Tim Romero (author of linked podcast) misses the point. Paul Graham has a far better take on this topic: http://www.paulgraham.com/usa.html
The problem isn't that Japanese companies "missed" the opportunity to do software at X point in time--the problem is that they are culturally Japanese. Japan is about consensus-based design, there is a high penalty for making mistakes, and individuality/creativity is not valued. 出る杭は打たれる - "The nail that sticks up will be hammered down". Certainly you can develop software this way, but the Silicon Valley model of rapid iteration, "if it doesn't work rewrite it", "move fast and break things", etc. gets you there sooner.
But what about Matz and the Ruby core team, you say? Certainly they are a zany, free-thinking bunch. I attended Ruby Kaigi last year and noticed that most Ruby core team members are hired by domestic Japan-focused tech companies, think Cookpad, Tabelog, etc. They are kept for their HR value to attract tech recruits and allowed to spend all their time on whatever open source project they want. Meanwhile the peons work on the actual product.
Source: I've worked in software in Japan for 15 years; 5 at a Japanese bank and 10 as a CTO of my own company. We run circles around our Japanese competitors.
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.
> To report a bug would cause the programmer to lose face
I'm sure there are some bugs that are directly caused by a programmer, but what about all the other inevitable bugs like a vendor upgrading their library, breaking changes, etc?
At least in SWE, the biggest "mistake" here is conflating that bugs and mistakes are the same thing.
To report a bug would cause almost a hara-kiri I can assume. It's a crazy evil society where anybody had to commit hara-kiri. It doesn't help anyone if you kill yourself. If you want to atone for your sins better do it by working harder. Is that a major difference between East and West? Is that a benefit of Christianity that you can be forgiven for your sins, once you admit them?
Being culturally Japanese didn't prevent Japanese companies from achieving dominant positions in electronics, automobile manufacturing, video games, shipbuilding, steel production...
Why is software so special?
I would argue that the kind of risk aversion that holds back software companies is just as evident in most of Europe as it is in Japan. It is also the norm in a bunch of American companies in the tech industry, especially the older ones. And China, which is also often brought up as a society where individuality is not valued in business (due to "Confucianism" and similar tropes), is not held back by similar risk aversion and has a massive software industry that can compete internationally.
Software is different from those other industries cause it's not really a science. No one's figured out a "right way" to make software, and it seems being well-made is a very small factor in the grand scheme of whether a piece of software becomes ubiquitous/successful. So throwing things at a wall and seeing what sticks works better than iterating and perfecting a technique.
Video games are a strange exception I suppose, because they are software. But it seems like above-described culture obsession with not making mistakes doesn't apply so much to the arts.
I think video games are the exception that proves the rule. They are software, yes, but due to their primary use being casual enjoyment (as opposed to productivity), being well-made is important for them -- video games have to run well, look pretty, have few bugs, etc. precisely because there's no other reason to play them other than to enjoy them. Meanwhile productivity software offers something else into addition to being merely well-made.
Have you played Japanese pc ports on steam? I… recommend against it. Most of the ones I own only work under specific conditions if at all. Several are literally unplayable, or require you to downgrade your video drivers to something from 3-5 years ago. The best of them often still drop frames, requiring fan patches to maintain 60fps. Sometimes you fall through floors. They are never patched
The primary quality of what people consider "well-made" for a video game is the artistic part. Sure, nerds will compare graphical fidelity, resolutions, framerates, etc. but most people just care about whether they're enjoying themselves.
Shadow of the Colossus infamously ran pretty badly on its original release platform (PS2). A good amount of the N64's library ran at sub-20fps in certain areas. Bloodborne launched with 60-second load screens. Hell, remember GTA5's multi-minute loading screens that people put up with until someone reverse engineered the code and found it was mostly pointless JSON parsing? Dark Souls was 30fps on PCs for years (without a mod that caused bugs of its own) until the remaster came out.
There's lots of examples of badly-running and/or ugly and/or buggy games that are still regarded as great.
There was a big video game market crash in the USA in the 80's before Nintendo arrived on the scene. The issue was precisely quality (see: E.T. the game)
But again, how is that a Japanese problem specifically, when the same risk aversion exists in corporate culture in other countries (as I mentioned)?
The reality is that only a few countries are dominant in software, and they dominate to such an extent that they draw in software engineers from the rest of the world. That didn't happen because all the other countries have a Japanese business culture... A convincing explanation of why the world is that way needs to be based on traits that are not unique to Japanese culture.
Right when the basic product can not change much it makes sense to improve it and the process to manufacture it gradually. Really Japanese quality movement was much about process quality. You produce the same thing but with a gradually improving process that brings in better quality and better efficiency of production.
But because software is "virtual" it can change-shape in whatever way as long as it does something useful. The software products change rapidly whereas we are still in the beginning of trying to understand what are the best processes for producing it, why because process depends a lot on the product. So if the product changes fast process must change too. And when it changes fast it is difficult to improve it gradually.
Move fast and break things gets you established in the marketplace early and can bring some advantages from that. I have yet to see evidence that it leads to companies that survive for the long haul. At some point, people want products that actually work and don't completely change with a given update.
What allowed Japan to modernize in the 1980's was a huge full society effort (with a lot of government backing) to remove the hierarchism from their work culture. All the great management innovations they started at that time are of the form of bottom-up knowledge transfer and distributed control.
If the software development is done under their "office culture", instead of the "factory culture" that was changed, it would be a very good explanation for why it's not very innovative. But I have no idea if that happens.
Software is special, because in software the blueprint is the building, the process of design and development is one, or rather can be one, but that requires a cultural shift.
Well, maybe they might have been slower to innovate in software, but I think a middle way between their conservatism and the free-wheeling ways of Silicon valley would have served the industry well. Imagine if software engineering was approached in The Toyota Way[1].
As it is now, software engineering us not respected as real engineering because we neglect to apply general engineering principles. We throw away battle-tested solutions and learnings very few years in favor of shiny brittle toys, in the name of "disruption".
I guess the only thing that would change things is something I call some kind of consistent regulation. Too much gate-keeping like in traditional engineering or medical fields can stifle innovation. However, software used by a certain minimum number of users, and software used in certain fields like medicine, transport, finance, need to be subjected by regulatory agencies to certain objective (hence automated?) tests to ensure they meet some minimum reliability, usability, security, and performance requirements. In this regime, small-time SaaS software, or hobby projects, would not need any regulation, but flagship software products of Facebook, Google, SAP, Oracle, Tesla, etc would need it.
The average software company (or software department in a company) in the US does not move fast and break things.
Examples include large older tech companies like IBM and Cisco, defense contractors, and banks.
Even Google's CEO has been talking about how the company has become unproductive and unable to move fast, because that actually matters now that the outlook for the economy is not great.
I find the "it's just Japanese culture" explanations to be weak sauce. Examining the market and political forces at play will usually get you closer to the truth.
The fact is, Japanese companies have taken big, creative, successful risks in many industries. If you look at the developers who work in console gaming, they find ways to push those machines far beyond what they were designed to do. If you've been in Japan for a while, I'm sure you've seen the amazing levels of creativity that exist here in both engineering and the arts.
It's not that Japanese are more risk-averse than Americans. It's that the risk-reward equation is different in Japan. On average, people are behaving rationally in both places.
If you want to change the "culture", change the risk-reward balance. That's why I find it more productive to look at market forces and game theory than attribute things to "Japanese culture:".
Most Japanese didn't historically own desktop computers due to the average Japanese home not having enough square feet to easily accommodate them, which is why they adopted smartphones ten years earlier than every other country except Finland. If they had had a couple more decades to play with computers before software took off as a big industry, maybe the culture would have shifted somewhat.
They could have used gaming consoles with more mainstreamed 'developement kits/setups' instead? That existed at least for some Sony Playstation, or something like this? https://en.wikipedia.org/wiki/X68000
I mean, going farther back, the original Japanese version of the first Nintendo console was called the Famicom, which was short for Family Computer. It had a keyboard accessory, disk system, modem, tape drive add-on, and a port of BASIC.
Well, sure, cell phones took off everywhere in the 90's, and the ones out of Finland were certainly among the more popular. GGP was specifically about smartphones, though.
True. Also, in Japan, you cannot argue with your manager. There is a very strict hierarchy.
Another posible reason, very few people in Japan speak English (at least that was the case when I worked there in the 90s). That might have affected international penetration of their software. Perhaps this is why we don't see much software from Germany either.
Germany is huge in Digital Audio Workstation software. Ableton Live, Bitwig, Steinberg (Cubase and creator of VST plugin format), Native Instruments, Celemony (Melodyne=autotune), and the former EMagic, whom Apple bought to make Logic Pro. Also plugin makers U-He, Synapse Audio, zPlane. Probably more that I’ve missed.
Paul Graham (who is about 4 years older than I am (eek)) and I apparently remember the 1980s and 90s very differently.
At the time, Japan was going to eat our lunch while it simultaneously kicked our butts. Its government-academic-industrial complex was how to do things. It was the technological and industrial leader, the one everyone was chasing. (This was possibly entirely due to the automotive industry, but.... And consumer electronics had something to do with it.)
Then the wheels started to come off. In the software world, the Fifth Generation project came up with...a PROLOG machine. At the same time that Lisp machine companies were going belly up all over the US. And then Japan disappeared entirely. From what I saw and heard, the Japanese software industry became very insular and remote from the rest of the world. And then, the lost decades.
The process described in the article does a better job (and a more relatable one) at explaining what I saw than "they were Japanese", I think. But then, I don't know what the last 20 years have been like, since I haven't needed to follow the Japanese software industry. (Well, there was Ruby, but that was a pretty big mistake.)
How is it that Japanese manufacturing practices made such a deep impact on the Agile software development movement in the West (The New New Product Development Game, Kaizen, Kanban, reduce muda, shu-ha-ri, etc. etc.), yet failed to improve work practices at home?
Kanban as originally used at Toyota was always used to coordinate pre-determined "production line" factory processes. It wasn't used for creative iteration in the way you see it used in a western software development context.
Similarly, kaizen was all about continual factory process optimization, not optimization of a codebase through iterative refactoring. In fact, the mindset is very much "write the code once, never change it, and keep it running it forever basically without upgrades or patches."
"Reduce muda" (we might say "reduce toil") is laughably out-of-place in Japanese software shops. To give an example, an acquaintance worked as "DevOps" at Rakuten. His job was to release new software versions by opening up 200 SSH terminal windows.
Yet, Japan had no issue building a wide range of popular console video games but had far less success in PC games. Language seems like a better explanation as US and Japan had quite different UI constraints in their native language.
Japanese game devs didn't put nearly as much effort into PC games as they did into console games, outside of some early exceptions like the MSX and the Sharp X68000. The main issue was that there really wasn't a domestic market for PC games in Japan, and Japanese game devs focused on serving the domestic market first. So you had half-assed efforts like a single outsourced developer "porting" Mega Man 1 and 3 to DOS (they weren't anything like the real MM 1 and 3). PC ports of most popular console games weren't common until the Xbox 360/PS3 era.
I mean, official PC ports weren't common, but clone games on PC were extremely common. I can't tell you how many PC games* I played as a kid that amounted to "a re-skin of Super Mario Bros" or "basically Donkey Kong", etc etc
*(often free or packed into "1001 games for free with this magazine" disks)
There are often other reasons. For example, in Korea console gaming isn't or wasn't a thing. It's PC gaming. Why? Because to Koreans in the 80s and 90s, playing something made by Japanese, their former occupiers, was unacceptable. Because of that they played games on PCs until that became standard for Korean gaming.
