Chuck Moore's fireside chat to SVFIG on 11/23/93. A video of this presentation in the store.
Chuck Moore's Fireside chat to SVFIG on 11/23/93
I talked about this only a month ago and not much has changed in the last month. SO I will be a little bit repetitious but not too much. First I want to show you the latest incarnation here. Here is the computer. If you remember last time it was a kludge of wires, now it is a smaller kludge of wires. It is attached to a battery, it's running off this battery. It will run off this battery for about 12 hours. I have a little Sony monitor that runs off its batteries for about 8 hours so I can seemingly endurance of any conceivable....
This is one of my latest images, a carefully crafted logo made of rectangles. If you look at it closely you will see that this picture is more noisy than the picture running on Dr. Ting's machine earlier. This is running at 4.3 Volts. It is voltage sensitive. That is sort of the remaining flaw in this incarnation of the chip. You pick the voltage you like and you better run at it. At 4.5 Volts it is pretty clean at 4.7 you get a perfect picture. And it deteriorates as you increase the voltage. Strangely as you decrease the voltage something runs better. I think it is probably the stacks.
This is the keyboard inherited from my NOVIX system and it is increasingly unsatisfactory. The microswitches seem to have a most pronounced bounce that I don't remember from the old days, but anyway, I have a color pattern, that's RS189 as accurately as I could represent it. It shows eight of the fifteen of the colors I have available. Those are the bright colors, except for gray which is dark white. I have dark versions of all those colors. The bars that Ting had up earlier and I had up last time are still in memory but are pretty inaccessible. This is the latest memory card. It is amazing how quickly I can turn out memory cards that are different than the ones before. Superimposed on this I can put up the DUMP which I could not put up last time. The DUMP is now quite reliable. Much more reliable than the color bars themselves. Based on the fact that it runs at a higher voltage. You can rank your software based on the voltage that it will run at. These run real well.
Low memory is at the bottom of the screen and high memory is at the top of the screen as it should be. I can scroll through memory. I am going up, and I can get back in into the program. This is program now. This is the address and data in binary. It is an interesting way of displaying binary. I'm tired of hex (laughter).
Also you will notice that I have more windows than Windows (laughter). Each word is contained in its own little window. I find it is a very pleasant thing. It minimizes the disturbance in the picture that you are trying to debug. You superimpose a dump on whatever you are debugging. It is easier by far to use binary for this application because what I care about on this machine is the pattern. And in this machine that pattern is particularly intricate. Numbers are represent as alternating zeros and ones so the number zero is a pattern AAAAA and so forth. (giggles) The numbers increment smoothly in the binary kind of fashion but the patters increment sort of backwards. The patterns are there in an attempt to minimize ripple carry hardware. I get request from people to eliminate that feature any time they think more than casually about it. (laughter)
The pattern on the right with alternating Fs and 0s is quite typical of uninitialized DRAM. I am scrolling at a rate determined by the program. The code on this machine is very compact. Refreshingly compact.
The major problem with this chip is that the stacks are only 3 deep. And working with 3 deep stacks is a real hassle. In working with this code I I had to save three locations in memory in order to have enough I don't think I could do anything with two. If I had four ...
Page under construction
2512 10th Street
Berkeley, CA 94710