use a floppy, loose fitting hex drive mechanism out to be drawn and quartered.

So the question I have is, where is the accountability for poor engineering? In the world of professional engineers, if I make a stupid mistake that causes people substantial problems or things can't be built properly, my company has to pay for the problems. We call it "errors and omissions" and we carry insurance to cover those types of things. Where is the accountability for other types of engineers?

Return from the Abyss

It has been nearly 10 years since I wrote anything that was published "online" purely for the sake of writing. When I worked at Red Hat, among the many things I did, I helped market the Red Hat brand. Part of that involved writing a column for an "electronic newsletter". So, I wrote a few hundred words on whatever happened to be on my mind at the time the marketing folks reminded me it was time to send out our next newsletter. In many ways, I've missed that basic one-to-many communication that allowed me to rant, complain or praise as appropriate. To that end, I've decided to take time out of my schedule and continue in that tradition.

Since I brought up Red Hat, I think it is fitting to spend the remainder of this entry discussing what I miss from my Red Hat days. More than anything, I miss the free spirit that built that company. When I started, the entire engineering team fit in one car, albeit cramped, to go to lunch. While the people that made up that early team were as diverse and unique as you can imagine (we had virtually nothing in common except our passion for Open Source and Red Hat), the bonds between us during those early years was very tight.

I remember fondly the events that led to my getting hired. I was a graduate student at Montana State University working on a masters degree in engineering when I got this idea that I should get a job or an internship or something. At the same time, I was using Slackware Linux not Red Hat, but Red Hat was the up and coming distribution. There was a guy named Donnie Barnes who worked at Red Hat and posted heavily on various mailing lists and USENET groups. At some point in time, he mentioned that Red Hat was looking for people to come work for them. I sent a email off to Marc Ewing, the co-founder of Red Hat. After a couple of back and forth email exchanges, he asked me what I wanted could do, I told him I was really good at breaking things and he offered me a job. It was kind of surreal so I tried to catch him on the phone and talk to him in person, which I eventually did. It was a brief call during which he advised me that Red Hat "paid dirt and I had to bring my own shovel". Turns out that "dirt" means $30,000/year which sounded real good to me at the time as I was making about $1,000 per semester as a research assistant.

The stories that I could tell about the early years are many, but I'll end this entry with the story of my first day at Red Hat. After finally finding Marc's office in the suite of mini offices that Red Hat occupied on Cornwallis Drive in Durham, NC, he showed me around the place ending up where I was going to call home. I was introduced to Donnie in person for the first time and that Donnie and I were going to share an office. Let me clarify the term "office" a bit, the office was a room that was so small, when Donnie wanted to leave the office, I had to leave too. Donnie showed me to a pile of very random looking computer hardware and told me to dig through to find a working computer that would be my workstation. I didn't know it at the time, but this pile of hardware was the Red Hat test lab that I was going to be in charge of very soon. Some digging revealed an Axil 311 which is a Sparc clone that was paired with an enormous 19" color display (known as the Hulker) that had 256 colors but was so dim that even in near total darkness, I could usually only make out shades of black on it. And let me tell you about "Stump Boy", my first chair at Red Hat. The most ungodly uncomfortable chair ever conceived by human beings. I believe that chair would have been banned under the terms of the Geneva Convention had it been used in times of war. I ended up using that chair for most of my first year at Red Hat and then proceeded to drag it around to all of the offices I ever had and even back to Montana when I returned years later. The chair has since been removed to an undisclosed location...

Thanks for reading this far, I'm sure I'll put up many other interesting stories along with rants, raves and general rambles. And much to Alan Cox's delight, I'm certain there will be many wangerisms posted over time.