I have been terrible about blogging lately, but it has been for a good reason. In February of this year I started a new company, called Grokky. I cofounded it with my former HubSpot colleague Patrick Fitzsimmons. We have been absolutely heads-down learning and building product, but I wanted to quickly share the idea for the company on this blog.

The quick version of the pitch is this: While at HubSpot, Patrick and I realized there was a big problem with companies understanding the quality and competency of managers. This resulted in managers with some bad habits (myself included, I am still learning how to be a great manager). Grokky is a software platform to make management measurable and transparent. If you want to keep up on the latest with the business, I suggest you subscribe to the Grokky blog.

I’d like to say that I will be better about updating this blog, but I know that is not the truth. I will share periodic updates and more personal thoughts here, as I always have.

Market

Commonwealth’s market is a welcome addition to Kendall Square which has, to date, lacked any kind of place to grab eggs, butter, or orange juice. Unfortunately, Commonwealth is not comparable in scale or ambition to Bee’s Knees which in my mind is the epitome of an urban high-end market. Some elements of the market are a win, like the refrigerator cases filled with home made ice cream, fresh dairy, and seasonal vegetables. But other attempts in the market fail, like the antique cast-iron stove filled with outrageously overpriced jars of Hellmann’s Mayo ($4 for 8 ounces) and bottles of Heinz Ketchup. Although you can grab a baguette and some house made jam, there is no delicatessen, butcher, or cheese monger available.

Dinner

Commonwealth is a farm-to-table concept, which is all good and well, but the menu errs on the side of oversimplified. The best starter may be the home made Omni-Parker House style rolls served warm with good butter. The scallop ceviche satisfies but the meager portion hits low on value. The bowl of mussels is a better option for a seafood starter that satisfies. Main courses are split into a la carte proteins and sides. The meat and fish options are all in the $20-30 range. All sides are $10, resulting in an easy $30-40 per diner, just for the mains. Again, not a great value, but forgivable. Of the main dishes sampled, all were cooked and seasoned properly. The downfall of dinner is the largely uninventive menu. Everything being a la carte, there are no composed dishes. Items listed on the menu are exactly that, no herbs, spices, or accoutrement. I prefer to dine out for items and preparations I can not achieve simply at home, and Commonwealth does not meet that bar.

Dessert

The desert menu is another a la carte offering. You choose from a selection of starches (waffle, brownie, pain perdu), ice cream flavors made in-house, and toppings (fudge, caramel, fruit). Picking out the perfect layers to build your sundae is quite enjoyable. The portions are on the small side (continuing the value question), but the flavors are all very good, especially the homemade ice creams.

Lunch

Kendall Square has no shortage of lunch spots, but Commonwealth is a fantastic new option. There are typically 5 sandwiches and salads available for take-out or dining in. Of three take out attempts the sandwich portion varied, it may be that the kitchen is still dialing in the right size sandwich for the $9 price point. Sit-down lunch offers the same menu of sandwiches and salads for $12. Sandwiches are served with sides of both duck fat fries and salad greens. The ‘steak and cheese’ sandwich is simple but tasty, with deep beefy short rib flavor. The pork belly bahn mi is full of bright acidic veggies, but the pork belly varies between soft and juicy and dry and mealy. The roast beef has a welcome kick of horseradish mayo, but is otherwise just an average sandwich. Overall, Commonwealth is an affordable and enjoyable option for a work-day lunch.

Service

Market aside, the service at Commonwealth is fantastic. The bartenders are knowledgeable and talented. Though on my night visit they were behind with mise en place and took a bit too long to prepare cocktails. Servers and runners were courteous and efficient. The team was helpful but not overly chatty, interactions were pleasantly terse. The Market staff seems to vary from jumpy to unaware. A greeter near the door is quick to chase you down and explain the footpath to follow for market, take-out, or sit-down, even if you happen to know where you are going. A couple of team members behind the market counter are standing around at lunch time, unsure of their role. Generally, though, the team is happy and accommodating.

