Tuesday, February 14, 2017

the computer

We have the computers that we have.

If we were, say, more godlike, we might approach the computer as, oh, a kind of inert object.

Consider its structure: it is a box - we know something about its internal features, but only very generally - with two wires protruding from it.

Well, actually, what we know about its internal features is: it's a circuit.

Voltage, applied to one of the two wires, may, if the circuit is closed, emerge from the other one.


If, by applying voltage to the one wire, electrons are caused to travel through the circuit, they do, as they travel through the circuit, some kind of work.


Somehow the computer can tell when it has been unpowered and now it is powered up, and somehow when it detects that kind of change it knows what to do.

What does it do?


Is there such a thing as a computer with no records?


No. There is no such thing.


A computer is, by definition, a record, and a record is, by definition, records.


However ...

We are talking about a computer which is an inert object when powered up.

By this we mean, when powered up it always begins in the same state.

If we call a certain thing a session, that thing being powering up the computer, having it do some work, and then powering it down again, this computer keeps no record of anything that happened during a session.

And this is the computer you have.


But, you need to keep a record of what you did during a session. You need your computer to keep a record of what it did during a session. Now the computer is not a box, it's something inside the box, and it is connected to other things in the box that keep records.

These are now our primary concern.


I mean, they aren't.

As long as our computer is and remains powered up, in a session, it does or at any rate can maintain a record of, let's say, things it might need to know ... from earlier in the session.


This may be your computer, but it is not the computer you have.

When your computer powers up it looks, in things it is connected to that keep records (even when not powered up) ... this needs more explanation ... Let's call the computer we've been discussing a core. A core can record changes. It's a collection of bits, and certain kinds of events can set those bits to either 1 or 0, and as long as the core is powered up, those bits remain in the last state they were set to, but as soon as the computer is powered down those bits return to some kind of original state, and the settings are lost. The other kind of thing in the box is also a collection of bits, and those bits can be set to either 1 or 0, but when that thing is powered down, the bits remain in the last state they were set to. Maybe there's a third kind of thing, in which each bit is simply set to a state, originally, and the bits remain in that state whether the thing is powered up or powered down, and the value of one of those bits never changes and cannot be changed. Actually, a core must include some of the latter type, and some of the first type, seems to me. The latter type, the third type, would be what the core knows to do when it powers up.


The core knows what to do when it gets powered up ... in this sense: it knows to get the session manual ... from intersession memory ... and follow the directions in it.


I don't know the details.


Let's talk about a brand new never used computer.

You power up the core and it gets the manual from intersession memory, and follows the directions, but ... there has been no previous session. Well, maybe there has been. It depends on how we define a first session. Here we're defining it as you starting a new computer for the first time.

The manual is already, somehow, in intersession memory.

The computer gets the manual, and then it walks you through a setup process, for first sessions, and then it gives you a screen with options on it, and these options allow you to start to put data in intersession memory.

That is the computer you have.

Now we can talk about putting data in intersession memory.