Comment by po | Hacker News Reader

Comment by po | Hacker News Reader

Parent Full thread

As someone who worked at Sony in the US and later founded a successful startup in japan, I feel I can speak with a bit of authority here… The main thing to realize is that this article is talking about Japanese 'tech giants' missing the boat, not HN-style software startups.

For many years Japan dominated and succeeded with well-built consumer hardware. The best and brightest minds became hardware engineers at Sony, Toshiba, Fujitsu, etc… When the industry shifted to software and hardware blended together in a seamless fashion - an ecosystem bigger than the device itself - those minds were particularly ill-prepared for the fast paced, just-ship-it style development cycles that software demands. The product looked good on paper but the lifespan of the consumer relationship was inferior because it didn't include the user's newly found software lifecycle.

We all know how Microsoft missed the boat when the internet came along and disrupted their office/windows strategy. Now, imagine a whole country filled with Microsofts. Japan has floundered in denial for a while.

The good news for Japan is that the in the last few years I have seen a huge growth of people diving into it. There are a ton of coworking spaces and incubators popping up, and people seem to be focusing on software now. I actually think the big earthquake was a bit of an awakening for people here. More people want to work at a company doing something cool. The younger generation seems to understand that hardware is no longer king, even if they don't have the mentorship and guidance they would have in the US. It may take a bit of time but I think it is already under way. They have amazing work ethic and drive and I am hopeful that Japan turns the ship back towards innovation.

Sony didn't just 'miss the boat', they did what they could to sink it and damage their brand.

Before they had a fantastic reputation, their audio equipment was good to excellent for a small fraction of the price of the competition.

Rootkits, lawsuits and me-too products seems to be what they specialize in nowadays.

Toshiba, Fujitsu, Matsushita and many others (also in other branches than just electronics, for instance Toyota and Honda) have done a lot better and definitely didn't aim the shotgun at their own feet.

Another thing that hit Japan was that they were no longer able to compete on price with China and Korea.

> Rootkits, lawsuits and me-too products seems to be what they specialize in nowadays.

Didn't wait for that. The issue with Sony is that they're historically utterly incompetent on the software side of things. I swore off Sony products after getting a NetMD. The software for that POS was so absolutely terrible it makes iTunes look good, snappy and user-friendly.

That 'shoot yourself in the foot' strategy is exactly what you get when you put old school hardware guys in charge of software. Sony had years and years of success with proprietary connectors, batteries and exclusive content deals. Consumers don't really like it but they quietly tolerate it if it means getting the newest features. Put those guys in charge of your software strategy and lawsuits and rootkits is exactly what you'll get.

Also keep in mind that the rootkit came out of the 50-50 joint venture Sony BMG and was the brainchild of Bertelsmann as well. The guy who went on-air to defend the decision saying effectively 'if you don't know about it, why should you care?' was in charge of that division and originally from the Bertelsmann side. The old-school publishing industry has almost the exact same relationship with software.

By then they'd already started their slide.

I remember Sony for things like TA3650, the Walkman and their (excellent) reel-to-reel recorders. By the time they got into media (and don't forget betamax) they'd already lost their spirit, even though financially they had not yet peaked.

I think we agree. You remember Sony for things which have no software component. The only software Sony was very good at was embedded software. I still have an original AIBO which sort of exemplifies that.

I just realized that from my window here in Tokyo I can see and recognize the tops of 2 large buildings: NTT (Japanese hardware) and Oracle Japan (foreign software). This is why I was so encouraged by the large acquisition of a foreign software company by Recruit this past week.

Part of the problem is software development just isn't seen as an integral part of company operations. A friend of mine is a developer in Japan. He makes less money relative to his peers than he would in the US. He gets about the same amount of respect from the company as the guys on the loading dock - management doesn't really understand what he does and they view him as a replaceable part.

I asked him why he doesn't move to another company, something which is becoming more accepted. He shrugged and said they were all pretty much the same. He's planning to leave software behind as soon as he can get promoted, and they'll replace my friend with a guy whose only experience is a doujinshi game that was written by himself in college.

It all boils down to a population problem. Japan is ageing and software demands huge work force. Only India and may be China in the near future can bridge the gap. It is not accident that Silicon Valley has more Indian engineers than any other foreign nationality.

I am an Indian (from Goa). Your idea that the millions of Software Engineers from India are going bridge the gap is misguided, IMO.