Other reasons PC gaming might not be a thing in Japan. It takes a lot less room for a console (especially NES/SNES/Gamecube) than it does for a PC. You plug the console into your exiting TV. A PC needs it's own space with it's own monitor and desk. Japanese houses and especially apartments are small.
I'm sure there are plenty of other reasons than just that a PC costs more.
Language (i.e. character sets) is not a good explanation for differences in software outcomes. It is possible write Japanese entirely in Hiragana charset (~50 chars) which can be represented in 8-bit alongside ASCII chars, which is how early Japanese computers like the PC-8001 worked. The JIS standard for double-byte representation of kanji also characters was also made in the 1970s. So that doesn't explain why Japan fell way behind in software the 80s, 90s, and 2000s.
Language is more than charger sets. Japanese is often read in vertical columns, then right to left. The space of UI’s that feel natural in both languages are just different.
Not true. Nearly all text in Japanese print media, ads, brochures, websites, video games, etc. is left-to-right horizontal. Only books and some magazines are written in vertical columns. Anyway, this again is not an insurmountable technical hurdle that held back Japan's otherwise burgeoning software industry.
Well, yes, to some degree this does affect localization in the US for software that was originally built in Japan. However, given that the Japanese character sets are more demanding than US character sets, it would usually mean that software made in JP is easier to port to US, and not the other way around. Emoji for example came from Japanese mobile phones ("e" pronounced "ay" means picture, and "moji" means letter)
You misunderstood, the fact that the language does both means the design space is different. Stuff that would seem natural in Japanese would seem very odd in English.
Similarly there are far more people fluent in English so small dev teams can reach a much larger audience without considering translation issues.
My impression is that the technical mess from Unicode(Han unification, nonlinear byte count <-> letter count relationship) and difference in types of words used("parts of speech", whether a noun or verb or a short sentence is used in labels) are larger problem than character sets and reading directions for Japanese translations.
That's a good point that leads to another claim, and I don't think those are exceptions at all; Nintendo and its products is quintessentially Japanese if we didn't focus on the payload content bundled in their software. CEO always in suits and shows fatherly smiles at most, the OS not allowing browsers or sideloading, the corporate having zero hesitation suing modifications on first sight, etc. They are completely in line with the rest if not leading the puck.
And I have an explanation: Japanese education imprints and emphasizes for the whole duration of it that work is strictly for community goals, education is for learning officework punctuality and not for skills, and fun at work is nothing but slacking[0]. Some attribute those to agricultural tribalism; I don't know, but I think these imprinting statements fits behaviors describe in GP and of Nintendo(et al.).
"Move fast and break things" would be just display of laziness, ease of use would be just enablement for saboteurs, iterative design would be just poor planning, under this thinking. UI improvements based on subjective individual opinions would be considered an attitude incompatible with the concept of work, especially for business applications. No wonder that GP runs circles around others when the impediment is in individual successes that products offer users, which the whole culture is united to always look firmly away in all professional settings.
This become less apparent in games and gambling because business goals just happen to align with joy and ease of users, hence Nintendo seems unaffected from consumer standpoint.
0: Such words as "comfort", "efficient", "precise" are used as euphemisms and self deceptions when pursuit of fun or ease cannot be avoided
I think both attempts of explanation are valid. 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 rigid and unfriendliness of Japanese and (German software in some degrees) can IMO only explained through the long isolation of the mass market as Romeo outlined in his podcast.
Hi Johnny - considering there is no one answer here, no one, including you, is missing the point right? If weighting I'd generally put the largest on time-based rather than impact-based seniority, but it's just one of many factors and there isn't really a way to pick any one.
I think if a Japanese person created something like Javascript, it would bring dishonor and shame to himself and his family for 3 generations. Even his grand-children would have to bear the burden.
Therefore, something like Javascript would never emerge from Japan. Only an American would be able to create an abomination like Javascript.
I’m having a hard time squaring the circle of “consensus based design doesn’t work for software but it does work for cars.” By all accounts, here in the US, we regularly bounce software architecture decisions around a room and land on a best design based on inputs from many. Likewise, our cities and cars are founded upon (id even say drowned by) consensus based design.
Maybe I’m misunderstanding the thesis here, but it seems contradictory. I do think PG has a point with an ingrained culture of quality being hugely important. I just don’t think that necessarily contributes to bad software. In fact, I’m quite certain it doesn’t, since companies like Nintendo manage to produce absolutely beautiful pieces of software regularly.
In the US, consensus would be achieved through brainstorming, open debates, etc. and at the end maybe a vote, or maybe the leader just says "let's try it my way, and if it doesn't work we'll do it your way."
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.
Don't know if it still works the same way, but "rubber stamping" used to be real - everybody would stamp a document with their 'hanko' [0] (name stamp). Those who were not completely on board would make their stamp at a slight angle.
What is "bad software"? It is not software that has bugs. It is software that does not do the right thing. Any amount of money and effort invested in SW quality and process quality is simply wasted if the software does not do what its users need it to do.
Now tradition prioritizes quality. Do it the right way, like it always has been done. Tradition is not a good ingredient for software industry. If your motto was write software "like it has always been written" you are missing the boat because there is no need to write the same program again and again.
There's no reason to write the same software again - because then you could just use old existing code instead of retyping the same characters again in your editor. Déjà vu. Groundhog day.
Software allows more gradual consensus-building throughout the product lifecycle, while car companies have to reach consensus earlier and between more stakeholders simultaneously.
Speaking as a Japanese-American, I personally think the biggest problem with Japanese programmers comes down to the Japanese being too crafty and elitist.
What do I mean by this? Basically: If you're a wizard at peeling the skin off an apple with your knife you never need an apple peeler; consequently, you never get a good peeler industry going. Meanwhile, Americans manufacture the best apple peelers because they can't skin an apple with a knife to save their own lives.
Likewise, Japan doesn't /need/ or /want/ the conveniences and luxuries afforded by genuinely good programming. Japanese can and will get by with the hard(er) way and won't care.
Tools exist to address an inconvenience. Good tools exist to make a convenient inconvenience even more convenient. And good toolmakers exist to enable such better conveniences. None of this can exist if there aren't inconveniences in the first place.
This also extends to the hardware side of things, too. Take flash memory, for example. Flash memory was invented in Japan by Toshiba, but nobody cared because Japan was too good. Meanwhile, Intel and the west seized upon the invention and made bank.[1] We all know where Toshiba is now.
This I feel is why Japanese programmers continue to be utterly inept (most still can't produce unicode programs), and why the wider industry in Japan continues to be irrelevant on the world stage from a programming perspective. There is no Japanese equivalent to Microsoft or Google or Amazon because Japan's too good for that.
I'd also note the hereditary love for trade/manufacturing secrets which transformed into adoption of extremely proprietary licensing. Developers prior to ~2015 hardly ever considered publishing their code or even format specs. This secrecy has deep roots and it really annoys and pisses me off. I'm not narrowing down on corporate software - even hobbyists who write "dating games" would never publish code even if it's something ridiculously simple. I can hardly remember an example of FOSS from Japan except for CrystalDiskInfo. But I can tell for sure, when you were using japanese software from before 2010s you could FEEL it, the proprietaniness, the secrecy, the know-howness, the one-of-a-kind-ness, handmad(e)ness :)
I still remember trying that "dark internet" sharing program... SO secret and proprietary, so much security by obscurity - it made be laugh, cry and facepalm.
I've worked for Japanese companies several times in the eighties and nineties and this was definitely a thing. We were an American division and yet we couldn't even get source code or art for the game we were supposed to be making for them. This continued when I worked for a Japanese video card company. They had good programmers but they jealously guarded their code even from other Japanese programmers in the same company and surely they wouldn't give any to us... working at the same freaking company. Another thing about Japanese code that was notable is that in my experience it was always bespoke and custom and quite often spaghetti. It could be brilliant but it had that hand crafted aspect, and was inscrutable outside of the mind who created it. There was no 'computer science' or any attempt to make it engineered - no self conscious thought about how to make software better other than to meet the target and often that target was defined by performance so the end result was no slouch but it wasn't what you would call elegant. This approach becomes very hard to scale. And I believe that is a key element in why Japanese software lagged. Great craftsmen but they did not evolve into engineers but of course I think this is probably changing because my snapshot is old.
I'll also speak to 'elitist' of the top comment. One of the first times I had felt prejudice was when I wrote a very good piece of code for their video card which accelerated some video processing by 4X or more. Seriously they (the Japanese) were surprised and started treating me like a miracle talking dog as they just couldn't seem to get their mind around me being able to figure out their systems and contribute something impactful. Kept asking me - how did you do that?
Alternatively maybe they just didn't think a 4x speedup would be possible? Or their internal culture discouraged people making improvements to code they didn't originally author? Or they thought you did impressive work?
Mostly by eliminating a lot of individual copying steps - also understanding how the texture swizzling worked and writing directly to that format and combining all the steps and writing very good code that used MMX. Often these types of optimizations are things that add speed at the expense of generality but I still think they are useful.
Another example is OpenToonz[0], which is one of the animation tools used by Studio Ghibli and developed in collaboration with Studio Ghibli, another Japanese company called Dwango, and an Italian company called Digital Video. Although, Digital Video does have a premium, paid version called Toonz[1].
Less broadly known than Ruby but the late Itojun[1] and the KAME project[2] also put an inordinate amount of effort into adding IPv6 and IPsec support to the BSDs.
It's never really occurred to me before, but Matz is Mormon, right? I wonder if that informed his decisions differently than if he were heavily influenced by Buddhist or Shintoist thinking, as the larger population does. It would be fascinating to see if there is any correlation between alternative religions and approaches to software in Japan.
Feels like there is a heavy influence of Buddhist thought (mindfulness) on the silicon valley community.
Perhaps people that come from minority spiritual communities have to "think different?"
It's more of a work culture, I think. Remember the swordsmiths, potters, builders, etc., their dedication and know hows. People are used to this and respect this. Noone blames them. And over centuries the culture of masters, schools and trade secrets only hardened. It's not a surprise these practices are hard to let go of.
I'm not sure, I'm someone that has studied a little Buddhism, and lived in Japan, but I don't live in SV and haven't lived in Japan for almost 30 years. I am just pointing out that it is notable that Matz is probably the most well known Japanese software engineer, and he is Mormon, and wondering if that influenced his thinking. And, if "that" was because he had a very different spiritual experience in Japan. Or, perhaps, wondering more generally, if "anyone, anywhere having an non-majority spiritual experience" puts you into a mode of thinking that leads to exciting outcomes, such as inventing an amazing language like Ruby!
I think that's the main reason. Copyright and trade secrets stifle software development. Any development really but it's most visible in software development.
This example is complicated by games also being art and how it can be hard to separate both (even when it's just graphics and audio that you keep on a more closed license).
See how Western games are still overwhelmingly closed source, despite the now known rewards : like Civilization 4 (2005) opening most (but not all) of its C++ (dll) source code, as well as easy to mod XML & Python scripts resulting in a still alive modding scene even today, 18 years later, and after 2.5 (much more restrictive) sequels !
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.
