Simplifying creating an account on behalf of someone else could help, it seems. Then the author could likely have helped out the woman directly, leveraging the trust accrued by her own account.
But if the author had her own mobile phone with her, it seems she could have used this plus existing account creation mechanisms anyway?
"Star analysts who change firms suffer an immediate and lasting decline in performance."
This sounds a lot like "regression to the mean", that is, people performing better than their "mean" performance level for a time but longer term trending back to it.
The main innovations seem to be the example-based scraper and the focus on having end users decide where to apply the transformations. (What are some similar projects?)
Having access to more wealth gives people more faith that the government will be around a long time, will not do crazy things with the currency. That's not the same thing as the currency having more intrinsic value. (Also, it would not seem to make sense to talk about backing a currency with taxes denominated in that same currency, would it? But I'm unclear what you meant by "provide value".)
(What I've heard about bitcoin makes me a bit skeptical, but I want to be skeptical for the right reasons.)
It seems like this in particular applies to most government-issued currencies as well (except that they have very good marketing and market share). Each currency itself has no intrinsic value and depends on "marketing" (reputation of the government minting it) and market share (number of places accepting it). If people lose faith that a government will do a good job managing its currency or even think that the government will disappear soon, the worth of the currency (in terms of what people will give you for it) will probably fall.
A couple people have mentioned the query plan changing in the middle of the night. Is the scenario a single query getting executed two different ways by the same running system? Or is it from upgrading to a new version of PostgreSQL and hitting different behavior? If it's the first scenario, would using prepared statements help somewhat with avoiding unexpected changes?
// Edit: Sorry for the naive question. I see that other people are saying that is the first one.
The Stack Overflow discussion points out that the patent hitting rollover menus was going to soon be (and is now) expired: "A long, legal way to say this patent is public domain on 10/5/2010 and these guys are probably trying to snag a nickel before it goes out."
There's no irony in advocating using a richer array of types in a dynamically-typed language. In fact, static vs dynamic typing doesn't seem apropos here since it's also common in statically typed languages to use strings for everything, leading to the same problems.
But if the author had her own mobile phone with her, it seems she could have used this plus existing account creation mechanisms anyway?