Steve Yegge has another humorously insightful and long post about how hard it is to do big projects.
He uses two examples. The main topic is about legalization of marijuana, which funny enough is not the first end to prohibition. Actually, this time, with each state trying something different, maybe there’s some healthy competition creating a better outcome than just trying to solve the problem all at once everywhere.
The second example is of a bucketing concept for credit card limits. This would help consumers avoid fees by stopping them from spending in various categories once a limit was reached. Its hard to believe credit card companies in general would want to do this since it would mean giving up an average of $34 per card holder and growing. Nowadays, if a bank really wanted to do such a thing, they could just partner with Mint.com, which is the totally awesome site that has solved (many of) the issues in Steve’s credit card example. I don’t see any banks jumping at the chance to save users from fees.
Steve’s obviously incredibly bright and I enjoyed his article even though I don’t think either of his examples are realistic. Making stuff is hard to do.
Speaking of competition, was anyone aware of the fact that insurance companies are exempt from the anti-trust laws? They’ve been skimming 30% of the healthcare pie and they don’t have to play by the rules. Its a good deal if you can get it I suppose.
Starting a new project so I thought I’d look for some standard requirements gathering questions to make sure I had all the bases covered. Google didn’t help me find anything good, but I did stumble on Steve Yegge’s hilarious (seriously LOL) and insightful blog post making the argument that “Requirements Are Bullshit“. He’s right for his point of view which is in mass marketing, but when building in-house custom apps which probably outnumbers mass market apps 10 to 1, requirements are pretty important. I will agree that you have to build something you would want to use. When I put in x-teen hours a week running my app for testing, any feature that’s not easy to use will go on the upgrade list pretty quick since I don’t want a hassle from my own software.
I’m curious enough to ask more open ended questions than most people want to answer anyway, so I’m sure not having a specific list will keep working fine.