We have to remember that it's the writing, the graphics, the music in Japanese video games that make them great, not necessarily their programming. They are amazing works of art, but unimpressive technical demonstrations.
I disagree, they were often impressive technically but what I feel was at a great cost. They had some great individual programmers but they did not seem to develop technologies that could be shared and amplify and did not value systems or computer science or sharing. Contrast that to say a John Carmack who lifted an entire industry in the west with his thoughtful engineering and complete openness (all ID code was shared and formed the backbone of a lot of other game engines even today). Who is that Japanese Carmack? They had great programmers and great tech but it was hard to scale and they did fall behind. I sense a comeback now. They have swallowed their pride and are using Unreal Engine for example. That's a big step.
The movement code in mario games is better than anything in the west though. But you could call that type of code "art" as well, it isn't really a software engineering feat. Maybe you could say that Japan has great software artists, but they aren't that great at software engineering.
Can you expound upon this? I've not heard this talking point before. I do know Nintendo has a history of pushing the technical limits of their own hardware, but I'd be interested to read more about some specifics.
When you only care about how something looks and plays but care nothing about how it happens under the hood you can end up with some really ugly code. I am not claiming this doesn't happen in the west either but in general when I got a glimpse of Japanese game code it was 'ugly' compared to western even if it technically was great (fun, performant etc)
I would call it "good enough" rather than perfect.
The usual meaning for "perfect is the enemy of good" I've heard is the danger of spending forever to perfect something instead of releasing and learning from experience.
That is a really interesting perspective, thank you for sharing!
The idea of Japanese people honing a skill/job to perfection is definitely a stereotype where I am from. Perhaps it might be a justified stereotype which ties into this. Definitely the apple peeler example rings true.
It is justified. That's why Toyota/Lexus is superior to all other brands in terms of fit and finish, and quality.
I can literally see the difference in stitching quality between Lexus and any other brand vehicles. Even poor Porsche gets brutally mogged by the LC500.
My favorite is the lathe type apple parer (peeler corer slicer). Probably an overkill for eating an apple, but perfect for cooking a few. Invented by David Goodall in 1864! Not because Americans are bad at peeling but because people made all sorts of things with apples and because of “yankee ingenuity”. The first apple parer patent was granted in 1803!
https://shop.kingarthurbaking.com//items/apple-peeler-corer-...
Ha! It’s been a while since I’ve seen that type of peeler. Makes sense that it was popular in the 50/60s my grandparents had one. I recently bought a peeler that I don’t like and I’ll try the Jonas one.
Touching apple skin with my teeth gives me the same sensation as fingernails on a chalkboard. I don't have this issue with other fruits, just apples for some reason.
I feel that based off my experiences using and interacting with Japanese software. Very frequently, they exhibit exotic bugs and quirks that make me sincerely question the aptitude of the programmers involved.
Maybe the programmers in question are smart as in book smart, but they clearly don't know how to reflect that in practice.
Also, it truly is ridiculous that so many Japanese programs still can't unicode.
I think starting with a certain complexity a polished product is more of an execution issue than it is of quality of programming.
In my experience, many programmers that are book smart can contribute to great working software, but can also produce crap when left to their own devices ("who cares about UX, look at this slick 3NF data model I fleshed out") or managed incompetently ("We fixed all the bugs that were in our bug tracker, therefore the software must be complete").
One can be competent and still suck. The thing with software is that it's quite easy to reach mastery in some very specific domain, while lacking any competence in other important areas.
That's my story with the dish washer: I find it is a lot easier just wash dish as soon as it is used, it does not take more than a few seconds for a plate, and my dishwasher are only used as a place to dry plates.
I had thought about the same concept you mentioned here when comparing eastern and western culture, and I thought it might come from diet, but I didn't backup with any research.
Dishwashers make sense if you have a lot of dishes, and make a practice of filling the dishwasher before washing. For a single person, this can take many days. The total amount of water and time ends up being less by using the dishwasher.
However, if you have a small number of dishes so that you can't afford to wait 3-7 days between washings, and need to clean them daily, the dishwasher makes no sense.
Are people really leaving dirty dishes in the dishwasher for a week? That sounds like it would smell terrible and anything left on the earlier plates would be dry as cement by then.
I'm not sure what you leave on your dishes, but generally it's just a few crumbs, maybe a bit of butter, if it's obviously too much for the dishwasher - oatmeal does get as hard as cement for example - you need to rinse it off first anyway.
Dishes have been known to sit in my house for 3-5 days depending on how many we use. The only reason we don't run it every day is to save on washing pods.
Ideally, I would probably run it every day right before bed and then unload it first thing in the morning. As is, I forget, then there are too many dishes to fit in the machine and I need to wash them by hand to cook with immediately.
I know I'm a little late replying here, but I wanted to follow up anyway. I have seen this video before and used the powder when I lived in the US. Now that I've moved to the Netherlands in 2022, I cannot find the powder easily in stores. I don't really want to order it online as it's so cheap I might spend more on shipping than on the product itself. The pods work "good enough" for our purposes.
Maybe the appliances sold here are also better optimized for pods? I do have a brand new dishwasher so that could be a factor as well.
My family will not start the dishwasher until it is completely full or we have otherwise run completely out of clean dishes of some category that they want to use. Tends to stink quite badly eventually, yes. I have failed utterly to persuade them to do otherwise, so mostly handwash my own dishes and let them do their thing.
Are you putting dishes directly from eating into the dishwasher? Most people rinse it off, especially anything that could dry out. The dishwasher is more to automate the application of soap and water, not to literally remove every piece of dried food from a dish.
It's highly dependent on the dishwasher you have then. Maybe he has a more powerful one because mine often does not get food that's stuck onto the plate and it just have to wash it again. I'd rather rinse it before putting it in than after, then.
I choose to wash it immediately after use, either cooking, or after eating. I just find it easy to do. If I keep dishes in the sink, it will accumulate, then I really find no will to wash it. I have thought about why I keep doing this, is it something to do with diet, so I am mentally lazy, but physically not XD
Another thing is I don't like the smell of washing liquid very much, I choose not to use it at all, even when washing dishes with hands
If it's any comfort: Switzerland invented* the quartz watch. But swiss watchmakers were too good for that.
* at least according to how swiss tell this story. https://en.wikipedia.org/wiki/Quartz_clock#History differs substantially. (but does mention Longines ca. 1966, which may have been how the story I heard arose? In any case, these days Longines has long since been bought by Swatch...)
In the case of Kodak (my father used by an engineer there and his neighbor at time Kodak also in charge of digital cameras and was warning about digital cameras in the 1970s) and other incumbents like this ,for the current Csuite there is NO incentive to cannabilize the product and take risks and lose revenues (think film, processing, etc) to prepare for the future.
Kodak with digital cameras.
Nokia with smart phones.
etc.
Absentee leadership with no understanding or care to monetize it - including the pricing on the first Xerox Computers.
Also many deca billion dollar spinouts from Xerox Parc:
1. SGI / silcon graphisc based on Jim Clarke's work on 3D chips.
2. 3Com / ethernet based on Bob Metcalfe's work.
3. Adobe based on John Warnock and Charles Geschke on vector graphics and printing.
etc.
> Absentee leadership with no understanding or care to monetize it
Its not that. Its that pushing digital photography and other stuff and monetizing them would kill their existing business. However you look at it, its a risk.
Hindsight is 20/20 of course, but I feel like it should have been fairly obvious that computers were the future, and their existing business (photocopiers) wasn't exactly going to die, but it was going to become less important as digital documents because normal. We still have photocopiers today, after all, though now they're integrated with digital printers and scanners: people still use paper documents, just in different ways than merely copying them. Someone more far-sighted should have seen this, and seen the value of their PARC work. Obviously, someone there at Xerox did, or else why did PARC exist in the first place? So somehow, they managed to have the vision to enable spending tons of money on this research center and the brilliant minds there, but then the higher-ups were too stupid to actually do anything with all that research. It really doesn't make sense to me.
> Someone more far-sighted should have seen this, and seen the value of their PARC work
Even if someone sees that and is somehow able to sell it to the board, that would also need to be sold to the shareholders. And there comes the real challenge. Its not easy to go to shareholders and say that you are going to start a new product line that will kill the profits of another, but ask the shareholders to trust you, because this is going to be the 'thing' of the future. On one side billions of dollars of investments in stocks, backed by a gigantic market held by a flagship product. On the other side predictions of future.
The investors and even the markets wouldnt like that. In a sense, capitalism hampers innovation by providing the wrong incentives.
One of my favorite pieces of software, EmEditor, is made by a company that was started [0] 1995 in Japan. Though maybe it’s telling that they’ve been headquartered in the USA since 2001 ;)
Well, not exactly sarcasm but a snarky way to say that Ruby was written by a Japanese programmer, so he can't be completely incompetent.
Perl is my preference of the three (Perl, Python, Ruby). I agree with 'There Is More Than One Way To Do It', and disagree with 'There Is Only One Way To Do It'. And Perl has the best support for regular expressions. Ruby is OK, but I learned Perl first, so it feels more natural.
I don't know exactly why, but whenever I write something in Python I get extremely irritated. I'm always worried that I'm going to clobber the namespace somehow. Some functions like 'str' are extremely short and common, and happens to be a variable I always use for temporary variables. The forced indentation messes up the code when I want to debug something by commenting out a bunch of lines.
But sometimes there is a builtin module that makes something easier to do, so I am faced with a 'Kobayashi Maru', a no-win situation
I've never used Nim or Crystal. Rust is too complicated for me to bootstrap and not particularly compatible with older architectures as far as I remember (I believe had trouble with compiling 32-bit)
FWIW, Nim has various trade offs for you. It has an alternative comment syntax that goes like
echo 1 #[ deactivated ]#, 2
While `from modA import nil` does exist, it is not very idiomatic. Idiomatic is more managing the global namespace with the type system/overload resolution which it sounds like you would hate. However, shadowing outer scope names with inner scope names (like your `str` example) *is* considered idiomatic.
I enjoyed reading this. Thank you. Specifically how you beautifully put “Japan is too good for that”, in a way that both compliments and criticizes at the same time.
IMO the Japanese failed at Software Development because they never thought of it as distinct/different from and more important than Hardware that is used to run it. It requires a certain degree of forward abstract thinking which they unfortunately missed. They then compounded this mistake by staying on it for decades.
As an example, in the late 90's i was part of a group developing a Office Software Suite for a Japanese Company. The specification they gave us was very clearly done by some Hardware Engineer who had just learned OOD/OOP. All classes were named like hardware part nos. eg. MDK_C1001, MDK_C1002 etc. All methods were declared pure virtual in C++. We had to convince them to throw out their specifications, sit with us and re-design it together (and implemented using Visual C++/MFC framework).
I think this is on the right track, Japan has a fantastic hardware oriented manufacturing culture, but this can be a serious impediment when it comes to approaching software development. One symptom is that software teams are seen as part of a product line team rather than having their own org.
Famously this was a major issue at Nokia and Blackberry, who institutionally believed that their competitive edge was in the hardware. Device specifications would change late in the product development cycle due to constraints on manufacturing or cost targets, and the software devs would just have to work around it. Adopting software platform, like Symbian at Nokia, was primarily seen as a cost and resource saving move to speed up development using a generic solution.
