The Era of Black Boxes
melikyan.blogspot.com
melikyan.blogspot.com
When I was a kid I'd make a bit of money repairing TVs by cannibalizing a few other 'corpses', such an endeavor would be nearly impossible today.
I fixed a computer a few weeks ago, the motherboard was put together with el-cheapo capacitors that had all blown their tops until there wasn't enough of them left to stabilize the circuitry they were in.
Easy enough you'd think. What would have taken 5 minutes in the pre-smd time took me a couple of hours and it was still sub-standard work.
Now I've been 'out' of electronics for a long time, almost 30 years now, but in a way I'm really happy. I'd hate to be an electronics hobbyist today, the state of the art has advanced so far that you're probably best off buying an FPGA kit and using that instead of rolling your own circuitry.
Arduino seems to capitalize on that feeling, to allow people to take a small step back from those black boxes and to open them up again.
But there's a lot of fun to be had with Arduino. And non-digital electronics is still fun. Just go look at some of Forrest Mims' circuits in his original books. Lots of fun with a 555 timer.
Also, see the good Make: Electronics book from O'Reilly: http://www.makershed.com/ProductDetails.asp?ProductCode=9780...
thanks!
Part of it is surface-mount components, part of it is that the hardware is being pushed down the commodity stack. But I wonder what kitchen innovations we are missing because few people tinker with the bare metal anymore.
For 30-odd years hackers built their own lightsabers. Now they've pretty much stopped. I can't decide if that's inevitable or not.
I think that most people that are interested in 'tinkering' with things are doing it in software instead of hardware.
One of my favorite classes in school (CS major) was a microcontrollers lab class, wiring up simple processors and microcontrollers and then programming them in assembly. At that level, everything is transparent, there's no hazily-understood stack of abstracted layers written by others.
I think the reason I found it so refreshing was that transparency. The entire system could be viewed and understood at once.
On the other end of the spectrum you have something like Rails, where things happen seemingly by magic, and when they break, it's impossible to know why as a novice.
http://en.wikipedia.org/wiki/Brazil_(film)
"...the real Harry Tuttle (Robert De Niro), a renegade air conditioning specialist who once worked for the government but left due to the amount of paperwork."
I remember him sneaking into homes like a cat burglar in order to fix things without going through the arduous bureaucracy put in place to keep unauthorized repairs from happening. That is close to becoming a reality for anyone who wants to fix their own or their friends' stuff, with the DMCA, etc., standing in their way.
I'd like to make a few additions: While schematics aren't included anymore for most devices you can still get the repair manual by calling the manufacturer. A pasted schematic can't hold all the information that goes into most electronics these days.
Apple didn't need to "borrow" opensource tech. The XNU kernel is designed and developed by Apple. They gave it away. The FreeBSD userland stuff is also free to give away. Borrowing, in quotes, implies that Apple has done something underhanded which is not the case. This is the big difference between the GPL and BSD licenses. Only one is truly free. Also, the whole modern OS thing is old now. They've owned NEXT technology for 12 years. If they didn't buy them work on their next OS would have carried on anyway. But I doubt they would need 12 years to complete it.
With the development or Arduino, Beagleboard, and Bugboard getting into technical development has never been easier. This is vastly different than 20 years ago when the most you could hope for was a box with some blinky lights from a 444 chip. Now, even novices, are creating some pretty cool stuff.
See p284 of Horowitz and Hill :-)
I agree, you can do that, but the description was fairly clearly a 555. Besides that an lm444 is a 14 pin dip and a 555 is 8 pins (see comment by yardie).
But what is available today goes way beyond the kits I had when growing up.
I remember seeing them at CES a few years ago and they were generating a lot of buzz in the tech blogs.
I do remember seeing these. They're more expensive but do look really quite polished. Certainly more attractive to look at than the tangle of wires I have hooked up to an arduino at home.
