3 ms·

I like these lists, but the ones I like the best are the lists that also provides examples and an explanation with the falsehoods. Pointing out a falsehood with

I like these lists, but the ones I like the best are the lists that also provides examples and an explanation with the falsehoods. Pointing out a falsehood without providing a reason does little to change my mind. Falsehoods about phone numbers was recently on HN [0], and I liked how it came with examples. Then it's possible to have a meaningful discussion about each one, and if it is something we need to handle. For instance Phone numbers are always written in ASCII, gives an example where that isn't true. Then we can discuss if that case is meaningful for us to support or not. I cannot build a system that will handle all possible names, emails, addresses, timesystems, currencies, locales or whatnot there is a falsehood article about. So when writing an article like this, if you want it to have impact please explain the point of each bullet. [0]: https://news.ycombinator.com/item?id=19214035 https://news.ycombinator.com/item?id=19214035

MatekCopatek 8y ago
No disrespect, but I think the thought process of your comment is exactly what this post is trying to address. While falsehood lists that cover emails, phone numbers or CSV files are basically just a much better specification than the one you were following initially, this falsehood list works on a different level. You can't simply go through all the items and make sure your code covers each one. You're supposed to read it and come to a conclusion that economics is less similar to mathematics and more similar to sociology. That it's not a set of rational rules but something inherently more complex and chaotic. It can still have a huge effect on your code - not in the form of additional corner cases you need to cover, but as questions you may ask yourself about the purpose and morality of your work. At least that's how I understand it.
- throwawaymath 8y ago
  My reading of the parent comment is that they understand everything you're saying, and their point is that the subject matter is thus a poor fit for the "Falsehoods Programmers Believe" moniker. The phone number example they gave works very well for demonstrating misconceptions. Economics doesn't work as well for that, for precisely the reason you stated about it being partly sociological. In fact it's sociopolitical, which complicates matters further.
throwawaymath 8y ago
Agreed. It would be nice if the author first explained what they meant by economics being "morally neutral" or "racially neutral" before declaring them falsehoods. For about half of these I can follow along with what the author's saying is false (though still, there's no justitication...). But the rest seem very weaselly or subjective.

MatekCopatek 8y ago

No disrespect, but I think the thought process of your comment is exactly what this post is trying to address. While falsehood lists that cover emails, phone numbers or CSV files are basically just a much better specification than the one you were following initially, this falsehood list works on a different level. You can't simply go through all the items and make sure your code covers each one. You're supposed to read it and come to a conclusion that economics is less similar to mathematics and more similar to sociology. That it's not a set of rational rules but something inherently more complex and chaotic. It can still have a huge effect on your code - not in the form of additional corner cases you need to cover, but as questions you may ask yourself about the purpose and morality of your work. At least that's how I understand it.

throwawaymath 8y ago
My reading of the parent comment is that they understand everything you're saying, and their point is that the subject matter is thus a poor fit for the "Falsehoods Programmers Believe" moniker. The phone number example they gave works very well for demonstrating misconceptions. Economics doesn't work as well for that, for precisely the reason you stated about it being partly sociological. In fact it's sociopolitical, which complicates matters further.

throwawaymath 8y ago

Agreed. It would be nice if the author first explained what they meant by economics being "morally neutral" or "racially neutral" before declaring them falsehoods. For about half of these I can follow along with what the author's saying is false (though still, there's no justitication...). But the rest seem very weaselly or subjective.