in another time and space …

Over the past month of relocating, I’ve had to put up a lot of my stuff on ebay (alright, it was kijiji.ca). This is a very stressful thing for just about anybody and part of the stress come from the strange weirdos you will meet in the process. Two examples of this are worth mentioning based on the kijiji.ca listing for my car that I expected to sell for $10,000.

Continue reading →

I’ll be honest – When I first saw this ‘White Hot Duck Bath Toy‘, I thought it was a cool, useful product. At the time I probably thought it was cute, would keep the baby engaged and I could trust it because it was more objective (the colour changes to white when it is too hot, which is a clear yes/no answer thus not leaving it to me to make a subjective judgement).

And it all changed when I saw it being used for the first time.

First, it took time to turn white and Second, as one puts their hand down in the tub they already had a 98% accurate sense of the temperature since the hand touched the water anyway. Also, it is stupid to assume a duck in the water will keep the baby engaged and make the bath any easier.

I realized I had been ‘dumbed’. By dumbed, I mean I was made to trust a more deterministic instrument because well, it is more deterministic than my internal thermometer.

Just because it has one better attribute (accuracy in this case) does not make it a better product. It is humbling as a product maker to know what the ‘purpose’ of the product is.

Baby products are an ideal category to search for such ‘dumbing’ products. I wonder if there are other examples of such products. Weight loss products, everything on tv shopping channels, security products for enterprises like hdd encryption tools?

PS: I am not making any judgement here about the financial success of the White hot Duck Bath Toy It might as well have been an awesome lineup for the manufacturer fuelled by the irrationality of first-time parents.

Steve Ballmer revealed that there are 500 new features in the Mango release of Windows (here).

I fail to understand who outside of geek bloggers and their readers would appreciate the value that “500 new features” brings them. Their target customers are unlikely to understand any of the 500 new features and instead IMO they would appreciate qualities like faster boot times, secure computing, better personal data management and seamless use of online services.

I am truly surprised that this company still thinks in terms of features and not what customers want. I would instead speak about how windows could help people manage their task-list, connect their digital cameras and share their pictures online with ease. Or I would scream about how windows would never make them lose their personal data through rockstar algorithms. Exactly the kind of marketing they did for Windows 7 after it was deemed a bad launch.

Disclaimer: I may be biased, but it is hard to take this guy seriously (video).

I was thinking about the design of a traffic light this morning while waiting at one. It occurred to me that since their purpose is primarily providing safety, their design would have to include some fail-safe algorithms. Ideally it would require that at any intersection, all traffic lights would have to talk to the others and respond collectively in the event of the failure of any one of them. At an level of a single traffic light, if there is a failure, you would want a blinking light (ideally) and make sure that it does not stay green in case of any failure. At the intersection level, the neighbouring traffic lights must also change behaviour. For instance, the adjacent roads (ones that are at 90 degrees to the failed light) should not allow pedestrians to walk through.

To add a bit of complexity, you’d want to think of scenarios where there are two failures. In those scenarios it is easy to settle for a very conservative solution (turn everything to blinking lights and assume people will respond like an all-way stop sign). And I suspect that traffic lights today follow the same rule as they were likely to have been designed some 40-50 years ago.

It would be interesting to see if this net throughput could be improved while preserving the same level of safety.