by ayoisaiah on 11/7/24, 4:08 PM with 324 comments
by kippel on 11/13/24, 1:00 PM
- Graduates don't apply to a job, but to a company. The company decides where to place then. The first years are generally a rotation between departments until it is decided where to allocate them. This means that they will be often misplaced in positions for which they don't have the necessary background nor motivation to learn or contribute because in 6 months they'll be somewhere else.
- It is not uncommon for developers not to have a CS or coding background. They learn on the job how to "program" but lack best practices, etc. and figure out things as they go.
- Standing out is frown upon
- People who are good at their jobs are generally rewarded with more work. One can know who the manager's favourite is by who busy the person is. At the end, the good ones end up burned out, over-stressed and brain death.
- Looking busy or hard working is more important than the outcome.
- Combine the above, and there is no point to do a good job: There is no reward other than more pressure to deliver but on the other hand, as long as you look busy, not delivering is not "punished". Clear what option most people will take.
- As already commented, decisions are top down and often very conservative replicating old methods digitally.
- Many customer facing products will try to cover as many cases as possible to avoid complains, perceived discrimination or causing trouble, for example. This results in over bloated software, websites, flyers full of information, etc.
by quanto on 11/12/24, 10:04 PM
It was not an accident that software did well in the most hippy region in the US, San Francisco. On the contrary, hardware development, due to much more constraints from the laws of physics and economics, has been done well in Japan et al as careful top-down planning is the edge, not individual-level agility.*
I am a little surprised that the author, who is active in Japan, is off the mark. I regularly talk to many engineers/entrepreneurs in the region, and the cause-and-effect are quite easy to see and are unanimously agreed upon. Kudos to people there who are trying to change the software development culture for the better.
* Elon's ventures seem to challenge this conventional dichotomy as he attempts to bring both agility and top-down leadership into his firms. More power to him.
by jankcorn on 11/13/24, 2:33 AM
Of course, the attribution of causes to this is highly subjective and I expect every person to come away from the elephant with a different interpretation.
In my case, the very, very top down 'age hierarchy' culture was (and continues unabated) to crush any ideas and proposals that come up from younger and more competent engineers. In the last 30 years with Japan, I have met only a small handful of people that are willing to take input, let alone change direction, from someone younger than them. (a trivial example was a fellow company director of mine that was born 5 _days_ earlier than me. In 4 years working together, not once would he take anything I said seriously. Hmm...)
Give the number of excellent Japanese software engineers that I know, the burden of this "culture" is (to me) quite tragic on its impact slowing down national progress in an important global field. If anyone as ideas how to get around this, I would love to know and learn.
by miki123211 on 11/12/24, 10:13 PM
In Japan, home computers never really made sense until it was far, far too late.
In the west, you'd buy a PC (or a home computer) to play games, edit documents or manage your business. The latter two were pretty much impossible in Japan, as the computers of that era couldn't handle the complexities of the Japanese language and character set. Gaming was all that remained, and if you only wanted gaming, you could just as well get a NES (known in Japan as Famicom), which was much better suited for the purpose.
Computers eventually caught up, but some of the cultural impact remained, still making them less popular than in the west.
This is one of the reasons why Japanese were so good at consumer electronics, they just needed that electronics a lot more than we did, and the devices needed a lot more features, as "just plug it into a computer to do the complicated stuff" wasn't really an option there.
by timoth3y on 11/13/24, 12:56 AM
There are a lot of great comments, but I'd like to collect and respond in bulk to the ones about keiretsu, since there are a lot of misunderstanding about them.
1) It's not the keiretsu, it's X. It's not just the existence of keiretsu in isolation. They keiretsu produced incredible innovations in the 60s and 70s. It was the combination of the keiretsu control and the shift to domestic markets (which the keiretsu also controlled) that killed the PC software industry in its infancy.
2) Amazon, Apple, Facebook, Google are like keirestsu. They really are not. In fact, they operate on a radically different philosophy. Keiretsu would be willing to lose money (and a lot of it) in order to keep business in-house. Their supply chains were owned, integrated, and exclusive. Employees did not leave their keirestu group It fostered innovation as long as they were export facing, but it fell apart once they started to focus on the domestic market.
Thank you again for reading. I'm happy to discuss.
by lmm on 11/13/24, 12:33 AM
Unicode is uniquely worse for Japanese than for any other real-life world language. You can't make an application that displays Japanese correctly using Unicode, unless you implement mumble mumble font selection mumble ranges vaporware, which no-one (except web browsers) actually does. Or you can sacrifice the ability to display Chinese correctly for the sake of displaying Japanese correctly, but no international software maker will do that.