Right; when you are good at one thing, you tend to look at everything else through that framework only. If you then have a strict Hierarchical Organization/Society (i.e. thou shalt not question) which enforces that thought pattern, you miss out on "the next big thing".
The success of American Software Companies is a direct result of limited structure, controlled chaos, lots of money to experiment with and loose financial controls pioneered by DARPA and transplanted to Silicon Valley.
This particular point is addressed directly in that the ready access to money made Japanese corporations stagnant. Why innovate if you can continue to get rich doing what you already do?
This is the current failure mode of many automotive companies. As institutions they see software as something you sprinkle on top of hardware platforms using a process optimized for hardware.
It's interesting how that ties into some of the famous missteps from Japanese video game companies in the past.
Nintendo went overconfidently into the 5th console generation with the cartridge-based Nintendo 64, allowing Sony to step in out of nowhere and eat their lunch since Nintendo alienated developers who obviously wanted the format where they weren't bound to the small storage sizes on cartridges. Including their bread-and-butter partnerships like Capcom with their Mega Man games and Square with the Final Fantasy series, both defecting to make PlayStation games.
Albeit their lunch was eaten by another Japanese company, but Sony made an overconfident hardware misstep themselves a couple generations later after feeling invincible with the success of the PS2, releasing the PS3 which was technically impressive from a hardware standpoint, but difficult to develop for, causing them to alienate developers (see Gabe Newell's famous comments) and lose steam for a lot of that generation.
Bit of history many people are not aware of - in the 1980s, Japan developed its own indigenous operating system, TRON. Towards the end of that decade, the Japanese government wanted to adopt it as a national standard. The US government started threatening trade sanctions, which gave Japan cold feet. (Some accused Microsoft of being behind it, although others insist that’s untrue.) How would have the Japanese software industry turned out without US threats, or if Japan had ignored them?
MSX ran the MSX-DOS operating system - which was developed by Tim Paterson under contract to Microsoft. It was a port of MS-DOS 1.x to the Z80. ASCII in Japan wrote the device drivers but the core OS was developed in Seattle.
Please correct me if I'm wrong, but your comment seems to imply that the US directly intervened/prevented the Japanese from making their own operating system.
Curious, I read the wiki page and if I'm reading it correctly, at the time DOS was already dominant in Japan and TRON/BTRON being incompatible with it made it economically unviable, which is the main reason why it was cancelled as a national standard.
> the Office of the U.S. Trade Representative issued a preliminary report accusing BTRON which only functioned in Japan of being a trade barrier and asked the government not to make it standard in schools.[12][13] TRON was included along with rice, semiconductors, and telecommunications equipment in an April 1989 list of items targeted by Super-301 (complete stop of import based on section 301 of the Omnibus Trade and Competitiveness Act of 1988).
> According to a report from The Wall Street Journal, in 1989 US officials feared that TRON could undercut American dominance in computers, but that in the end PC software and chips based on the TRON technology proved no match for Windows and Intel's processors as a global standard.[14] In the 1980s Microsoft had at least once lobbied Washington about TRON until backing off, but Ken Sakamura himself believed Microsoft wasn't the impetus behind the Super-301 listing in 1989.[15] Known for his off the cuff remarks, in 2004 governor of Tokyo Shintaro Ishihara mentioned in his column post concerning international trade policy that TRON was dropped because Carla Anderson Hills had threatened Ryutaro Hashimoto over it.[16][17]
It seems clear the US government did pressure Japan to kill the “BTRON in schools” project. It also seems clear that the project was facing its own internal obstacles (like the DOS compatibility issues you mention), quite apart from any US government pressure. The respective weight of those two factors in leading to its demise is a matter of opinion.
> How would have the Japanese software industry turned out without US threats, or if Japan had ignored them?
Good question. The same question can be asked about the defense industries of the UK, Germany and many other NATO countries. In some cases, the US killed their industries through threats and bribery that were later revealed as scandals.
> But the rest of the word was moving in a different direction. The rest of the world was moving away from dedicated hardware and towards innovative software running on standard hardware platforms.
This piece is about Japan vs the world, but looking at countries other than the US I'm not sure this narrative works.
For instance did the UK become a Software behemoth ? France ? Germany ? Brazil ? India ? Perhaps each country has their own failure story, but if we see the same result accross the board, are the individual specificities significant ?
Most advanced country have seen a few successes and has a decently sized software development power, but not that much bigger than Japan, and certainly not to the scale of the US. The real story would be to me "the US ate everyone's lunch"
Bull for instance. Ubisoft and a bazillion other video game company. Alcatel / Lucent / Thales for industrial stuff.
For the deepness of my ass I would say failure was tree fold :
- France like a good old top down approach. That … mostly don’t work for software.
- France decided to listen to our American friends and sell the familly heirloom to the market. ( I’m bitter on this one )
- France like to do it the “right” way, and that is, of course, the French way. That lead us to build our own internet in the 80’s ( see : minitel ) ( see : point 1 )
By that definition, Japan has its own software companies too. They are also big in the gaming world. I didn't check the numbers, but I'd bet it's bigger than France.
The US has won the Software (ie: not the games, but the core Software market) because it had pocket deep VCs that understood tech. Everything else is secondary. This is of course, my opinion.
Yes, even in games this is the case. Market leadership in gaming followed market investment. Japan had a large manufacturing presence in games from early on which propelled the corresponding software industry. And the US was pioneering in funding pure software businesses - gaming became part of Silicon Valley's VC pool in the 80's. In other countries, they faced difficulty raising the capital needed to grow as productions got bigger in the 1990's, and mostly got absorbed by US investment, either through acquisition or talent poaching. Ubi was a rare exception to this trend.
I'm pretty sure a story similar to this was playing out throughout the software world.
Yes. Also France made a real big effort to protect its agriculture, wine and luxury industry.
LVHM grew and flourished, but there was no love for the software industry. I kinda think there was no alternative scenario where it would have worked out, culturally most people really don't give a damn.
Also having a smaller internal market makes it harder to bootstrap new ideas to a decent size before getting it out. Even now successful startups might target France and Spain, but going beyond seems to be a very hard sell while the EU is supposed to make it easier in theory.
I’ve seen the market thing play out first hand : my former company was dominant in France ( in a tiny niche market )
But that was bringing just enough to get by. Then we expanded to Italy and Germany with great effort. Each country was a whole new market with different regulations, customer expectations and so on.
We ended up spreading ourself very thin.
In insight, we also made mistakes, but that was not helping that each next 500 deployment was a large, hard to quantify effort.
> - France like to do it the “right” way, and that is, of course, the French way.
Having a national identity seems to be a double-edged sword. Brazil's soccer team, for example, is expected to win the Brazil way, dance, samba, which even for professionals may be a subconscious distraction from seeing a particular game situation clearly.
All this talk about why the U.S. is better at producing software could someday turn into the same burden.
> For instance did the UK become a Software behemoth ? France ? Germany ? Brazil ? India ?
UK, France and Germany all are pretty big in software, as also in gaming. Japan is also big in gaming, but they started with consoles and hardware. Only in the recent decade became their software-side better, but it's still kinda crappy to be honest.
Brazil and India, were they not to poor to made it big in the early IT-years? Even today they are busy with catching up. Though, India has found their stance in IT at this point, at least as contractors.
India has been a significant force in IT since at least the 1990s. Look at Infosys, Wipro, Tata, etc - all huge multinationals that give the IBMs, HPs, and Accenture's of the world a run for their money.
Japanese mobile phones kicked the ever lovin' crap out of western phones for a very long time. The Japanese got into social media way before the West did. And Yahoo Japan actually still exists and seems to be doing okay.
And that's not to mention the success of Japanese video games.
The big problem that Japanese software has is simply that its market is too much smaller than anything which caters to English. This is why video games can do so well--language is of secondary importance.
To me the main difference is they failed to push an operating system.
As you say, they were really ahead in mobile, and docomo had an honest try to push imode in europe for instance, but it couldn't take off as the markets were too different.
The iPhone not supporting Felica gave the maker a small bubble to thrive but android was too far behind. Now they are barely surviving.
I'm probably misreading you a little bit, but here's my two yen.
There were a bunch of silly operating systems in the history of computing, and Japan had a few too. TRON was pointed out in another thread, which still exists in some embedded stuff, even in modern stuff you may have used before, for example the Nintendo Switch joy-cons.
But TBH, there can only be a handful of operating systems that actually matter. So you'd have to outpush some extremely pushy companies in a very tight space, and the market for commercial operating systems is almost dead anyway.
>> Most advanced country have seen a few successes and has a decently sized software development power, but not that much bigger than Japan, and certainly not to the scale of the US. The real story would be to me "the US ate everyone's lunch"
They did. By virtue of being the only advanced economy that didn't need to focus on rebuilding after WW2. They have been capitalizing -rather successfully- on that breathing room since then -be it for the space race or software industry- and the advantage has only grown.
In the 2020s I largely invoice in USD, despite being a european immigrant, because almost no one else pays.
(even if there weren't a substantial rate difference: it is far less painful to stay awake until Pacific Time to coordinate with people who grok remote development than to attempt to deal with people in my time zone who do not)
The euro-america centric view in claiming that India failed in software is simply absurd. India has giant software companies powering one of the most advanced governance, finance and business societies and not to mention a major destination for the west to outsource software jobs!
We're not talking about straight failure. There's giant companies all over the world and sure they'll produce software.
But the FAANG/MANGA/Whatever acronym for the five/six near monopolies aren't from India. If I'm in Korea and tomorrow I want to use a cloud provider, I won't choose an Indian one short of wanting to market to India. But chances are it will be an US one. That's the gap I am pointing at.
I see where you're coming from but that is a very narrow view only taking into account B2C service type software, and that too for a sample size of you. Not to mention the fact that FAANG hires a ton of Indians
Is Cloud Services B2C ? or computer OSes ? Do your Indian giants give their sales team laptops with an Indian OS managed through an Indian Directory system backed by an Indian Cloud, and make them build spreadsheets and presentations on an Indian Office suite ?
I know there are serious efforts to cover all this ground, but have they landed ?
I’d actually see China landing first on that scenario, and while the future is too murky to say anything, if they ever get through the tunnel they’re in now, they’ll have rebuilt the whole stack literally down to CPU technology.
The phrasing in the previous comment makes me think that you are talking about a B2C aspect. The point in trying to make is that software industry is much larger than the specific examples you mentioned, which are also very "English speaking user facing" for lack of a better word
> For instance did the UK become a Software behemoth ? France ? Germany ? Brazil ? India ?
Other than Brazil, all of these countries have dozens of tech unicorn startups, and numerous established success stories. The UK has nearly 50, India has over 100. Germany and France have somewhat fewer. But those countries may have a more cautious VC style, with less American influence.
In Japan's case, they kicked the ball into their own net 25 times.
Because of the Keiretsu market power coupled with Japanese culture (social status, financial risks, career risk, labor mobility rates and bankruptcy laws) very few startups existed, especially in this space even though the capital requirements are far lower.
This left all the market power in hands of Sony, Fujitsu, Toshiba, Sumitomo, etc about 30 companies. And in the case of the major hardware vendors they also owned content libraries and firms eg - sony / sony music, bmi, toshiba warner brothers etc. they were extremely protective against piracy and especially digital piracy - this killed them.
We see the same in US, look how much innovation happened in virtually unregulated tech and software but not highly regulated Detroit autos - DOT, DOT, NLRB, OSHA, etc.
