Hive Blog

Engineers and software developers are among the brightest people I know. Yet, they can also be the most uninsightful at times.

If you’ve ever watched the Angry Video Game Nerd (AVGN), you’ll find plenty of examples of “what were they thinking?”. James the Nintendo Nerd takes gamers back in time to look at the bad video games of the past, although expressed a lot more colourfully with how-to-say “artful language” spliced in. If you think engineers lacking insight is a new phenomenon then this demonstrates otherwise. In fact, if you look through history you’ll find tons of engineering’s biggest mistakes. Mistakes are one thing, lack of foresight is another.

Mistakes can range anywhere from failing to see the obvious flaw in the design to ignoring the behaviour of users or even completely missing the market altogether. I’m going to pick a few examples to illustrate.

SIP (Session Initiation Protocol), the leading Internet VoIP (Voice over IP) technology on the planet, was designed by some of the brightest people on earth. However, due to purists’ design principles, SIP missed the obvious flaw.

The Internet could be drawn to show every computer sitting at all ends of the network happily exchanging information with any other computer and sending data through the cloud in some kind of utopian model. Except, that model is fundamentally wrong. The only computers that can send information to any other computer are servers. The rest of those personal computers are sitting behind hardware or software firewalls and barriers that intentionally obstruct communication in the name of keeping us safe from Internet hackers and malware. Although there are a few brave souls who connect themselves directly to the Internet and ride out the storm, they are the exception to the rule (or they did so accidentally without realizing the danger).

In the real world, if you want to communicate P2P (peer-to-peer) you need to bootstrap the communication through an intermediary and sometimes, due to extreme protectionism, you’d have to tunnel the communication through an Internet relay.

So why am I picking on SIP? Because the SIP protocol intentionally ignores reality. “Firewalls” are not part of the protocol and are their issues they bring are not to be solved in the SIP protocol layer. Sure, SIP is an elegant protocol as a result. In fact, it’s very clean compared to other technologies, *cough* H.323 *cough*. Except, without taking into account firewalls, you have to add extensions and other protocols to make the entire thing work in real world situations. The initial solution was SBCs (Session Border Controllers). SBCs kludge the SIP protocol inserting themselves into the network path so they can ensure all traffic flows between peers correctly. Then engineers added on STUN, ICE, TURN, rport, outbound SIP, edge proxies, GRUU, broken SIPS (secure SIP), VPNs, SIP “friendly” firewalls (which are often opposite of friendly) and now we have ourselves a huge mess. If you aren’t familiar with all of these terms and their respective meanings, don’t worry because many people in the SIP industry don’t know either.

At the end of the day, you end up with competing interests between all these solutions with each having their place at the table when it comes to future designs and protocols to ensure technologies remain backward compatible to their solutions and protect their market. From here on out, the problem cascades deeper and deeper. If you factor in bleeding edge companies implementing protocols still in an early draft state (because these companies need solutions to problems now and not years from now) that just ads salt to the already blistering wound.

Suddenly, what was an elegant solution is a complete and total mess. Trust me! I know firsthand having worked in this domain for many years. Still, SIP is infinitely better than the alternatives (unfortunately that doesn’t say much). And moreover, it works. Well, most times.

For an easy picking, and perhaps lazy blogging, I’m going to pick on Microsoft next. What were they thinking when they developed the Surface PC? I mean, it’s wonderful to have a huge table computer that can display pictures, video, maps and email. But at $10,000, it’s a high end gimmick at best. There is only one reason for this technology and that is to look cool and get a bunch of reporters together to marvel at its wonder.

The old fashioned favourite lame product is Microsoft Bob. Released in 1995, this was Microsoft’s attempt to win over those too challenged to use a computer by making cute little cartoon characters to help guide you through using a computer. The solution is not to make cartoons for grownups and help ensure that a computer gets thrown out the window! The real solution is to make the operating system easy by being genuinely easy to use! Although, I’m not sure I’m proving a point since I doubt that an engineer came up with this flightless turkey.

The entire dot com era was proliferated with engineering ideas and big money. Anyone with an idea and with a dot com extension would find themselves funded. Born out of that era came ideas like Flooz and Beenz. For some reason, having real currency wasn’t good. As a result, an online currency was born and seemingly just as quickly, disappeared from history. I guess credit cards, having been the standard through time, were not good enough!

Another failure was Microsoft Passport. The idea is sound and frankly, needed; create a single login account and use that for all your services online. The trouble is pretty obvious though if you think about it for a few moments. Who do you trust to be that single login? Microsoft is not the best liked or trusted company in the industry and they have a reputation, for right or wrong, that they want to own everything. Microsoft introducing this service is a bad idea. Again, engineers missed the obvious right from the start.

So what does this have to do with dinosaurs and engineered steak knives? Sometimes a tool is just completely wrong no matter how good the design. Moreover, dinosaurs are long gone so there isn’t much need for killing them in the first place, thus engineering a better steak knife is a bit pointless. The meat lover in me does wonder however, what dinosaur steak would have tasted like? Surely, it would take a well engineered steak knife to cut through the shank of a tyrannosaurus rex!