What Software innovation has India done? Do you know of any compiler companies from here? Or any Operating System companies? How many linux kernel engineers does Infosys or TCS have? Both companies are sitting with billions in cash. Do you think any senior directors there are thinking - let's create a microchip, even if it's way behind Intel or AMD? I remember reading an interview of Bjarne Stroustrup - he said AT&T gave away their compilers to their competitors (like HP) knowing fully well it would strengthen them. Is any company in India on the same level as AT&T, who can see the bigger picture ? Maybe Tata's come close (not TCS). Indian senior management are like glorified buniyas - who when they put down 50 paise, expect Rs 1.5 back. It took Standard Chartered to come to India and sponsor a marathon. That should tell you a lot about senior management here.

On a larger scale is any Indian company thinking we can build a commercial aircraft? Only HAL builds fighters - but that's a govt company. Soon Microsoft is going to create an opportunity for everyone to erode their Desktop OS share by walling off developers? Do you think TCS or Infosys is building a Linux Distro like Canonical/Ubuntu? Do you think they have the balls to get into that market and try and take market share?

The fundamental thinking of India seems to be flawed. Everyone wants a piece of the outsourcing pie. Synergies and bullcrap like that. No products come from here. I believe we are already suffering a much worse fate than Japan. Videocon is a collaboration with a Japanese company (since the talk was about electronic majors from Japan). Where are the real indigenous products from India, so that we can have the same fate as Japan? The only advantage the 1 million "good" engineers you claim is cost. Once the cost advantage goes, people are going to be jobless, not bridging any gaps.

I went to 2 leading BPOs with my compilers for Survey Programming and Cross Tabulation. The VPs said the thought of writing such products never crossed their minds. They were not interested.

If you want to see the difference in Indian thinking - compare with Germany. Open source is available to everyone and that means countries too. I have read many articles about Germany using Open source in govt offices and there was an article about France and Libre Office.

India cannot go down the same road as Japan, because we dont have the products in the first place to go down that road.

I don't know if your "bridging the gap", means supplying engineers to other countries. If yes, then this is the typical thinking that creates the net negative cash flow in economies like ours.

Software innovation and engineering go hand in hand, but not necessarily the same thing. Innovation will and have to come from innovators, some of whom would be engineers; but execution requires engineers. No India is not on the innovation train yet, but that is not an impediment for creation and supply of engineers. You seem to be trying to argue about certain lack of innovation in India and whether engineer migration is good for India. Those are separate subjects altogether.

fellow goan here, and i have to say that while you are mostly correct, things are beginning to shift. there are a sprinkling of homegrown "product" companies popping up in bangalore and pune, and a small but increasing participation in various open source communities. hopefully by the next decade indian software engineers will be motivated by seeing what they can build, rather than by where they can make a quick buck solving someone's outsourced problems.

I will believe this, when I see products which are not your typical social products coming from here, not your next twitter app clone variants, after people have done their first walk through using Rails.

What would be challenging products showing some tech know-how (note there is hardly an innovation in any of this, but demonstrates some level of tech competence)?

1. Take the g++ source code or clang, intercept it at the point it creates the parse tree and create a code navigation(cscope for c++) or lint like tool.

2. Chapter 6, PAIP has the bare bones of a flight search engine in lisp. You need to add more parameters and a different metric function. Open source the solution to disrupt current monopolies.

3. Google is solving the language translation problem.

4. Disrupt the big 5 in audits (PwC etc). Create an accounting /audit package that uses similar ideas as git. Use SHA1 entries to track accounting entries. PwC, screwed up Satyam's auditing and got away with it (because of bribing?). In the US, Arthur Anderson was destroyed as an entity for fudging Enron. Since we are used to bribing the govt to pass laws, TCS or Infosys could throw weight on the govt to pass a law to accept streams of accounting records using "git push". Audit data is now a weekly or monthly disclosure, using git push to a govt location. Much lesser chance for companies to fudge data once you have a much tighter cycle, tamper proof because of SHA1 checksums.

In the US, Arthur Anderson was destroyed as an entity for fudging Enron.

Actually, it was simply killed by the initial indictment, which for an auditor is a death sentence. Several years later the Supreme Court unanimously reversed (http://en.wikipedia.org/wiki/Arthur_Andersen_LLP_v._United_S...), but by then it was way too late. I.e. this was a typical abusive political Federal prosecution; Enron in general gave the usual suspects an excuse to wreck horrific damage, including most especially SarBox which was the final nail in the coffin of the VC funded hi-tech startup business model.

