3 ms·

"I think 95% of industrial espionage effort yields zero if not negative benefit" Citation needed. That makes extremely little sense.

baybal2 8y ago
Why? I think it's quite clear, to know what valuable knowhow to steal, you have to know quite a bit about the subject as well. After doing that much research, it may well be that you know enough to do it on your own.
- wadkar 8y ago
  You need to provide evidence for your claims, specially more if you’re going to trow around numbers. For example why 95%? Why not 5%? Or why not 80%? Please understand that unsubstantiated comments are not constructive for a fruitful discussion. Take a look at the HN guidelines for more information.
  - baybal2 8y ago
    This is what I personally think, and given that I worked in a more serious part of tech industry than running diary websites for a big part of my career, I think my though has truth credit to it. Though, my career in serious engineering companies is not that long, I don't see a thing in any serious piece of engineering that ever qualifies as non-obvious. And under serious pieces of engineering I mean things from SCADA for Baosteel's main plant, to widgets for Airbus and EADS. Nor I ever qualify this engineering work as very laborious. Doing "Big engineering" is more about bureaucratic and legal work than doing actual engineering from my own personal impressions. Moreover, whenever "big engineering" is involved, and you hear names like WW groups, EADS, Siemens, Rheinmetal, you don't "innovate." The cost of screwup is so high, that there are even special people employed to watch that nobody dares to replace existing working parts with their "innovation." In general, when doing such work, you want to use most proven and conservative designs over any improvements.
  - orbital-decay 8y ago
    >For example why 95%? Why not 5%? Or why not 80%? I'm probably going to assume too much, but I guess @baybal2 comes from a Russian background, and 95% is a meme/figure of speech there, meaning "the overwhelming majority", and doesn't refer to a specific number. It's (possibly) just a case of English as a second language or misunderstanding.
rleigh 8y ago
It obviously depends upon the context. But for many industrial processes and designs, the design and documentation alone is only the bare minimum needed to replicate it. The rest is in people's heads and hands, and needs to be passed on by training over weeks or months. Here's a trivial example: in a physical chemistry lab I was working in, I was training a new worker how to do a particular analysis. The ISO 9001 documentation said to pipette 1000µl of the sample and 1000µl of a standard into a 2ml GC sample container, using an electronic pipetter. I got repeated accuracy to within 5 decimal places; the trainee got about 1 d.p accuracy, using the same equipment and samples. Because skill and technique matter, take time to learn and master, and those aren't in the instructions. This applies to any process. Having the instructions aren't enough. No matter how many designs you rip off, you still need the right people with the right equipment and the right skills and techniques to fabricate it correctly, or duplicate the process exactly, etc. Industrial espionage can only take you so far. The effort required to rip off the designs and documentation might be more than the effort required to do your own design and establish your own processes. Particularly when you're otherwise inheriting other peoples design choices and you need to completely reverse engineer their thinking. In a way, it's similar to how writing software from scratch can be quicker than making a change to a huge legacy codebase. You're free of the design constraints, and you don't need waste weeks ploughing through someone else's mess to find our what those constraints are. Another problem is the starting materials. Have a read up on the effort required to manufacture FOGBANK, a material used in nuclear warheads. The materials and process were completely documented, but the people who made it were long retired. It took over half a decade of effort to rediscover and replicate the process, not only to learn the techniques but also to discover that the process relied on an impurity in one of the materials which wasn't documented.
- orbital-decay 8y ago
  >The rest is in people's heads and hands Not just in the heads and hands, but in the available infrastructure and supply chain as well. A classic example: SR-71 airframe documentation would probably have been nearly useless for Soviets, since they didn't have good titanium machining capability at the time (they decided to go with steel for MiG-25 airframe, for example).

baybal2 8y ago

Why? I think it's quite clear, to know what valuable knowhow to steal, you have to know quite a bit about the subject as well. After doing that much research, it may well be that you know enough to do it on your own.

wadkar 8y ago
You need to provide evidence for your claims, specially more if you’re going to trow around numbers. For example why 95%? Why not 5%? Or why not 80%? Please understand that unsubstantiated comments are not constructive for a fruitful discussion. Take a look at the HN guidelines for more information.
- baybal2 8y ago
  This is what I personally think, and given that I worked in a more serious part of tech industry than running diary websites for a big part of my career, I think my though has truth credit to it. Though, my career in serious engineering companies is not that long, I don't see a thing in any serious piece of engineering that ever qualifies as non-obvious. And under serious pieces of engineering I mean things from SCADA for Baosteel's main plant, to widgets for Airbus and EADS. Nor I ever qualify this engineering work as very laborious. Doing "Big engineering" is more about bureaucratic and legal work than doing actual engineering from my own personal impressions. Moreover, whenever "big engineering" is involved, and you hear names like WW groups, EADS, Siemens, Rheinmetal, you don't "innovate." The cost of screwup is so high, that there are even special people employed to watch that nobody dares to replace existing working parts with their "innovation." In general, when doing such work, you want to use most proven and conservative designs over any improvements.
- orbital-decay 8y ago
  >For example why 95%? Why not 5%? Or why not 80%? I'm probably going to assume too much, but I guess @baybal2 comes from a Russian background, and 95% is a meme/figure of speech there, meaning "the overwhelming majority", and doesn't refer to a specific number. It's (possibly) just a case of English as a second language or misunderstanding.

rleigh 8y ago

It obviously depends upon the context. But for many industrial processes and designs, the design and documentation alone is only the bare minimum needed to replicate it. The rest is in people's heads and hands, and needs to be passed on by training over weeks or months. Here's a trivial example: in a physical chemistry lab I was working in, I was training a new worker how to do a particular analysis. The ISO 9001 documentation said to pipette 1000µl of the sample and 1000µl of a standard into a 2ml GC sample container, using an electronic pipetter. I got repeated accuracy to within 5 decimal places; the trainee got about 1 d.p accuracy, using the same equipment and samples. Because skill and technique matter, take time to learn and master, and those aren't in the instructions. This applies to any process. Having the instructions aren't enough. No matter how many designs you rip off, you still need the right people with the right equipment and the right skills and techniques to fabricate it correctly, or duplicate the process exactly, etc. Industrial espionage can only take you so far. The effort required to rip off the designs and documentation might be more than the effort required to do your own design and establish your own processes. Particularly when you're otherwise inheriting other peoples design choices and you need to completely reverse engineer their thinking. In a way, it's similar to how writing software from scratch can be quicker than making a change to a huge legacy codebase. You're free of the design constraints, and you don't need waste weeks ploughing through someone else's mess to find our what those constraints are. Another problem is the starting materials. Have a read up on the effort required to manufacture FOGBANK, a material used in nuclear warheads. The materials and process were completely documented, but the people who made it were long retired. It took over half a decade of effort to rediscover and replicate the process, not only to learn the techniques but also to discover that the process relied on an impurity in one of the materials which wasn't documented.

orbital-decay 8y ago
>The rest is in people's heads and hands Not just in the heads and hands, but in the available infrastructure and supply chain as well. A classic example: SR-71 airframe documentation would probably have been nearly useless for Soviets, since they didn't have good titanium machining capability at the time (they decided to go with steel for MiG-25 airframe, for example).