Decor

Wood pallets. Everywhere. On the walls. On the ceiling. The tables themselves. Personally I think they over did it but some of my friends were not in agreement. Besides the pallet-overload, the Commonwealth space is nicely laid out. There is ample space for the different areas of the restaurant and market. Though there is a bit of a navigational issue around the market counter. At lunch time many people will enter the register line from the wrong side. The space has a wide-open and industrial feel that is well suited to the character of the restaurant.

As software developers, we’ve all been there. Sometimes we are either too tired, stressed, or drunk to be trusted with a production deploy. For that reason we took to protecting our deploys at HubSpot. Nothing complicated, just a simple math challenge to make sure your brain is (fully) functioning before you start rolling code out to production. I posted a python example on this blog less than a year ago, but I have recently updated it. Now instead of a method, I’m using a decorator, this feels more pythonic.

Save yourself from a slipshod deploy, start using Deployment Beer Goggles.

I am excited to see the detailed plans, but before that happens I’m going to take a stab at how the system works. I’ve spent some time putting together the pieces of the puzzle that Musk has revealed. Other articles have been written about the Hyperloop, but most seem to focus on an evacuated tunnel system (vacuum) and magnetic levitation. Musk has not really addressed these comments, but I’m pretty sure the Hyperloop will not rely on vacuum or MagLev.

Part 1: Vehicle

When Musk described the Hyperloop as part air hockey table, I believe he was indicating that the vehicle would ride on a cushion of air much like an air hockey puck. In the case of an air hockey table, jets of air shoot upward from the surface of the table reducing the friction on the puck. However, for the Hyperloop I don’t think the track will have air jets, if for no other reason than having air jets along 400 miles of track is cost prohibitive. Instead the train will be a Ground Effect Vehicle. A GEV uses wings to create an aerodynamic interaction with the ground and maintain a low altitude. In this way, the vehicle will ride on a cushion of air achieved from forward momentum and not from jets of air on the track.

If you think of the vehicle as an airplane that is never meant to be much more than a few feet above the ground, Musk’s mention of the Concorde makes more sense. In an evacuated tunnel, there is no air and therefore no air friction. Although creating a vacuum is easy on a small scale, maintaining a vacuum for 400 miles of track is insanely difficult. Instead the GEV will ride on a cushion of air and it will have to deal with air resistance. The Concorde was a major feat of engineering – the planes cruised at Mach 2 (1,500 mph) – and the fuselage was designed to be extremely aerodynamic. I believe the Hyperloop will use a vehicle design that incorporates a lot of the same aerodynamic considerations as the Concorde, but being a GEV will only ever “fly” a few feet above the track.

Part 2: Propulsion

Musk commented that the Hyperloop would be part railgun, which is probably the most straightforward clue to the propulsion of the system. In a railgun, an projectile is launched through an electromagnetic interaction with parallel conducting rails. The railgun design is fairly straightforward: A circuit is created by sending positive current up one rail and negative down the other. An armature completes the circuit and the force is directed at right angles to the plane of the tracks. The electromagnetic interaction propels the armature down the rail. In the case of the Hyperloop, the vehicle will likely integrate the armature and therefore is both the projectile and the “launcher”. Because the track and vehicle form a single loop of current, this is likely where the name Hyperloop derives from.

The railgun is not a new piece of technology. It was theorized in the early 20th century, and the Germans built the first working railgun as an anti-aircraft weapon in World War II. The WWII railgun was capable of propelling a 1-pound projectile at 6,600 feet per second. As a comparison, the Hyperloop will only need to achieve ~1,200 feet per second to travel SF to LA in 30 minutes. Of course, the mass of the projectile is also much greater with the Hyperloop. But assuming modern composites being used to construct a lightweight vehicle that carries no propulsion system, and modern electronics controlling the railgun, the Hyperloop achieving this velocity is plausible.

The Hyperloop track system is unlikely to be one continuous 400 mile electronic circuit of. Instead, each section of track would be its own fully contained railgun with the projectile travelling between them. Using solar panels and batteries, each section of track can build and store the energy needed to propel the vehicle. Not every section needs to be active all the time, the system can be designed in such a way that some sections are propelling, and the vehicle is coasting through others. Therefore, even if there is a long solar charging time for each portion of the track, sections can alternate and allow for near-continuous operation.