Arduino, and products like it, as others have mentioned, offers a great way for someone to get into embedded systems. You can get a decent setup for around $100US that offers way more potential than opening up an old phone or TV.
I've grown up in the era of the black box, and I was still able to find things to pull apart and figure out when I was a kid. Even better, technology had gotten to the point where things like LEGO Mindstorm allowed me to tinker with something without upsetting my parents by breaking something they used. Later, when I got into music and started playing guitar, I made some money by figuring out how to repair tube amplifiers, a decidedly old technology, but one that continues to thrive in the guitar world. I purchased several tube amps on eBay for ~$20 each. I reverse engineered them and turned them into guitar amps. I built a tube amp from the ground up, even putting together the turret board and drilling the chasis.
I do agree that there is value in keeping things open, but the argument that there isn't anything out there for the next generation to open up sounds a lot like another "Well, back in my day" story.
For my own case, had it not been for the internet, I would've never gotten into tube circuits. I never would have been able to make heads or tails of even the most basic tube amp design, even given the schematics. On top of that, I probably would've either killed myself or caused serious harm to myself if I did try to poke around those circuits without having researched it. Tube circuits can have deadly amounts of electricity in them even when they're unplugged and turned off. The same is true for old (and new?) TVs, I believe.
I'm 20. PCBs were pretty much ubiquitous by the time I was old enough to really start digging into things. That never stopped me, and I don't believe it would stop someone that's 13 now and has the inclination to take things apart and figure out how they work.
Of course there are the more industrious (and more annoying to their parents) who go on opening things that weren't broken and mess around in them. Those people are the ones who will actively go out and by arduinos. But it seems the first group is a dying breed.
I can certainly sympathise with the OP and the older generation of programmers but I can't help feel that it's more a case of standing on the shoulders of giants. If we're to forever tinker (even if it is fun) with basic electronics or low level programming we'd have a hard time making any progress. As painful as it is I find it best to accept that something is black-box and just work with it, warts and all. Having said that, it's really useful to have the source code for the frameworks/libraries you're working with. Delving into it is similar in ways to the tinking the OP speaks about.
In assembler you place every instruction, every part.
A high level language that gets compiled (or even compiled twice by a jit) outputs millions of instructions, many more than the largest hand coded assembler program.
Nobody in their right mind (or of a (in)sufficiently long lifespan) would even attempt to do such a thing manually.
Just like modern cpu designs could not be duplicated without a functioning computer make the masks.
So, 'old' programmers feel the same sense of being removed from the machine instructions and the complete knowledge as electronics people do from being removed from their components and visible, traceable wiring.
Not that long ago there was a posting here by a man who built his own computer, out of ttl logic circuitry from the ground up.
It's an excellent exercise in gaining control of the machine.
But as soon as you want to do something that is competitive in the marketplace then 99 out of 100 times you end up using that high level instruction grinder anyway.
Only guys like dark_shikari and people that are busy with smaller embedded systems get to enjoy that feeling on a daily basis. For everybody else that is just as alien as an electronics person that insists to build a complicated piece of equipment out of individual resistors, diodes and transistors.
When the first thin film circuitry appeared in the mid 70's (or at least, when I saw it for the first time, I must have been 10 or 11) the only thing other electronics hobbyists were wondering about was how they were going to repair them.
That's just as inconceivable to me today as it would be to perform surgery on in integrated circuit.
Regarding the IP issues: there is a lot more investment involved developing the software and chipsets which runs a modern cell phone compared to the rotary phone you took apart. Without the patents and copyrights protecting Apple's rights to the MacBook in front of you (and me), it would never have come to exist, because there would be no economic incentive for Apple to invest the billions that created it into R&D in the first place.
1. http://en.wikipedia.org/wiki/List_of_common_microcontrollers
Its more the DMCA and maybe-enforceable licenses you "agree" to when buying something that try to prevent you from figuring out how things work. And that comes perilously close to "thought police" IMHO.
