Hallo,
I have been wondering about this for a while; how do you program a computer? I see how its done as soon as you have something to type on, but how do you go from having a lot of logic gates to being able to install a small program?
Printable View
Hallo,
I have been wondering about this for a while; how do you program a computer? I see how its done as soon as you have something to type on, but how do you go from having a lot of logic gates to being able to install a small program?
Back in the old day, you have to use plug boards, then they programmed a plug board to read punch cards, and then they used punch cards to program ROM's, then they programmed ROM's to read the prograsmmign from tape. Then they used the tapes to read data directly from keyboards and save them to other tapes. Most people dont appreciate that computers have been around for over 100 years.
It all depends on your definition of "computer", and whether you are talking about the concept or an actual, working implementation.Quote:
Most people dont appreciate that computers have been around for over 100 years.
Start here and keep following the links. It should help.
Well, the first turing complete general purpose reprogrammable digital computer is over 65 years old. http://en.wikipedia.org/wiki/ENIACQuote:
Originally Posted by laserlight
So it doesnt take much to expand the definition of computer to older designs. Such as http://en.wikipedia.org/wiki/Jacquard_loom
True, but whether these "older designs" are computers in the modern sense is debatable. For example, the abacus is an extremely old design of an aid to human calculation, and this is an important step towards modern computers which aid human calculation in far more powerful ways. So, the abacus can be considered a precursor to the computer, but is it a computer in the modern sense? Likewise, is the Jacquard loom a computer, considering that it does not do actual computation, though its mechanisms have implications in early computing concepts? I would rather draw the line at ENIAC and its contemporaries as the first computers in the modern sense.Quote:
So it doesnt take much to expand the definition of computer to older designs. Such as http://en.wikipedia.org/wiki/Jacquard_loom
1937 is the date of the first digital computer. I cannot remember the name though. It was called computer back then as it is today.
Anything earlier than that, I personally don't call "computer", despite what others (more instructed even) may say. I tend to attribute that to an undesirable over generalization of the term. I wish there were more intent in being precise. They were mechanical devises at most.
http://en.wikipedia.org/wiki/Z3_%28computer%29
older still
That's 1941, not 1937, and certainly not before 1908.Quote:
What about this programmable device?
http://en.wikipedia.org/wiki/Jacquard_loom
abachler mentioned it too.
It's a loom, Salem. Not a computer. Jackard Loom. A precursor, a source of inspiration. But its "just" a loom. And an heck of cool invention for its time I'm sure. A mechanical device that made it easier to draws different patterns without much human intervention.
It's not my place to contradict some of the possible opinions on this matter, not crediting myself as an expert or historian. But I do it anyways.
Calling these things computers is not different than calling car a horse driven a wagon. The evolution from the latter to the former was also a natural process and the two always shared similar objectives; to move people from one place to another faster and more comfortably. And yet we don't speak of a wagon as a car. We could extend this to planes and call them cars too. Or bycicles.
I believe it would be possible for many historians to keep an analytical eye and still avoid oversimplifications. Personally I always felt calling these devices computers always took more from them than it gave. They were wonderful inventions that any of us here would struggle to come up on his own without few, if any, backup information to start with. And yet tagging them as computers makes them look like crude ancient artifacts that were just earlier attempts of a weaker mind.
They were not. They were full fledged inventions meant to solve a specific problem at their time.
I thought the thread was about programmable "things".
s/card/paper tape/core memory/DRAM/....
It's all just
- get the next symbol from the storage medium
- do something
- rinse and repeat.
The only thing which has really changed is the number of symbols in the store, and the speed with which they can be processed.
Or do you consider programmable devices as only things which involve electricity?
What about Babbage's difference engine?
Will your awesome quad-core machine look as impressive in 200 years?
Or will everyone roll around laughing at how slow it is, how big it is, and how much power it consumes, and 1000's of them together still don't have the mental capacity of an ant.
Quiet good arguments. But then don't ask me. Ask the writers of the article for instance. Nowhere they call it a computer. You'll also find the exhibits don't call them computers.