Part 3: Safety

The Hyperloop was described by Musk as “uncrashable.” I think this may be a bit of an exaggeration, the rules of the universe make it possible to crash anything if you try hard enough. But in its default operating state, the Hyperloop could be an extremely safe vehicle. The vehicle rides on a cushion of air, making it impossible to derail. Small vertical wings could be used to stabilize the vehicle horizontally against the track walls. If the track were to lose power, the vehicle would simply land on small wheels used to support it during takeoff and landing at the terminals. The only major issue I can think of is the track coming apart, if that were to happen it is possible the vehicle could launch out. Computers could monitor the track and shut down the system if any section of track lost its integrity, but there is still opportunity for a crash to happen. Perhaps Musk has some sort of tether system designed as part of the railgun’s armature to prevent the vehicle leaving the track. Overall, the system could be very safe with fewer variables than the car or air transport systems of today.

Conclusion

There is no question that the Hyperloop will be an amazing breakthrough in transportation – regardless of the design – if it can achieve the stated challenge of travelling from SF to LA in 30 minutes. I have spent quite a bit of time reading what Elon Musk has revealed about the system, and although I may be proven wrong I am quite sure the Hyperloop will not use an evacuated tunnel system or MagLev technology. I believe the Hyperloop will employ a Ground Effect Vehicle and solar powered railgun track sections to propel a lightweight, aerodynamic passenger compartment. The components of this system are all approachable, though constructing them into a massive 400-mile system is anything but. I am very much looking forward to the official announcement by August 12 to see what the mad genius Musk has in store.

Have your own thoughts on the Hyperloop? Join the discussion at Hacker News.

TEAM:

MARKET: Small ideas will not garner much interest. You have to have a story line that will last 5+ years and innovate throughout to stay relevant. The core premise needs to be big enough to carry multiple story lines, the team, too.

EPISODES: A consistent upward trajectory is unlikely (and boring). There will be ups and downs. There will be emotional moments. Take as much ownership for wins as the losses.

Hybrid Retail is a physical outpost of a predominantly online brand. This has come into fashion recently as companies like Bonobos and Warby Parker establish brick and mortar locations. The storefronts are a bit unusual in that they do not keep stock and orders are fulfilled through the same warehouse systems as online orders. Another business I will add for consideration is Tesla. The Tesla showrooms do not actually perform any sales function, all the reservations and orders are done online.

In the case of Tesla, the Hybrid Retail model was forced rather than designed. Automobile dealers have state and federal laws regulating new car dealerships. Tesla gets around this by allowing customers to pre-order their car online. All sales legally take place in California and the car is delivered to your home.

For a car, buying on a computer at the store makes sense. You are going to wait for delivery anyway, there’s no chance you’ll be able to drive off the lot with your car. Arguably, this model could work for Warby Parker as well, because you are buying prescription eyeglasses. For a clothing retailer like Bonobos, I think this model is much less practical.

Traditional retail is all about getting what you want instantly. Warby Parker beats Lenscrafters on price, but not on time-to-happiness. Same for Bonobos vs. Gap on style. When you offer a product that nobody else has (i.e. Tesla) the Hybrid Retail model can work. But when you have a commodity good like eyeglasses or clothing, the convenience factor is huge.

Amazon and Walmart are investing heavily in same day delivery programs that will provide large catalogs, low prices, and great time-to-happiness. Can Warby Parker and Bonobos offer amazing in-store experiences that get customers to look past the convenience issue? Only time will tell if the Hybrid Retail model can really work, though I would bet against it.

In my experience, there is no better gut check for your own commitment to a startup than to try to talk somebody else into it. Discuss your own passion for idea. Explain the current team and roles. Talk about the big picture and where this startup will be in 5 years. To recruit somebody (good) you have to be clear and convincing.

It does not matter if you are the CEO or the janitor, you should be able to get someone else excited about your startup. If you can not, you probably should move on.

