Thread by @deniseyu21

Just learned a real-world example of the cost of “meh, we can refactor this later”:

Japan is the only modernized country in the world to run on two independent electricity grids, by historical accident. When the 2011 tsunami happened, half the country was knocked off-grid...

And the other half was unable to help out, because the two grids run on different frequencies.

How did this happen?!

In the 1800s, Tokyo entrepreneurs bought a 50 Hz generator from a German company that would later become AEG. Osaka bought one from the US that ran on 60Hz.

This design is a reflection of the political facts of the era: power was consolidated in the hands of local authorities. Centralization came later.

If this isn’t a perfect physical illustration of Conway’s Law I don’t know what is: en.m.wikipedia.org/wiki/Conway%27…

In the world wars the Japanese government floated the idea of unifying the two grids, but ultimately the idea was dismissed because — you guessed it! — it was too expensive. Also, the cultural rivalry between Tokyo and Osaka didn’t help.

When people moved across the bifurcation line, sometimes their kitchen appliances didn’t work. So manufacturers started building appliances that ran on both frequencies. Conversion facilities are being built, but can’t support “production load” yet. Real-world shims and patches!

The cost of not addressing the bifurcated grids comes to light every time a natural disaster strikes the country. In this particular example there are bigger systemic issues and no easy solutions.

Not all technical debt has such a high cost when ignored, but I couldn’t help but think that this story is a nice example of how hard it is to compare tangible present benefit (feature work) against less-tangible future deficit (tech debt).

Source: japantimes.co.jp/news/2011/07/1…