Edit: In fact scratch that. You just addressed my issue. It's different points of view. We disagree in that I think the terminology should be less simplistic. If a computer is indeed just a programmable device then I'm afraid my current keyboard is a computer, and I do coffee in a computer and cars aren't just cars, but instead computers. And we fly in computers and, more interesting, a ladybug is a computer.
In 200 years things may as well have different names. That has been always like that. A wagon used to be called a car too. Not anymore.
I think that we need some clarification from h3ro. This thread is about "how do you program a computer?" from scratch. So, we need to know what does h3ro have in mind with "computer", and then whether it is from a historical perspective or with respect to current technology only, or both.Quote:
I thought the thread was about programmable "things".
I consider the use of electricity to be one of the aspects of a modern computer. It certainly need not be one of the aspects of a programmable device.Quote:
Or do you consider programmable devices as only things which involve electricity?
I love it when people write my name. It always makes me smile to see the goofy name.Quote:
I think that we need some clarification from h3ro.
My problem is that I dont see how a computer (or anything else with ICs) is programmed.
I understand how given one input (eg button pressed) you get an output based on a logic circut. But how do you go from having a mechanical input to having a program as input?
What?Quote:
I understand how given one input (eg button pressed) you get an output based on a logic circut. But how do you go from having a mechanical input to having a program as input?
If you can process logic in a circuit then you can also cause voltage to flow or not to flow to another circuit via an IC. There isn't any difference between a standard switch and an IC when it comes to turning something on and off. Usually mechanical motors that will interface with computers are already pre-built and designed to be compatible with PCs or with the board/CPU that will be used to control the machine.
So normally something like:
m_pPaintArm->Rotate(.625f);
Would be sent to the driver that interfaced with the hardware. The driver would then interface with the IC that controlled the arm which would then interface with the motor that actually rotates the arm. As long as all components are using the same protocol and interfaces at the software level and as long as everyone can talk to each other on the hardware level you can control any type of machine from software. If you do some googling you will find that many of the ICs come with a C compiler, assembler, assembly refs, C libraries, etc. All that is left for you to do is buy the EEPROM burner which can be very expensive. Most of the EEPROM burners interface with your computer via USB but some older ones still use the serial port. They are not by any means cheap - google for some Phillips ICs and you will see what I'm talking about.
I'm not sure what you are confused about.
I think he wants to know how to bridge the step between a few logic gates and the highly complex chips of today, which execute complicated instructions, reschedule them, predict jumps, preload caches, etc.
Then I woudl like to know what h3ro's level of electronics knowledge is. Do I need to start at explaining how electrons work, or can I just assume you know the difference between a D-type and a J-K type flip flop.
You can take this course and study it yourself.
I think I understand what h3ro is asking. Assume you have a small microprocessor, some pins on that processor are used to address memory, and another set of pins are there for the data. You connect these pins to some kind of memory chip. Then when the processor resets it will put some binary number on these address pins, that will cause the memory to put some data on the data pins (there is more going on but I'll try to keep it simple.) That binary number that is on those data pins is some kind of operational code. This code is 'wired' into the processor to do something. Then the address pins might be incremented and the next value read from the memory and that instruction would be executed, and off we go.
Now how do we get this program into that memory? Basically you wire up some kind of circuit that accesses these same pins on the memory. You put address 0 on the pins (all off) and then put the data on the data pins (with switches or whatever) and then strobe (just another pin with voltage) the memory. That will write that data to the memory at address 0. Then switch to address 1 and repeat. We actually had to program a small microcontroller (uC) like this in college. It was very tedious and it took 20 minutes to enter a program into the uC that would flash an LED.
The above comment about the punch card reader is just a way to mechanically automate this process. Once you have automated the process a little bit then you write a program to automate it a little bit more, so then you can move from punch cards to programming little microprocessors to read from a magnetic tape, and then from a floppy disk, then from a hard disk and then you simply use each evolution in the tools to make better tools.
If this isn't what you were asking then I'm sorry for oversimplifying.