Last night, I had a beer with my friend Nick Francis (CEO of HelpScout) when he said the quote above. We were talking about one of our favorite activities, go kart racing. The F1 karting track near Boston left us wanting more from the experience. It started with an annoying process of filling out paperwork and waiting for our group to be put on the track (often an hour wait even with nobody on the track). After our third or fourth visit, we were expecting a better process. We had purchased racing licenses and were spending money on each race. But, when we left the track, we didn’t feel special, we were left with a bad last impression. And we did not want to go back.

Another example of this is the cruise industry. When I get on a cruise ship, there’s a great first impression with music and free champagne. But when I am disembarking, they hand me a bill for the week’s charges, and assign me to a waiting area until unloading time. After sitting around waiting to get off the ship, I trudge single-file through customs and immigration, and all this at 7am. Again, it leaves me with a bad last impression.

So what businesses leave you with a great last impression? In my opinion, there is one company doing this right: Uber. Getting an Uber car is easy with the mobile app, but the magic is really when I arrive at my destination. I smile knowing I don’t have to pull out my wallet. I get out of a luxurious black car, curb-side. I feel like a million bucks when I use Uber. It is a fantastic last impression and I look forward to using it again.

Consider the last impression you leave your customers. The business world is over-optimized for first impressions, because of the old adage  ”you do not get a second chance at a first impression.” But, if your customer’s last interaction with your business is not also excellent, they will not be coming back. Nick is right, first impressions are overrated, leave your customers with a great last impression.

On December 9, 2012 I was out for a run in the morning, about 4 miles from East Cambridge into Boston, over the Longfellow and back up along the river to my building. Just before I got home, I had to cross Route 28 and while clearing the center median I caught my back foot and fell in the road. I was running full speed, so it was more of a tumble than a fall, on asphalt. I quickly got up and out of the way of the oncoming traffic. When I got to the sidewalk I realized my finger felt strange, and when I looked down I could see it was dislocated. The middle joint was out of place and turned about 45 degrees sideways.

While still on the sidewalk, I decided to see if I could reset it. I pulled a bit and the finger slid back into place. I was feeling pretty good about my amateur orthopedic skills. I went inside and showed Krystal, but by that time my finger was neither out of shape or swollen. I decided against going to the hospital for the time being. I iced the finger and we got a splint from the drug store to immobilize it.

1 Day After Injury

The doctor and her very nice PA decided to set me up with a hard splint which is basically three-quarters of a cast, going from the tips of my fingers up to the middle of my forearm. It was awkward and uncomfortable, but necessary to keep the joint protected and allow the bone to heal.

A few weeks later, we removed the splint and took more X-rays. Luckily the bone was healing well, but the joint was still not aligned properly. I was sent over to Occupational Therapy where they made me a hard plastic splint that bent into a sort of claw shape. The idea was that a constant bend would help to align the bones. A couple more weeks and more X-rays showed that this did work.

4 Weeks After Injury

I’m thankful I did not have a worse injury, and this one just proves that accidents happen and no amount of caution in exercise can keep you completely safe. I also learned that anytime there’s a dislocation, you should definitely go to see a specialist. Even a small fracture can complicate things greatly. The longer you wait to get it taken care of, the worse your chances of recovery.

For about a month, my ability to type was very limited. It was an interesting challenge. I still had to write code and emails, so I found ways to simplify. I did a lot more copy-paste and reused code wherever possible. I found ways to think through the problem with pen and paper before moving to the keyboard. I wrote terse emails, I think I’m actually going to stick with that. A month of one-handing typing was bearable, but only just. Though now that I’ve been through this, it was not as bad as I had envisioned.

I am not an expert on hand injuries or therapy, but I’m happy to help with any questions readers may have. Leave a comment or connect with me on twitter @abdinoor.

The great Patrick Fitzsimmons had written a similar script for our C# applications, and his included a math challenge with every deploy. I believe this was modelled after the Gmail Mail Googles feature that prevents you from sending email late at night, drunk. Anyway, I loved this little feature that Patrick coded.

The other day, George was deploying some code here at BabbaCo and acidentally pushed to production instead of staging. Not a big deal, but it brought the math challenge concept back to mind. So I just updated our fabric deploy script with this code to challenge the user with a simple math problem when deploying to production.

Consider plugging this into your own deploy script, or other places where it might be useful. Enjoy!