I also remember taking apart rotary phones and realizing that you could dial them by hitting the hangup lever at the right frequency and number of times. If that had been "forbidden", I think it's very reasonable that my natural curiosity had just been rejected at an early age. This story reminds me of the TED talk about how you can't buy a chemistry set anymore.
Chemistry sets are readily available, just as they always have been [1]. The big difference is that now, video games are a hell of a lot more fun than any chemistry set.
He gave an illustration of an application of such technology for his neighborhood. The street near a school had a speed limit that was often exceeded. He wanted to find a way to bring the attention of this to the city. Purchasing some parts from Ramsey Electronics, he built a radar gun (using licensed ham radio frequencies), attached it to a post near the street, connected the output to a macintosh, and delivered plots of the speed of cars over a period of time. The city then put speed bumps there, making it a safer street.
Nowadays I mostly work with software, the interesting components are algorithms and data structures. They are abstract and intangible but just as fascinating.
Even back in the analog day, it was a top priority among manufacturers to patent their technology if at all possible and, in this sense, even the hobbyist was legally restricted by such patents. Patents, though, do not hide the technology but only restrict what you can do with it. You could learn away to your heart's content as you took it apart and tinkered with it. That is the nature of an analog world: you can put your hands on and physically inspect and manipulate real-world objects that in themselves constitute the heart of the technical innovation.
Reverse engineering prohibitions really came about as leading forms of technology began to consist more and more of proprietary software consisting ultimately of source code with binary implementations. The law accorded copyright protection to such code as soon as it took a fixed form of expression and prevented others from either literally copying it or paraphrasing it in ways that copied its substance while making minor variations. Those who had invested major effort to develop the code went one step further to protect it, however, because the binary implementations would be released to the public generally, leaving them open to inspection by any technically savvy person who might reverse engineer the product. Thus were born the modern forms of license agreement by which such software was licensed only (not sold) and by which licensees became bound by contract not to reverse engineer the software. Legally, then, one could no longer "tinker" with a product insofar as it contained such software, though nothing would stop anyone in practice from doing so (he just couldn't legally use what he discovered through reverse engineering). This reverse-engineering prohibition, by the way, has nothing to do with copyright restrictions; rather, it is designed to plug a loophole in trade secret law by which courts had ruled that software that is otherwise protected by trade-secret becomes publicly available (and hence is no longer treated as a trade secret) if it can be legitimately reverse engineered - because of such rulings, software vendors in essence said to the world, "if you license our products, you agree that you can't peek into our world of trade secrets by reverse engineering our code."
These legal changes came about to protect what early software vendors perceived to be their vital interests in guarding their proprietary advantage gained through investing their development dollars and are a default part of the landscape today.
In my view, though, it is not so much the legal barriers as it is practical factors that have turned the products of today more and more into black boxes. Leading technology today is much less analog and much more digital and the tiniest of contraptions today can contain mountains of sophisticated technological development in the most miniaturized of forms. Who, sitting in a garage, has any hope of figuring it all out in the sense of opening up the box and analyzing it without the aid of sophisticated tools that no average person can even begin to afford, much less without the aid of technical know-how that is far beyond the reach of even the average technically-savvy person?
Even when it comes to earlier forms of software, again, back in the day, many of the engineers I represented came out of the old mainframe environment and were quite technically savvy in working with miniscule resource levels by coding in assembly languages, etc. They worked much more closely with the physical machine resources than is either needed or desired today. As technology has developed over the years, however, the working level of knowledge has become more and more "abstracted" to the point where interactions with the machine-level functions are now a specialized niche with which fewer and fewer people are even concerned, much less proficient in dealing with.
So the black box is there owing to primarily practical factors relating to technological development. The Steve Jobs's of the world (or their contemporaries) might have learned from their Heathkits, but that day is gone forever. It is all too complex now at the leading edge of technological change.
http://www.amazon.com/dp/1856175073
...to people who want to demystify current generation electronics.
What we have now is not as simple as an early telephone, but it still might be in the realm of understanding.