The result is that Japanese software mostly doesn't use Unicode (because it sees no benefit), and, more insidiously, the whole Unicode-first (and, increasingly, unicode-only) world of open-source libraries and languages is much less useful in Japan. So whether by accident or design, Japan is cut off from the global market, in both directions.
by kazinator on 11/12/24, 11:59 PM
The PC software industry was organized around American operating systems that couldn't even display Japanese text without having your path separating backslash turn into a yen symbol.
Pretty much nobody outside of the USA or USA-based multinationals developed a globally relevant PC (no non-PC) software industry.
There are only some rare exceptions to this like SAP (German).
All the American software was developed for English speakers, with internationalization as an afterthought. Not for the global market at all.
You could not take this approach in Japan, like oh, I'm gonna write a word processor for people here in Japan and then we will throw it over the wall to an i18n team to internationalize it and sell it everywhere else.
by Animats on 11/12/24, 9:27 PM
by hnaccountme on 11/13/24, 9:30 AM
Its a cultural problem Japanese cant create software. There is no sense of "Hacking".
For them everything has to be formalized with detailed processors, costing and KPIs. There is a huge bureaucracy to slow down development as much as possible and complain when things get delayed.
They look at the company as a factory that output lines of code, bugs as product defects and treat developers as factory line workers who are insignificant and easily replaceable.
by upfrog on 11/13/24, 1:33 AM
A lot of people in this thread have been mentioning the importance of risk tolerance in Japan's (lack of a) software industry. He gives some good examples of just how omnipresent that risk aversion can be; from getting funding, to renting an apartment, to finding a significant other, running a startup makes your life much more difficult in Japan than in eg the SF Bay. He also gives a bit more context on the matter of overall software quality, and I think that's an important point: writing assembly for small-scale electronics or cars or industrial machines is just as much "software" as writing a modern web app.
Also, while I'm not universally endorsing Japanese web design; dense UIs for the win!
by personalityson on 11/13/24, 8:20 AM
Highlights from the previous thread:
"I experienced this working for the US arm of a Japanese company. To report a bug would cause the programmer to lose face, so we had to waste a lot of time going through all kinds of contortions to lead someone to the bug without calling it out. We wrote a lot of "feature requests" that were really bug reports."
"In a Japanese company, people in general do not speak openly in meetings, because they are afraid of disrupting group harmony. Ideas need to be circulated in a series of one-on-one discussions--this is called "newashi" (https://en.wikipedia.org/wiki/Nemawashi). This means that for a group of N people, it's N*(N-1)/2 private discussions that need to happen. And everyone needs to be in agreement and comfortable that the idea is "right", and that there is nothing the slightest bit off with it. Only after all these discussions have happened and everyone is fully bought-in, there is then a meeting to "rubber stamp" the idea."
"While the risk-adverse and face-losing-adverse traits of the Japanese culture can explain the (in general) slow development and response of Japanese companies (not limited to software), they cannot explain the quirky, often ugly and not user friendly UI of Japanese software. Germans are a bit risk-adverse, too, though not comparable to Japanese, their software, especially enterprise software are showing the same rigid UI and in general difficulty to use. In fact, you can not use them without reading the manual or being trained and that is expected from the end users, too! In a stark contrast, user-oriented software today are very intuitive, offers pleasant onboarding thus every user can use them casually."
"The same holds for Germany. Beside the "no pain no gain" attitude, the pursuit of "perfection" leads to weird outcomes. For example, the music band Kraftwerk dissolved because half of the members wanted to make sounds that looked "perfect" on an oscilloscope and not how good they sounded."
by delichon on 11/12/24, 10:32 PM
by Circlecrypto2 on 11/7/24, 4:21 PM
by araes on 11/12/24, 11:10 PM
by lebuffon on 11/12/24, 9:39 PM
The author explores why with some surprises along the way. I would say it is like the "Mythical Man Month" from the Japanese side of computing.
The book also opened my eyes to the difficulty of mastering Japanese.
https://www.amazon.com/Fifth-Generation-Fallacy-Artificial- Intelligence/dp/019504939X
by ambyra on 11/13/24, 3:21 PM
Asianometry has a good one on this. They’ve become dependent on US software (and hardware), and programming is considered a low prestige job.
by Prickle on 11/12/24, 11:27 PM
A lot of this sounds accurate. Though, I am not sure if any keiretsu had an influence on my current employer.
by guardiangod on 11/12/24, 11:16 PM
From the creator of Windows' kernel:
https://www.windowscentral.com/software-apps/windows-longhor...