This is a weird point to do here, but in my opinion startups aren't the end all be all of innovation: they are needed in a system where incumbents are mostly static, but in different configurations the role of a startup can be accomplished inside behemoths or through joint ventures, or other settings.
To back my point, the Walkman didn't come from two guys in a garage. The GUI wasn't invented by Apple, linux wasn't born out of VC investment, NFC payments weren't brought to market by drop out college hippies.
> For instance did the UK become a Software behemoth ?
Scotland did. We all know that the video games industry just exploded in Dundee about 20 years ago, but the whole area has a rich software development industry across a lot of fields.
One reason games became such a big thing in Dundee was because that's where the Timex factory where they built the ZX Spectrum was. And because that's where the factory was, everyone had one because everyone's dad knew someone who could "get" one, bypassing the usual supply chain hassles that Sinclair had at the time.
> because that's where the Timex factory where they built the ZX Spectrum was
You could say that, but there weren't any problems getting ZX Spectrums in Aberdeen, Glasgow, Edinburgh either, and the whole UK was awash with bedroom programmers making piles of games.
What I think actually put Dundee on the map was David Jones, attending Dundee Institue of Technology (now know as Abertay University) in 1987, employing his friends Russell Kay, Steve Hammond and Mike Dailly to form the company DMA Design. They had a series of successful games on the Amiga, releasing the worldwide smash-hit Lemmings in 1991. This encouraged Abertay to offer the world's first computer games degree course in 1997. You can see how large an employer DMA Design was in 1996 in this interview [1] where they were working a little-known game called Grand Theft Auto they released in 1998...
I'd put the case that DMA Design and Abertay kickstarted the Dundee games industry themselves, they had a lot more influence than the Timex factory. Bathgate had the Motorola factory, Greenock had IBM, that didn't turn them into software hubs.
Neither Bathgate nor IBM made inexpensive home computers that could be picked out of the reject pile, sneaked out under a jacket, repaired, and sold down the pub for 100 quid.
The thing is that while a lot of people throughout the UK, by Christmas '83 literally everyone in Dundee had a ZX Spectrum. Probably half the reason they had difficulty fulfilling orders is that so many "minor faults" ones on the lines ended up getting sold in pubs under the table, that would otherwise have been reworked and packaged for retail.
And I mean, literally *everyone*. If you wrote a game and you gave copies to your mates at school, by the end of the week everyone would be playing it, particularly if it was good but even if it wasn't, because it was *new* and you hadn't played it a million times.
I don't know if this is specifically your childhood experience, but you're describing also my childhood too, and I didn't grow up in Dundee. There were plenty of first- and second-hand ZX Spectrums to go around, VIC-20s, C16s, C64s, Amstrad CPCs, and the schools had BBCs. The UK went home-computer mad. Remember when there were type-in listings in C+VG? The Usbourne computer books? [1] INPUT magazine? [3] Kids were swapping tapes like crazy, and writing their own demos and games, all over the country.
While Timex may have played a part and the density of Speccies may have been higher in Dundee, I still think Timex's main contribution to Dundee's games industry was to make David Jones redundant, giving him the money to buy an Amiga 1000 [3]
An anecdote from the period described: I was working on a product which had hit an unknown bug on the japanese localised Windows, so our partner sent an engineer with a bunch of install disks to visit me.
One of the first things he said was: look, this is the difference between the US and JP — you have three computers on your desk (he was counting the ICE), but I have to share "my" computer with two other developers.
At the time, I took it for self-deprecating exaggeration. After reading this article, I fear he had probably just been making a factual observation.
Milage vary, my own anecdote: until we got Macs (OSX) as dev machines, every companies I worked for in Japan allowed us to buy two individual machines: the standard Windows one (just like every other employee) and a linux "server" for our dev needs we'd ssh into.
And they usually looks and works like a Borg tentacles and assimilated production hardware, despite the "emulator" nomenclature, apparently by tradition from very early days of microprocessors when ICEs actually were alternate implementations for chips it emulates
Hi. I'm Tim, the author. I'm delighted to see the interest in this subject on Hacker News and happy to answer any questions.
To group and respond to a few common comments:
> 1) Is this unique to Japan? Most other countries have not developed world-dominating software industries.
That is true, but most other countries did not enter the PC era after spending several disrupting and dominating multiple global industries. So, expectations of Japan were much higher.
> 2. Was cloud computing really that pivotal?
Hell, yes! Being able to prototype and get a handful of paying beta customers for a few thousand dollars changed everything. It totally short-circuited the VC "experts" who would say "no one will ever pay for that".
It was a big ambitious well backed plan to pursue the silly notion (as if!!) of massively parallel computing with unheard of numbers of cores to jump start AI networks etc.
Are there any stories there on how that all fizzled out and came to naught back in the day?
(a) cf https://news.ycombinator.com/item?id=26953443 for one aspect. It didn't do much good to have massively parallelisable logic programs without any equally massively parallel hardware on which to run them. I ran jobs on top500 machines back then, and even there coordination overheads would kill anything too fine-grained.
(b) it took a lot longer for Moore's law to end for traditional ISAs than we had been fearing, back in the 1980s. Like many things in tech, maybe Fifth Generation had just been Too Early.
(a) is the nub of the question - the spiel from Japan at the time was that they had massively parallel hardware in the works that was on the cusp of delivery...
At the time I was tinkering about with a bunch of transputer boards, modding Minux, thinking about capabilities, and mangling Occam (the language) into something less awful (IMHO).
I was more curious about how this decade (roughly) long chapter in Japan's IT history didn't rate a mention in the authors article (aside from "Japan plowed ahead with hardware in the 80's" (IIRC) throw aways)
1) Could you share examples of bad, good, and open-source Japanese software? Outside of the gaming industry and except for Ruby.
2) Does they have some kind of the hacker culture?
3) May you know links which answer question: Why Europe/Brazil/Turkey/Russia/etc failed to build well-known and international software companies, FAANG-like.
> 1) Could you share examples of bad, good, and open-source Japanese software Outside of the gaming industry and except for Ruby.
"Outside of Gaming and Ruby" cuts out a lot. There is plenty of very good software, but not much that has made it outside of Japan. Nulab has made some cool stuff. I love what Mui Labs are doing with UI.
> 2) Does they have some kind of the hacker culture?
We are getting there. I think there are almost two "hacker cultures" in Japan. There are developers who model their lives on the Silicon Vally lifestyle, and there are people who just love programming. The overlap is not as large as you might imagine.
> 3) May you know links which answer question: Why Europe/Brazil/Turkey/Russia/etc failed to build well-known and international software companies, FAANG-like.
I'm not an expert in those counties, so I don't have a great deal of insight. During the dot-com book, companies like SAP and Nokia did great, but this time around the US has certainly been the winner in this winner-take-all marketplace.
>> 1) Is this unique to Japan? Most other countries have not developed world-dominating software industries. That is true, but most other countries did not enter the PC era after spending several disrupting and dominating multiple global industries. So, expectations of Japan were much higher.
Indeed. Read any US computer magazine c. 1980 +- four years, and the one assumption/expectation in the background of every article discussing the industry is that the "Japanese are coming". This belief directly caused several things to occur in the US industry, notably Commodore's pivot from a seller of high-priced desktops sold to schools and European companies to a focus on low-priced home computers starting with the Commodore 64.
Tim, as a Mombusho exchange student in Kanazawa (and Rotary before that in Sapporo) your post brought up some great memories of my time in Japan in the 90s.
Do you have an email addres? I want to ask you some questions regarding my startup approach with a big Japanese company we are engaging with.
I don't disagree with much of the story, but there are other factors too. A big one is that due to the complexity of the Japanese writing system, PCs never went mainstream: the older generations survived with handwriting, faxes and dedicated word processors, and the newer ones jumped right to mobile phones, meaning the PC hacking generation of the 70s and 80s that kicked off the dot-com boom was much smaller than in the West. Plus you had the usual Galapagos problem of Japanese computers being incompatible with the rest of the world (MSX, PC-98, DOS/V etc).
I'm also much less optimistic than the author about the future of the Japanese software industry. Mobile dev in particular is massive across Asia and Japan has no competitive advantage over (say) China, Vietnam, Korea, Indonesia etc here.
I don't think this is a big barrier, except for the fact that Western computers didn't have good support for Japanese from the beginning. Earlier, heavier adoption of PCs might have changed the course of history, but at the time it would have been difficult to choose a 286 with DOS vs one of the Japanese computers.
What is your source or personal experience for that? I don't think the typing system has any correlation. In fact you can actually type pretty fast with modern keyboard software, there are lots of shortcuts.
I type in Chinese with friends and family everyday and so do my kids. Hardly anyone uses voice messages. I find leaving voice messages on messaging systems more popular with Spanish speakers than Chinese speakers, and Spanish is one of the easy languages to write for native speakers.
I'm not so sure. Have you watched them type on computers and phones? It's a bit more complicated because you need an IME, sure, but it doesn't seem to make that much of a difference. I guess it's tempting to draw that conclusion, but I think it's much more related to culture. Even among European countries, and within countries but between age groups, the usage of voice vs text is pretty different.
It's a common misconception that Chinese and Japanese spoken languages share roots, it's just the script exported from ancient China. Since IME is more tied to grammar than to script I am not sure how much relevant will that be...
I worked in Japan for ten years in tech and these points jumped out at me:
>Software development was an exercise in box checking. You implemented a feature once the customer had asked for it and the contracts had been signed.
I recall a meeting where we were trying to set some requirements for a product based on what our potential customers would need. It was a requirement that all customers would need, not some unique thing that would probably change based on each customer. I understand pushing back against scope creep but as an analogy, we were building a PC designed to put our customer's peripherals inside. But we decided not to add a PCI slot even though the potential customers made PCI cards. By the end of the meeting it was agreed that since there was no signed-up customer yet, there could be no requirement. Our eventual customers of course needed that PCI slot to be added mid-design.
In general it takes a lot of consensus building to add a requirement in the first place, and once it's there, it's seen as the exact and maximum. Going beyond is seen as only adding risk. It's very difficult to ask for effort to optimise something within a given constraint. There's good and bad things about this.
>There was no real career path in software development. I mean, maybe you could move up into project management or over into sales, but if you were still actually writing code when you were 30, people kind of wondered what went wrong.
It was common to interview senior software engineers with a good resume, but find out they were one of two types in the interview. Type 1: a middle manager that has done nothing in the last ten years aside from passing excel files back and forth between junior software engineers and other managers. Type 2: considers "software design" to be making block diagrams in Visio (or worse, drawing things in excel as if it were Visio), and considers writing software to be beneath him. This type usually plans to outsource the actual development work
I disagree with everything in this article post-2000. There were no "software gatekeepers" preventing Japanese software from flourishing on the PC or Mac. I'm not sure what the author is talking about there.
So there's still the unexplained question of why Japanese software is still bad, in 2023. And I think the answer is the language barrier. Windows and Mac dominate the world's computer systems, and the development tools for those environments are English-centric. The translations are lacking or no existent. The struggle of a Japanese developer to make an app is greater than that of a primarily English-speaking developer.
Also, outside of USA/China/India, software developers are still not given much respect or seen as equal to say, mechanical engineers or electrical engineers. So there's little reason for Japanese students to major in software.
> So there's still the unexplained question of why Japanese software is still bad, in 2023.