As for Satyam, at least in US auditors proceed from the assumption that the company it not lying to them and they do some spot checks to verify (e.g. "You say you have 150 widgets of this type in that warehouse; let's go look at them." And "You say you have 29 crores in this account at that bank; call the bank and authorize them to tell us about the account."). If the company is sufficiently fraudulent, as I recall reading Satyam was, auditor malfeasance is not assumed, you check to see if the auditor's verification procedures were reasonable and if they ignored anything they shouldn't have, but otherwise you can't assume they too were corrupt.

Auditing is expensive from both sides, the auditor and the company being audited, where a number of people are tasked with supplying the auditor the information they need and demand (e.g. those verification checks). If a company is running a sufficiently sophisticated scam, e.g. their official books are very carefully constructed fiction, I don't see how their pushing out regular records would help. I.e. the way they pull off their scam is to continuously make everything look good, as opposed to trying to put together a false set of books just before the audit, which would likely result in enough errors auditors would have a good chance of catching them.

The problem with India is that the market has been saturated. All the really good Indian engineers have migrated to Silicon Valley, and the rest of the good ones have been snapped up by the likes of Microsoft India and Google India.

As many companies who have tried to outsource in the last few years and have been burned in the process have learned, the engineers they've been getting have a poor grasp of computer science and an even poorer grasp of English. Until the public school system is fixed so that the Indian masses can get a good education, the situation isn't going to rectify itself very quickly.

India produces more engineers as more leave. There are decent number of good engineers. You have to understand that output of engineers have increased 10 times in the last 10 years and thus you may find quality deficit in some; but in general the number of good indian engineers have gone up roughly 10 times as well. By the way, Indian outsourcing industry is alive and kicking well - it is a $120B industry now and still growing at 15% per annum - not bad for such a large industry at all.

There are certainly a large number of engineers being produced - that doesn't mean they're employable. When you have training companies saying that "75 percent of engineering graduates are not employable,"[0] you know you have a serious problem.

And of course the outsourcing industry is doing well - you have plenty of American companies eager to outsource back-office programming in order to make their next quarterly earnings report look more impressive. This is just one symptom of the MBAification of American businesses, something that cannot continue unabated if long-term success is a goal.

One company you won't hear about much when it comes to outsourcing is Apple. They have an office in Bangalore, but they don't have the sort of branches that Microsoft or Google do. Why? Because they care about quality, even if that means paying their engineers more. If they hire Indian engineers, they'd rather that they come and work in Cupertino, rather than trying to cut costs by leaving them in India. And look where it's gotten them - more than $100 billion sitting in the bank and some of the most popular and profitable consumer electronics out there.

0: http://www.rediff.com/getahead/slide-show/slide-show-1-caree...

Yes the 75% figure may be right and that is a problem. But when you are producing 4million engineers a year, that still leaves 1m good ones. That is the sheer number. The mass effect. Apple is an outlier - so is Steve Jobs.

I fail to see the apple / Steve Jobs connection here.

It all boils down to a population problem. Japan is ageing and software demands huge work force.

I fail to see how it boils down to that. The demographics of Japan are changing (we hear about it constantly in the western media but holy hell spend one weekend in shibuya and you wouldn't believe it) but there's no way it could account for the effects we are talking about. Japan didn't lose their way because there weren't enough young people to fill their software industry; they hit this problem because they didn't even know they needed to be training people up to be software engineers. The best engineers went into solving hardware problems, not software. It's a quality thing, not a quantity thing.

As an aside, comparing apples and oranges is sometimes instructive to see if a theory has a kernel of truth to it. There are 1.8M total people in Silicon Valley (including food service industry and the like) and at least twice that many in just the 20-24 year old category in Japan. There are also at least 10 million more Japanese people now than there were in the 1980's when everyone was hysterical about them taking over the hardware business world. If Japan hadn't been so successful in hardware and had motivated their workforce around software instead, they certainly would be a significant force today. I don't think an aging population plays a significant factor (yet).

http://www.wolframalpha.com/input/?i=silicon+valley+populati... http://www.wolframalpha.com/share/clip?f=d41d8cd98f00b204e98... http://www.wolframalpha.com/input/?i=Japan+population

Reply on news.ycombinator.com