>Cutler found what he terms "the worst code he has ever seen," some IME code developed in Japan. He states that the code had no regard for bugs and that it got to a point where they couldn't fix some of the overflow plugs.
In the very same article he said:
>However, progress on this project halted as Windows XP's security had gone from bad to worse. Cutler states that his team alone fixed over 5,000 bugs while turning over some of the system's code.
His team fixed over 5000 mostly security bugs in Windows XP, and he still thinks it's still better than "some IME code developed in Japan". The mind boggles.
by slightwinder on 11/13/24, 4:03 PM
I disagree. Software is very bottom-up-orientated. The culture that developers create is a very strong influence on the industry as a whole. And Japan had for very pragmatic reason a serious lack of this. PC usage in Japan was for a very long time very low. Partly because they are expensive and big, and most Japanese homes are small and have no place for them. And partly because Japan has a very elaborated culture around their language and analog tools, which was hard to transfer to the digital world. This ultimately resulted in highly specialized, small and focused devices, which left no room for people to grow the same deep software-culture that other industry nations had. And this resulted also in kids not being embraced by computers from early on, preventing the growth of a serious foundation for the next generation in the 90s+.
Funny enough, we see the same now happening with younger people in western countries, who also are lacking serious competence in PC-usage, leading to similar effects.
by blueyes on 11/12/24, 11:43 PM
by Tiktaalik on 11/14/24, 8:53 AM
It seems implied in this article that Japan has struggled in consumer or business software of the sort that startups in SV make. It would be interesting to discuss why that is while they became quite successful in games software.
by alephnerd on 11/12/24, 10:47 PM
South Korea and China both adopted the Keiretsu model for conglomerates due to Japan's Flying Geese doctrine, yet both still have fairly robust software scenes.
If I were a betting man, my hunch would be the collapse of domestic financing during the Asian Financial Crisis and Great Financial Crisis.
Japanese asset managers who concentrated on tech like SoftBank, Nomura, and MUFG had better options in Asia (South Korea, China, India) or in North America (USA) to invest in with better returns compared to Japan.
This is why SoftBank has always been a prominent checkwriter in those markets.
by 29athrowaway on 11/12/24, 10:53 PM
My view is different: software is a response to a need. And the Japanese have found ways to solve many of their problems without software. They are OK using a fax and it works excellently for them.
Sometimes software is a solution to a trust problem, or a reliability problem, or a synchronization problem. The Japanese are trustworthy, reliable and punctual in general and do not have those problems.
by langsoul-com on 11/13/24, 7:27 AM
by a1o on 11/13/24, 1:49 AM
No, If you do that you get the satisfaction from the shorter format and won't have fuel to get around to doing the podcasts of the same topic.
by hello_computer on 11/13/24, 3:34 AM
by treflop on 11/12/24, 11:47 PM
It's extremely severe in Japan. Percentage aged 15-64 is now nearing 1950 levels.
And it's only going to get even worse. 12% are young kids today. In 1950, it was over 35%.
by ekusiadadus on 11/13/24, 5:56 AM
However, even though Japan invented the Walkman and early mobile phones, they couldn't keep control of the smartphone market.
This shows that having patents for gaming machines worked well, but patents for music players and phones didn't help Japan stay on top.
by mlinksva on 11/13/24, 5:54 PM
- How Law Made Silicon Valley by Anupam Chander https://scholarlycommons.law.emory.edu/cgi/viewcontent.cgi?a... compares the US, EU, Japan, and South Korea policy including copyright, intermediary, and privacy and US policy was more permissive and very importantly, offered more certainty earlier, to internet entrepreneurs; I don't recall the Japan-specific details but the abstract includes the line "Innovations that might be celebrated in the United States could lead to imprisonment in Japan."
- Technology and the Rise of Great Powers: How Diffusion Shapes Economic Competition by Jeffrey Ding (not open access that I know of) uses Japan/US competition in IT as a case study in its critique of the idea that "leading sectors" determine economic and ultimately geopolitical dominance, roughly the idea that early innovators capture global monopoly profits which ultimately accrue to the state/military; instead it argues what's determinative is diffusion of general purpose technologies throughout an economy. In IT, Japan attempted to translate "leading sector" theory into reality with its 5th generation computing project, a widely known failure. At the same time, it was decades behind the US in both establishing CS as a discipline at top universities and making CS education widely available across state/equivalent universities.
I imagine there's substantial interaction between the above theories and the keiretsu system but I don't recall (which doesn't mean much) anything on that in the above two works.
by PhasmaFelis on 11/12/24, 9:40 PM
Funny how protecting the ideals of progressive democracy from Communism so often involved suppressing those ideals ourselves. Beat the Commies to the punch, I guess.