I think its to do with their protectivism culture. My previous job was doing digital distribution of advertising content around the world. The Japan market was difficult to move because they could pay someone to ride a bike down the road to deliver an advert. And they activately avoided moving digital as it could put alot of people out of work.
It doesn't serve any real purpose at all, but what it did was move a bunch of tape jobs into digital jobs. But it made what should be a simple process into a complicated process.
Anecdotal: in one of the Japanese companies I've worked at, we had to do a Japanese-speaking-only follow-up after every meeting with the global team, as the majority of the team were relying on a machine translator. Every so often this ended up requiring a follow-up to a follow-up where the local team has questions but unable to ask due to a language barrier. It slowed down the progress considerably.
Another thing I've found particular interesting is that this language barrier also seems to cause some technique or design pattern to be extremely popular among the local developers out of proportion to the English-speaking world. For instance, Clean Architecture is extremely popular in certain industry (e.g., SI). I believe, it was due to the book having a decent Japanese translation, the lack of Japanese-language commentary against it, and the tendency to follow manuals due to social norms.
>Plus look at Ruby, which was all the rage in webdev at one point. It was invented in Japan, and it had better documentation in Japanese than English.
Ruby has somewhat infamously had a divide between "Ruby programmers" and "Rails programmers". The fact that it was invented in Japan is fairly unrelated to anything about why it succeeded.
Ruby took off here due to Rails, and to a lesser degree _why's work way back in the day that made it so beloved.
I agree with all of your points except part of the last one. I agree starting a business could be harder than many places especially if there is a language barrier. But I disagree that it's difficult to move to Japan.
It's easy for a company to hire foreigners and get a visa for them. There isn't even a local market hiring test like there is in most countries. If you're skilled enough in the newish points system[1], the process is fast-tracked and you get benefits like your spouse can work without restrictions and you can get permanent residency in as little as one year.
Or you can get an investor visa[2] to start a business. The main requirement is having ¥50M of capital. If you can't swing that you may be able to get a startup visa[3] with relaxed requirements.
The taxes also aren't that high and while salary is much lower than US for tech people, the cost of living is also much lower. If you're coming from Europe to Japan, you'll probably enjoy a similar salary, with lower taxes and much lower cost of living. And slowly, there is a trend for tech companies to allow remote work from within Japan, with occasional visits to the HQ in Tokyo. I know a few people taking advantage of this to move to a low-cost area to live a quiet family life, or be close to skiing or whatever.
> Windows and Mac dominate the world's computer systems, and the development tools for those environments are English-centric. The translations are lacking or no existent. The struggle of a Japanese developer to make an app is greater than that of a primarily English-speaking developer.
The lack of translations is true for almost every country and culture apart from the natively English-speaking ones. Some western languages such as German or French might get more in terms of translated (or natively written) documentation and teaching materials but many smaller ones have next to none. I've also seen software development materials with translations to Japanese or Chinese and nothing else (beside English).
I can't remember the last time I read any software development material or used software development tools in my native language.
There may still be some cultural differences that could lead to there being more or less of a language barrier, of course. Other languages may be closer to English, at least in terms of the writing system. In some cultures people might be less inclined to learn English as a second language well enough to be comfortable than in other cultures. Japan, having a large economy with a large domestic market as well as a strong language identity and a relatively homogeneous culture, might be one of those. (I seem to remember that, anecdotally, some highly educated Japanese people I met in the past were surprisingly uncomfortable speaking English, but I'm not sure how much of that was due to actual skill and how much it was about confidence or a lack of experience.)
But the anglocentrism of development tools and culture alone isn't enough to explain a difference between Japan and many other countries.
I wonder if the Windows/C#/VS ecosystem translations are actually decent leading to Japanese developers not being forced to learn English? From what I have seen, Windows and Visual Studio have usable Japanese translations and documentation. I have seen code from Japanese devs that if the original writer had had better grasp of English, they probably would not have made the mistake by understanding the English and argument names.
This also means that outside of the MS ecosystem, it's a huge language barrier to overcome. It's probably not the main factor but just some foods for thoughts.
I think it’s more than just language but in this case I think the relative economic history is a factor: Japan had a large modern economy so there were tons of jobs working at domestic companies building with domestic tools. Modern China’s economy & educational system developed later and a lot of high tech work was dominated by American companies. I think that meant there was a generation where an equally bright engineer in Japan could think they’d have a solid job at one of the biggest companies in the world and never need to be highly proficient in English but their Chinese counterpart would make the opposite decision.
We’re at a really interesting point where that calculation may be reversing with China being a huge domestic market with strong pressure for independence.
China has more developers that speak English than Japan does and they’re ~all in Shanghai, Beijing or Shenzhen because that’s where the money is. Shanghai is substantially easier to get around with very bad Chinese than Tokyo is with very bad Japanese.
Overwhelmingly majority of dev in Shanghai (or China in general) are still native Chinese. "Easier to get around with very bad Chinese" isn't a factor, at all. Not even rounding error.
China is better than Japan in software development simply because they pay (much) better, which causes more people (esp. more "smart" people) to pursue this career for the money.
And the reason why Japan doesn't pay software dev as well is in the article.
That's true for Japan too, though. With 125M people, there are plenty of local incumbent apps and tech companies that don't feel any need to compete globally.
Sorry but this is BS. What's your background here? The author of the article has quite some credibility here having lived in Japan for decades, and worked in said industry.
> And I think the answer is the language barrier.
You need to learn about a dozen English keywords to use something like C++. The rest is abstract thinking. Windows, Visual Studio etc have been available in Japanese for a long time. By that logic, every country that doesn't speak English should suck at software development.
As I read the article, post 2000, the simply wasn't any path forward for a startup. Nobody would give you money. Nobody would take you seriously. There was no easy way to distribute your software. It didn't cross anybody's mind to do so.
> You need to learn about a dozen English keywords to use something like C++. The rest is abstract thinking. Windows, Visual Studio etc have been available in Japanese for a long time. By that logic, every country that doesn't speak English should suck at software development.
Better to say every country where the professional class can’t be expected to read English can be expected to suck at software development. I can’t think of an obvious counter example. Being mostly isolated from the mainstream of software development doesn’t mean you can’t achieve anything but it does make it harder.
FROM is a a great game company, but also a perfect example of what this post is talking about. Their code is REALLY bad. There was a cross game RCE they had to shut down all their servers for recently, the games are riddled with bugs, and the actual code structure itself (from reverse engineering) is spaghetti.
Kojima Studios has made one game using an engine from the Netherlands. Other Japanese games tend to be good despite technical flaws, like Fromsoft keeps releasing game ports locked to 30FPS.
On the other hand, Japanese game designs actually care about how the gameplay feels, whereas the West seems to think they're a kind of movie where sometimes you press buttons. (And indie developers think they're about making ironic platformers where you learn to talk about your feelings.)
Games, and especially console games targeting immutable hardware, are a different world. On a console you control the environment 100% and it simply doesn't matter what you do behind the scenes as long if it works.
Seriously? PlayStation 5 and Nintendo Switch remain top of the class for console hardware, and Square Enix, Bandai Namco etc are still huge names in games.
I wouldn't consider Nintendo a failure, rather they do things their own way. While Sony and MSFT both compete on the latest and greatest hardware to run AAA games, Nintendo is instead providing an entirely different gaming experience. Typical games reach success by making graphics better, framerates higher, and pushing the boundaries of what the system can do, Nintendo seems to have taken the approach instead of making modest hardware with software that is very tailored to that hardware.
The best example of this is Zelda: Breath of the Wild. Technically the game is not that impressive compared so some of it's RPG peers, but that's not the point. While no detail is by itself that impressive, the whole package is fantastic.
So many parallels with the Italian software industry. Plagued by low pay, clerical work and system integrators monopolising the entire industry for decades.
I also see cloud computing and Silicon Valley glamour as the turning point for the industry.
Yet Japan has had a flourishing video games industry since the 70s starting with arcade games. I would have liked the article to touch on such counterexamples and what is different in their case.
The article/podcast mentions how quality hardware engineering was still valued, and used Nintendo as a specific example of success. Sony was mentioned as being outside the keiretsu system, and afaik Nintendo has never been involved with one either.
This also correlates well with my personal knowledge of the industry. Japan has a great track record of game design, but they aren't exactly known for innovation in software specifically. Some examples:
Xbox spearheaded the consoles in the online space, and Nintendo is still far behind.
Japan runs circles around the rest of the world when it comes to making successful fighting games. Yet western modders are also known for coming in and fixing online play by adding network rollback.
I agree, I think a lot of people here in this thread gloss over the fact that /games/ involve a lot more than just code. There's writing, graphics, music. I'm sure we all agree some of our fondest memories are from Japanese video games. Japanese video games are successful because they are great works of art.
But what we're talking about here is Japan's software industry at large. The majority of software aren't artistic programs, they are programs written to enable someone's productivity. The quality of the code consequently becomes a larger or even the only factor, and Japanese programmers on average really suck at their job compared to their peers from other countries.
Nintendo does have a pretty strange corporate structure though. A lot of their games are made by separate companies part-owned by Nintendo whose headquarters is coincidentally on a floor of a Nintendo office building.
Such corporate structure is very common in Japanese big company because of how salary decided. Salary was basically decided by age (but it's changing/changed now). Some division like developing (for non-software company), QA, support, are often separated from main company to make salary table different.
You can successfuly export Japanese games, they're valued because they are different. You can't successfully export Japanese enterprise or user software because it really predominantly is designed for the Japanese market which is very unique. It's not even awful as the author claims, just different.
People always create these complicated narratives for why non Anglosphere countries don't create many consumer or business facing software products and the answer is simply because they're at a comparitive disadvantage both in terms of culture and scale.
My take is that Japan due to their focus on hardware has a very good low-level programming culture that is still relevant for console games. I dabbled with SNES programming recently and what Nintendo, Capcom, Konami, Squaresoft... could do with assembly in the 90's is very impressive. The failure talked in the article is about higher-level programming and application software.
Japanese games usually have a pretty low software-quality. They are full of strange bugs, lacking good UI&UX, are technical inferior to the western state of the art. Even AAA-Games have often insane bad details on non-game-parts. They often compensate this with good or at least different content, but as we are talking here about software, engine-features and -quality do matter.
To name some prominent example: the latest Pokemon-game released 2-3 months ago, is infamous for how bad the graphics were, how many bugs it has(had?) and how bad the performance is. And this is from an old established company, making billions with their games.
Pokemon is a pretty extreme example though. In my experience, Japanese games are generally less buggy than AAA games from the west. Though they are less graphically ambitious
This is a question mark for me. In nintendo in particular, many of the software engineers ended up being quite important in the company, although there were also game designers so may be the logic is different here?
Thanks for this. I'm probably going to have to move back to Japan soon, since the wife is from there and can't handle living in the USA
Rather than take an 80% pay cut to partake in the salaryman lifestyle, I'm going to have to start some sort of business. This podcast should have some relevant info, haha
Based on the experience of people I know, the 'easiest' way is to get hired by a company in the US (or wherever you are) that has a significant Japanese branch, work there for a while, and then ask to get transferred. But be aware that even 'better' working conditions can be much worse than you are used to.
That's a big long shot. The easiest way is to simply get a job at a Western or internationally-focused company here; that's what I did. If you're an experienced software engineer, there's lots of companies looking for English-speaking engineers.
