And yet, the myth of Japanese "uniqueness" is largely the product of Japanese national propaganda, serving many useful purposes for the state, for example increasing tourism and limiting permanent immigration. I find it quite fascinating in modern times this propaganda still lives on strongly throughout the world, as people say "things are so different in Japan."
You'd probably enjoy this meme:
But Japan is a condensed WTF/sqm if you ever visit you will see that it’s truly a special country.
There were a lot of pro-China articles and comments on HN about 5 or so years ago, China’s popularity has just taken a huge nosedive since then. Sexy Cyborg makes China look cool until the government shuts her down, making China look very uncool.
What specifically are you balking at?
His solution was to communicate much more actively with the Indian team, to try to squeeze more feedback that they would hesitate to give at first.
It is a mix of development, very different culture (Brazil has a lot of European and US cultural influence), but also safety, institutional efficiency, traditions. Idk, there is a lot of different aspects and things.
That’s the core point. Of course Asian culture is more different. Probably a better example would be for a Brazilian to visit some African tribe.
I visited African tribes, and of course it is different than a big city in Brazil, it is just as different as an indigenous tribe here in Brazil, or a more rural village.
Japan uniqueness, I believe, comes from its urban development. Rich, clean, polite, functional big cities is the uniqueness that I see in Japan that it’s not common in any other place in the world. I didn’t find that in LA, NY, Paris, London, Dakar.
Maybe Copenhagen (from the ones I visited) would match some aspects of it, but then it’s not as big as Tokyo, nor it is blended with the tradition as in Japan (or I am used to Nordic traditions more than Japanese)
I am convinced that Japan has some unmatched uniqueness which make it more different of other countries than other countries are among them.
Is the Consumer Reports that we're commenting on also Japanese propaganda?
It seems to fit well engineering, but it is not that good of a fit for software development - as they are generally not too great at the latter. I have only heard these things, so not even first-hand experience, but due to a senior people basically “have to be” at a higher position, there are software companies where the higher ups don’t know shit about what is even a computer.
Though I guess it's weird we didn't do that a lot of Chinese dishes. Soup dumplings instead of Xiao long bao. Idk how to even properly pronounce that.
> Idk how to even properly pronounce that.
And there's your reason why we say dumplings and not xiao long bao! :)From what I've heard, the Japanese has no qualms about importing English words or just making up "nonsensical" new English words to suit their needs.
Want more counterexamples? Probably literally every Italian food: spaghetti, pizza, ravioli...
To me, this is just more evidence for how strong this meme of Japanese uniqueness, coolness, "sugoi kawaii desune~~ ^-^" is. I'm not bashing people who think Japan is cool. But I do think it's weird how they can't recognize it's just a meme, to the point of fetishization.
[1] https://en.wikipedia.org/wiki/List_of_English_words_of_Chine...
Some languages had periods where they tried finding a new, “pure” name for a foreign concept/item, but these have varied success rates, and often sound idiotic to native speakers, and they don’t see the point.
If something causes friction or pain, but switching away / updating to a new system would incur a high cost or retraining time or break institutional knowledge, there seems to be an extreme bias to adapting to the pain rather than suffering the switching cost temporarily to improve the situation. There does seem to be a benefit though of extreme adaptation that occurs when a process remains unchanged for a very, very long time and actors in the system can develop extreme levels of experience.
The website was crap though, as most Japanese web services are, but that is another topic...
While the "single file" part is not great, using all global variables in an embedded system can be a good thing. I've worked on a critical piece of software for aircraft that did this deliberately, for safety reasons.
One of the most important criteria for that software is determinism, which means no dynamic allocation since it is hard to predict how the allocator would behave. The stack can also be very limited in size, so what remains are global variables.
It also makes debugging easier. You know exactly which variable corresponds to which address in memory. And with a probe, you can know the full state of the program at any time. That's also how we tested "impossible" conditions (for coverage): by overwriting specific memory addresses.
Many of the reasons why globals are bad didn't apply there. We had no concurrency, no resources to share with other processes, in fact, we didn't even have an OS at first, Nothing was dynamic, all arrays were of fixed size, including strings, which were array of characters padded with spaces to the appropriate length.
Of course, we documented usage of these variables, proper up-to-date documentation was not an option as it was certified software. It was one of the most boring job I ever had, and it is probably for the best.
Yes, we avoid dynamic memory allocation but using globals for everything?
Assuming C:
For inter-file access static variables with getters/setters achieves the same thing and makes it easier to split/test
I’ve never went as far as not using local variables in functions but you could use static here too to avoid the possibility of two functions sharing the same variable.
Perhaps what is missing is that the way we did testing was a bit unconventional. That is, we didn't have unit tests. Instead, we did it all through integration testing on the real hardware. And setting these globals was how we got coverage. We also had a particularly good traceability, so we knew exactly which tests to run after each code change and get our coverage back, which made up for the fact that we couldn't tests the functions independently.
It was a rather successful project, we got audited a couple of times with only minor issues. It certainly made testing more difficult, but we in exchange, it was much closer to real life conditions.
Note that it was a while ago, newer project use a more modern approach with a continuous integration platform, unit tests with an emulator, etc... We also use less C, instead using industry specific tools and certified code generators, multitasking RTOSes, etc... With C and ADA for the inner and legacy parts.
(Ironically you can look at say Bosch ECUs and will find much the same: thousands of global variables and lots of "odd" code, but that's because its generated from a dataflow language).
Anecdotally my late model Camry has very thorough floor mat retention mechanisms.
It would be a mistake to conclude this means the software doesn’t work well.