20M/year is a huge salary in Japan. If you're looking for American salaries, you will only find those in America. The cost-of-living outside America is generally lower and life is lower-risk (no medical bankruptcy), so you don't need as much money, plus people outside America don't demand and expect to own giant McMansions.
Well, you can make 50M or more per year in the USA, so there's no point in working if you're only taking home 20M. You're financially better off working in the US, saving some money, and then being an unemployed hikikomori in Japan for a few years (which is pretty much my plan at this point, haha)
Sure, you're financially better off in the US, if you don't mind your kids going through active-shooter drills at school, getting shot at sometimes by their classmates, yourself being shot at when you're driving to work by someone with road rage, religious and political nuttery all around, and society in general breaking down. You're also totally forgetting the extremely high cost of living in the US, the possibility of medical bankruptcy no matter how much money you make, I could go on and on.
Honestly though, if you actually like living in the US, you should stay there (this goes for anyone). If you're not unhappy there, there's no good reason to leave for a country that's so completely different and unfamiliar, where you're not likely to be happy at all. The reverse seems to be true for your wife, so I recommend divorce so she can go back to a place where she's happy. The two of you seem to be incompatible. I've seen other marriages break up over the partners being unable to agree on where to live or to retire; to me it seems to be one of the most fundamental compatibility factors there is in a relationship, though it's frequently ignored or overlooked.
Also, jobs paying over $350k are not common in America as you seem to think.
ah, she just doesn't like the way things work here, and she's scared to learn to drive
The service in the SF Bay Area is really amazingly dismal. It's embarrassing every day to be an American living here. People at restaurants, stores, etc are all angry and stressed out.
I'm hoping I can convince her to move somewhere else in the USA with me, since it'll be a lot easier to find some remote work in my same time zone
Somewhere else will still require driving for every little thing, because that's how America is. If you want to live a car-free lifestyle, America is simply impossible outside of Manhattan, NYC, but it's commonplace in any larger Japanese city.
And Americans are all angry and stressed out everywhere else in America too. It became a really miserable place to live in my experience during the pandemic, so I was really glad to get out and move to Japan where everything is civilized and people know how to act in public.
Well, we've had pretty decent luck with Utah, Florida, Texas, and other parts of the US lately when traveling. The service isn't on par with Japan, but they're nice at least.
I'm hoping I can convince her to learn to drive.. Although if we move near my friends and family, I could just have one of my minions drive her around every day
Hopefully I can work out a way to earn $8k a month or so doing something online. Then I could maintain residences in Tokyo and the non-HCOL part of the USA, and pop back and forth occasionally.
>Although if we move near my friends and family, I could just have one of my minions drive her around every day
Is she some kind of slave or something? Being in a place where you can't get around by yourself is not much fun, and doesn't feel very free. I can totally see why she wants to come back to Japan; you don't have to live that way here, and can easily get around almost anywhere (except the countryside) without a car. It's a totally different way of life, but it's something that most Americans I've found simply cannot comprehend at all. They simply cannot imagine life without a car (or a gun).
>Hopefully I can work out a way to earn $8k a month or so doing something online.
Good luck with that; most WfH jobs won't allow working internationally because of the taxation issues, though I suppose you could lie and just use your US address, though the company might notice that you always work through the middle of the night and have an issue, but I really don't know. I know digital nomads do exist; it's not much of a thing in Japan because there's no working visa that allows this, but a spousal visa probably would.
Yeah, it sucks man. I tried explaining this to her, but she's still scared to drive and spending most of her time alone
At this point I'm basically a nampa weirdo, hitting on all the Japanese housewives I come across to get their LINE info and try to arrange playdates for the wife
Very interesting article, but I almost can’t listen to the podcast version because of the narrator’s over-dramatic voice. The way he puts extra tonal emphasis on every sentence sounds like he’s narrating a trailer for a 1990s action film!
> If you want details and debate about exactly how Japanese software falls short, or if you just in the mood for some good old-fashioned venting about being forced to use it, check out Reddit or maybe Hacker News.
Is gaming not part of the software industry? What aspects of Japanese game development made it hugely successful compared to business and other software development?
I know the author likely has more experience in Japan, but I don't understand this lust some developers in Japan have being like American start-up workers. There is a lot of money in it as an individual, but start-ups have sometimes been great but have also been terrible as well. I know I'm kicking against the goads here on HN by pointing that out, but an objective look at it shows it to be true. Just remind people that wework and Theranos existed, or point out the most important IT companies today started out as start-ups perhaps but are nothing like them now, and now look very much like the old guard they "disrupted." Start-up'ing at least from the outside just looks like an initial condition, a spawn point, but the final situation that is recognized as success is a large company with many admins and an HR dept, or being bought by one. That looks a lot like the keiretsu but without the commitment to workers, so really the worst of both worlds but the best for MBAs and HR people.
I sympathize with the lack of respect software developers have there perhaps, that I understand.
Although the article is about the entire software industry of Japan, and not about the efforts of one individual, I feel like it does a great disservice to not mention Yukihiro Matsumoto, aka Matz, the inventor of Ruby? Since 1995. How many startups have been launched with that platform as their basis?
Also we should not forget the history of the MSX platform that is tied to the ASCII Corp. in Tokyo. It was an early 80s startup led by Kazuhiko Nishi who licensed MS Basic and introduced MS products to Japan for the first time. Later it became the de facto Basic standard for the MSX platform that kind of dominated personal computing in the 80s (similar to how Commodore and Spectrum dominated that market in Europe during the same time).
There is this cute photo of Nishi-san posing next to the young Bill Gates:
This might be true but generally we don't count, say, Python- or semaphore- using startups as successes of the Dutch software industry, unless they are Dutch startups.
If you're asking why they don't have a Google, Microsoft, Apple, or Facebook, could it just be because they speak a language that's only useful in Japan and not anywhere else in the entire world?
If you're actually in Japan they have tons of apps, tons of software, tons of games, tons of many things "software".
Certain things fit Japan better than others but for example Mercari (the ebay of Japan) is 1000x better than anything I've seen outside Japan.
Japan's feature phones ran rings around the rest of the world's phones until iPhone.
There are certainly other factors like possibly underpaying, not valuing experience, etc... but I think the basic premise of the piece is probably missing the forest for the trees.
What this comes down to is simply economics and incentives, and how those economics and incentives are changed within and from outside by other forces. That is it.
even the concept of "Galapagos effect"is easily explained by government created market distortions.
1. Japan was a closed country for~250 years, until the US "Gunboats of democracy" opened up Japan under Commodore Perry threatening to flatten Japanese cities. Prior to that there was restricted trade/trading posts with the Dutch, Portuguese, and then Germans.
2. With the US military threats, Japan side the treaty with the US and realized pandora's box was open, they say the Western operations, war and colonies and said, time to get moving. Thus, German and Prussia in particular suited Japan well culturally and in many aspects - so this tight tie up, creating or promoting trading companies or families (Zaibatsu) & forming their own colonies, China, Taiwan, Korea, etc.
3. There are many other factors within the run up to WWII and after but Japan was a major innovator, especially after WWII as central control imploded - think Honda, Sony, etc. the issue is that as PCs came into play Japan was highly centralized again & there were numerous issues of lacking vc or risk capital, social risks, financial risk, labor mobility ,etc. And even today most software devs cannot hardcore sell - imagine 1970, 1980, 1990s.
4. The native ERP market was highly customized (mostly SAP) using SIer's and they wanted it specific to their unique workflows - the old 80% of what I want is not enough - funny enough that is a competitive advantage again.
5. At same time, most of foreign program work was either L10N or integration ,sintalls, support/maintenance for the foreign software vendor - Oracle, SAP, MSFT, etc.
6. Btw the time we got to 1990s, Japan had many innovations in software, BUT this ran into a wall - it had a horrible startup ecosystem and so it was occupied by ossified rent-seeking incumbents like Sony, who fought their own engineers. Sony was the original Apple, Akio Morita the original Steve Jobs, including taking the most important case, Sony vs Universal Studios to the US supreme court and winning. Ipod, imac, etc. are simply the 2.0 or 3.0 version of the sony innovation products. I have interviewed a dozen of these engineers and managers going back to 2006. And also other interviews state side of how products like Slingbox/Slingmedia were just sony products and so on.
7. At this point the social risk, the financial risk, the career risks and labor mobility of joining a startup or high grow firm in nascent industry are greatly reduced, more programming resources than ever, smart hard workers, it is just the opportunities -- opportunities that were artificially restricted by ossified rent seeking incumbents using regulator capture.
The first video game I played was Pokémon blue, created by a Japanese game studio, and the most recent was Elden Ring, created by another.
Sony and Nintendo are two of the world’s biggest companies, and a key part of their business model is creating great games franchises, either in house or working with local studios.
This is all driven by an ecosystem of great software companies, making the best games in the world (or at the least, outside of America).
Sony closed SCEJ/Japan Studio and moved PlayStation to the US branch, where it's now being run by people only interested in AAA games following the Hollywood movie + collectathon + puzzles where nearby NPCs yell the answer at you after 30 seconds formula.
So I don't feel like they're demonstrating Japanese software quality at the moment. The Alpha series cameras are pretty nice, but not as much as they could be.
Speaking of Japanese cloud computing, years before Amazon created AWS, Japan had the NTT cloud. We used it for about a decade. At first it was terrific, but near the end it was becoming unreliable. Our hypothesis is other cloud providers were siphoning off NTT’s best cloud engineers.
One of the key issues of Japanese software is that most companies don't rethink problems and processes to suit software, they try make software around an existing process. Most Japanese companies will just try to automate, or digitize some type of manual process and keep the people involved, I'd say just to hang onto jobs because firing people in Japan is pretty much impossible.
Recently while visiting Japan, I tried to get a eSIM card through "Docomo" the national phone carrier via their "online shop". I know...but I had to use them for other obscure reasons.
It was incredible to me how long it took to get the eSIM card. Days and days of problems, hours on the phone. I hit so, so many edge cases it was hard to believe. I actually felt depressed how bad the experience was. Everything was so obscure, and so so many things had to line up in their system.
I actually debugged the software by talking to support people and gaining clues, I finally worked out how to hack the inputs to allow it to give me a eSIM.
The problems were mostly that I had a middle name, which Japanese people apparently never have, say your name is "Ella Mary Smith" but your credit card is "Ella Smith", the "people" won't accept it and their web app does weird things, like removes the space between your first and middle name when you put both of them into the first name field. You some how have to work out through trial and error, the system / process / people don't care about spaces.
The process involved many people "checking" my application and almost being able to arbitrarily deny the application. There was almost no actual automation, I think it would've just been way way easier to visit a store and let them fill out the forms how they wanted, however they've made this particular process "online only" and so you're stuck with probably one of the worst user experiences in the world.
Anyway, each time I applied, I had to wait about 24 hours to be rejected and resubmit the application. Sometimes the application would blow up or the website would time out, error and I had to resubmit all documents again and do a silly online facial identification thing, I did about 12 times. I truly thought I was going to go insane. I didn't leave the hotel for 12 hours trying to hack this beast and workout what was wrong, then multiple day of rejections until I got the combination of inputs right.
What was so striking to me was the way I, as a user had to invent my own solutions to allow the system to give me the eSIM.
For example, I eventually worked out that I must have a credit card which includes a middle name, even though there is only a place for a first and last name when inputting the credit card. So I just lied and said my credit card actually is written as SMITH ELLAMARY (which it doesn't), and the "person" who was checking the application, believed this fairy tale and finally accepted the application.
It was a classic example of so much software I've interacted with in Japan in recent times. It always has this kind of "best effort, just works, sort of works if you're a foreigner and you know how to work around the quirks" vibe to it that I don't really experience anywhere else.
> but your credit card is "Ella Smith", the "people" won't accept it and their web app does weird things
Ha.
I live on street with a long but not that long name (18 chars for what it worth). I recently encountered a site what didn't want to accept my address because the full address was way too long.
> credit card actually is written as SMITH ELLAMARY
And here I am who specifically writes BANK NAME in the CARDHOLDER input instead of card holders name where I can.
So they were cross checking the name on the card with the name on my passport, apparently they think it's a security check and some how not realizing that:
1. The name on the card is rather arbitrary.
2. Accepting names without spaces isn't really a good check IMO. If you wanted thing to line up exactly, spaces should be accounted for, but who cares? If it feels good...
This is a bit of a cultural thing I find common in Japan. It doesn't have to make logical sense, but it's a requirement and compliance is very important.
One thing that trips up even native Japanese speakers is the windows IME. It likes to change characters like spaces, dashes, alphabet letters and numbers to roman ones when you intend them to be Japanese "full-width" characters. In our office there, our address was often rejected because the final part describing the floor number should be "−４Ｆ" would come out of the IME as something like "-4F", "−４F", "-４Ｆ" (yes that last one is wrong).
The MacOS IME is much easier to use, and was the singular excellent use of the MacBook touch bar. It lets you pick the correct character, pick autocomplete suggestions, and convert between full-width, half-width, katakana, hiragana, etc.
>Recently while visiting Japan, I tried to get a eSIM card...
>... Days and days of problems, hours on the phone. ...
>... I truly thought I was going to go insane. I didn't leave the hotel for 12 hours
All this just because you just had to try to use a US-spec iPhone 14 Pro instead of getting a phone with a real SIM slot. eSim isn't really a thing in Japan yet, or anywhere outside the US, so support is iffy. Dual-SIM phones, OTOH, are normal and common in Asia. You could have saved a ton of time by just buying a 30-day SIM card at the airport like everyone else. The first time I went to Japan, I just bought a SoftBank SIM card at 7Eleven, popped it into my phone, followed the procedure in the instructions, and within an hour I was set.
For people with iPhone 14 Pro phones from the US, the solution is really simple: just use international roaming. If you can afford the top-spec iPhone, you can afford roaming charges.
>The problems were mostly that I had a middle name, which Japanese people apparently never have
No, they don't. Middle names don't exist in Japan, just like names written in Kanji aren't recognized in the US. Foreigners get around it by concatenating the first and middle names, or dropping the middle name altogether.
>like removes the space between your first and middle name when you put both of them into the first name field.
Why would you put a space into a single name field? Japanese names don't have spaces in the middle; they're single words (one each for first and last name).
>For example, I eventually worked out that I must have a credit card which includes...
Or you can just buy a SIM card at the airport with cash or a credit card (paid by scanning so your name doesn't matter).
>I am not going to waste your time or mine cataloging and complaining about the many, many bad practices, user-hostile design decisions, mind-boggling complex workflows, and poor development process that afflict Japanese software.
>If you want details and debate about exactly how Japanese software falls short, or if you just in the mood for some good old-fashioned venting about being forced to use it, check out Reddit or maybe Hacker News. This topic comes up pretty often there.
>No, for the sake of this podcast I’m going to assume that we are all in agreement that on average, Japanese software. is just … awful.
What are some examples of Japanese software? I'm not aware of any Japanese software I use besides a few random apps from the App Store, all of which seem fine to me. "Japanese Software is awful" is a common topic on Hacker News?
I worked for a startup that was based on the fact ( amongst other ) that Sony was making a good hardware product, but terrible software for it.
Mostly the UI, but also weird restrictions on the input / output while it was a 20k professional pieces of hardware in a highly normalized market. ( easy to see that Dolby was NOT restricting it for instance )
We had a good run. Thanks Sony.
Detail : the setup of that industry made it that Sony and other HAD to let some api open for us insects. That was our opening.
For those reading comments first the story starts in the third paragraph.
It could only me but the way it is written obstructs the consumption of the content. Two long paragraph went on before started to talk about the actual topic, after elaborating about how much the author worked and work on it and what is not going to talk about in a stage entertainer kind of inflated manner (the comment "I am not going to waste your time" is ironic then). It kind of derails the focus from the content in my opinion.
The japanese way of working doesn't work for software industry. Everything is so slow and there's too many different organizations cooking the same soup. Japanese way of working works best for things that already have been designed for you and just needs building. The only exception would be video games, but it seems game development is heavily followed and developed by a single company, compared to typical corporate / business software.
> Cloud computing drastically reduced the capital and time required to start a startup. In the dot-com era a decade before, starting an internet startup required purchasing racks of servers and paying system administrators to keep them running, but suddenly fully configured, maintained, and secure serves could be had for a few cents per minute — pay as you go.
> Suddenly Japan’s software developers didn’t need to explain their idea to a VC and convince them that it would sell. They could just build things and get people to start using them and start paying for them. And that’s just want they did.
Your hosting cost, whether you're on bare metal or the most opinionated git-push-your-rails-app is a tiny line item on the P/L. Rarely have I seen an instance where the pay-off period for spending, say, a month of one engineer's time (for a nice round number, say, $10k) reducing the hosting expenses is less than a year (say, $1k/mo in hosting expenses).
In just about every case you're better off spending that developer-month building a feature or delivering a contract.
So yeah, maybe cloud computing saved you, at the outside, $10-20k server parts and a rack in a colo or whatever.
TLDR with or without The Cloud, 2022 or 1999, your server costs are tiny in comparison to what you spent on payroll getting to market.
I think you got it wrong: Buying hardware and setting it up and keeping it running, just as described, is / was a high barrier of entry cost-wise (Not to mention the fact that people used to actually running hardware in a data center are scarce). Many (sometimes poor) people could not / would not take that bet.
With "cloud computing" (to me even before there was the word cloud around, but thats nitpicking) suddenly your only real investment needed was your own time, which to you is "free" if you can somehow make a living on the side (some people simply accept to live poor for some time, some had their family, etc.). The few Euros for a vserver (or "webspace" at the time) were doable for most if not all. Not to mention student programs, free credits to lock you down to hyperscaler X, etc. we all have nowadays.
Or comparing it to myself: I learnt programming at the age of 12 because we got a free environment alongside our ISP contract for executing PHP with a mysql database behind it. I would never have spent any money. Now I have been a professional software developer for almost 16 years which I say is thanks to that way lower barrier of entry that was created along the lines of ~ 1995 - 2005.
In pure dollars per minute, yes, but the primary benefit being alluded to here is a political one. A small team can use tiny discretionary funds and build a working prototype quickly. The same project is dead in the water if you have to go get buy in to allocate company compute, get multiple VP signoff on a project proposal, etc etc.
Cloud computing turned asking permission into asking forgiveness, and that's a real benefit politically.
The part you've quoted is pretty clear that the cost it's referring to is maintenance cost for on-prem servers:
> ... and paying system administrators to keep them running.
That agrees with your main point, which is that employee time is the main relevant cost. (Nowadays, cloud services seen to have gotten so complex that you still need an eternity of employee time to set them up, but that's another matter.)
The one thing I might add (don't know for sure, just guessing) is the cost of things might look different in Japan vs. here in the US. May be the cloud made a serious differential in the upfront cost as hardware is more expensive there, I don't know.
Buying anything in Japan (even an MS Office license) is at least more likely to require going out drinking with salespeople rather than just visiting a website.
In the 80's, the old entrenched conglomerates decided it'd be easier and more profitable to expand their hardware divisions than invest in a new unproven software market.
I had to split it up into three pieces but it worked
> me: can you summarize all the information I gave in three sentences
> ChatGPT: Japanese software is known to be of poor quality, but there are historical and cultural reasons for this. The zaibatsu and keiretsu systems that dominated the Japanese economy in the past prioritized hardware innovation over software development, leading to a lack of respect for software developers. In recent years, the emergence of cloud computing and the smartphone market have allowed Japanese software developers to become more influential and respected, and the industry is slowly improving.
This article describes how Japan's economy and the software industry evolved from the post-WWII years to the present day. It explains how the Zaibatsu, large corporate groups that dominated the economy before WWII, were dismantled by the American occupation forces and replaced by the Keiretsu, weaker but familiar forms of the Zaibatsu. It highlights how the Keiretsu focused on the domestic market, rather than expanding globally and developing a software industry, during the 1980s and 1990s. This had a negative impact on innovation and caused Japan to fall behind in software development. The article then goes on to explain how cloud computing and smartphones have enabled Japan's software developers to start their own companies, rather than working for large corporations, thus creating a more vibrant startup ecosystem. It also suggests that Japan's software industry still has a long way to go to catch up with other countries, but there are positive signs of progress in certain areas.
> Shakespeare only wrote 37 plays, Orson Wells only made 64 films, Mozart wrote 68 symphonies, but Disrupting Japan? Well, as of today, Disrupting Japan has 200 episodes.
This is the most alarmingly stupid bit of text I’ve seen on this website in years. It’s wildly disconcerting that there exists some sort of apparatus that allows people to buy top ranking on the front page with utter garbage that clearly has zero value.
Edit: There are two accounts that are happy to publicly disagree with me within seconds of eachother, and three-six that are are happy to downvote privately within those same seconds.
Maybe I’m magically guilty of having The Wrong Take or there are some people that will kneejerk use their network of accounts to downvote things that aren’t useful to their buddies.
No ... it's really not a giant conspiracy. You're just guilty of being a jerk on a very public website which doesn't take kindly to jerks.
I've downvoted all of your comments on this thread because they are completely unrelated to the subject of the article, and consist solely of a joke spectacularly flying over your head to the point where you're getting aggressive and throwing around conspiracy theories about downvoting rings.
I had written a sarcastic response, but I'll be nice instead. Don't you think two people could independently see that you missed a joke and come up with the very generically phrased response of "isn't it a joke?" to point this out? I mean it's four words. It's certainly a more plausible explanation than being some sort of grand conspiracy involving sockpuppet accounts.
The problem isn't that Japanese companies "missed" the opportunity to do software at X point in time--the problem is that they are culturally Japanese. Japan is about consensus-based design, there is a high penalty for making mistakes, and individuality/creativity is not valued. 出る杭は打たれる - "The nail that sticks up will be hammered down". Certainly you can develop software this way, but the Silicon Valley model of rapid iteration, "if it doesn't work rewrite it", "move fast and break things", etc. gets you there sooner.
But what about Matz and the Ruby core team, you say? Certainly they are a zany, free-thinking bunch. I attended Ruby Kaigi last year and noticed that most Ruby core team members are hired by domestic Japan-focused tech companies, think Cookpad, Tabelog, etc. They are kept for their HR value to attract tech recruits and allowed to spend all their time on whatever open source project they want. Meanwhile the peons work on the actual product.
Source: I've worked in software in Japan for 15 years; 5 at a Japanese bank and 10 as a CTO of my own company. We run circles around our Japanese competitors.