Reader Letters — The Spectrum Market in St. Petersburg and Compatible Computer Production

NemoFAQ #06

Dear Nemo!

I don't know whether you'll get this letter or not, but so much has been boiling up inside me that I've decided to write. I'm writing in the hope of getting answers to my questions. I've been in a state of complete ignorance about Spectrums and Spectrum-compatible machines for far too long.

The first, and probably main, question I'd like answered is: what's the state of the Spectrum-compatible computer market in your region (by which I mean St. Petersburg)? For my part, I can tell you that things in our city, and in the region as a whole, are pretty dismal, since there is no market to speak of. You can sense that there's demand for home computers, but unfortunately there's no supply. I remember reading your article about the Spectrum market, in which you proposed setting up production in the regions. All in all, it's not a bad idea—and today it's actually a very good one, though of course that's just my personal opinion. In principle, I'm ready to take the risk and get into this kind of production, provided I have all the information.

I'd also like to share some thoughts of my own on architecture and circuitry, and, most importantly, hear what you think. I should warn you that I'm a programmer, not a radio engineer, and microprocessor technology is just a hobby of mine. So please don't judge me too harshly.

Let's start with the main thing—the processor.

Installing a Z180 in the Speccy was a forward-looking move on your part. It's a pity that neither programmers nor users understood it. Why not try installing a Z380? Naturally, that would mean completely redesigning the machine's circuitry, but the capabilities we'd gain should make up for the cost of the project.

Now for the memory.

I don't know how things are in St. Petersburg, but here in our city a KR565RU7 costs 20 rubles apiece—that's 160 rubles for 256 KB. This gave me the idea of using 40-pin SOJ chips from PC video cards: 512 KB in one package, 40 ns, costing $1.5–2. You could kill several birds with one stone.

The video controller is a separate subject. In principle, there's nothing that needs changing there. However, there's one drawback to the homegrown Spectrum: its video output. More precisely, the lack of one is currently a major limiting factor. Of course, there's no need to give up RGB, but a PAL composite video signal generator is worth considering. After all, no one is going to let you rewire their beloved—and, naturally, expensive—imported TV just to connect “some computer in color” (to quote the majority). It would also be good to have a video input for synchronization with an external video signal, as well as a way to overlay the computer's image onto it. That would let the Spectrum serve as a titling machine for small TV studios, and for anyone with video equipment at home.

Well, that's about it. I'll be eagerly waiting to hear back.

────────────────────────────────────────────────────────────

Dear N!

  1. The situation with Spectrums in our region isn't much better than in Russia as a whole. One city, even a large one, doesn't tell you much. The situation across Russia as a whole is laid out pretty well in “Open Letters.” Have a read, if you can be bothered.

Dear Nemo!

Fragment 2/3.

  1. And who’s going to deal with your market? Your uncle, perhaps? He could take it on, but he won’t—it’d be too expensive for you. So you’ll just have to manage on your own.

  2. Set up production in the regions. We already have some experience in this area, more negative than positive. As it turned out, the main difficulty isn’t a lack of people willing and able to do skilled work—there are plenty of those. The problem is that there are no local organizations, no professionals. The public’s attitude toward this undertaking is naively amateurish, when the main difficulty is how to sell something, not how to make it. An attempt is now being made to change the terms of the problem somewhat. Specifically, to try to produce locally fairly simple but necessary devices for the Spectrum. The base boards are just a kind of testing ground—a way to figure out how things work, learn how to work together, and establish contacts. If this works out, we can barter devices. Then everyone will have an interest in selling both their own devices and other people’s. As for “being fully informed,” that is limited by time, both in terms of delays and volume.

In addition, the regions have certain difficulties with components and advertising support. In St. Petersburg, for example, PC junk brought in by the nouveau riche is piled knee-high. And if we’re short of anything, we get it abroad, at the Finnish dump.

  1. Z180. Progress isn’t always a good thing, and sometimes it’s even bad. If the market and users aren’t developed enough to match the technical specifications, then no matter how wonderful those specifications are, the effect is negative. Specifically, mass psychosis and machine worship. A sort of pagan cult of Papuans, whipped up by television and the yellow computer press. The efficiency and real returns of the technology, with all its fanciest technical features, are zero. As for the Z180, I can say that I don’t yet see any tasks or real applications for Z180-based architectures. Why build electron microscopes if they’re going to be used to hammer in nails? It’s inconvenient, to say the least. Hypertext systems and shells are sluggish, the OS is making painfully slow progress, and copyright is ignored. Ideologically, the Z180 is designed for 1 MB of RAM. With the current methods and ways of organizing work, it would take 15 years to write a program of that size. Users won’t wait 15 years; on the contrary, they’ll have thrown out their computers five times by then. In my estimation, new software is created at a rate of 2–3% of what’s already available per year, and the computer fleet is replaced at roughly the same rate. Given that, the architecture can’t be changed any faster than 3% per year either. THESE ARE OBJECTIVE LIMITATIONS.

“The capabilities this provides.” You mean the technical specifications, right? In reality, these “capabilities” are often impossible to put into practice. They are nothing but ILLUSORY consumer values. The development of the market determines the level of technical resources/specifications that can be used effectively, not the other way around. This is that same stifling PC chimera being foisted on us from across the ocean, with the help of our own dear nouveau riche. A pagan shrine with the idol of “Progress” in the middle. More “Potemkin villages.” Better

Dear Nemo!

Fragment 3/3

…might as well have planted corn or something—there’d have been less harm. All this applies mainly to IBM. But it applies to the Spectrum, too. Users are regularly irradiated by their TVs and then, like zombies, repeat: “One megabyte… One megahertz… One pixel…” Some even get a sweet taste in their mouths—that’s how far it goes…

5. “Now, about memory.”

The 20 rubles per chip it costs in your city isn’t the price of the chip itself, but: 1) the fact that it’s being sold; 2) the cost of finding a buyer; 3) the cost of getting it there. The wholesale price of this IC on market day is 2–3 rubles. In fact, the IC is even cheaper, but you have to: 1) find it; 2) bend down and pick it up out of a garbage bin. Generally speaking, components make up a far smaller share of a computer’s price than users tend to think.

As for “40-pin SOJ chips from PC video cards,” the problem isn’t their price but their QUANTITY. It’s unlikely you’ll manage to scrape together enough stock to manufacture anything. We’ve been through this before, around 1992–93. We put 556РТ14(15) chips in the floppy-disk controller. We really did buy the first 200 quickly and very cheaply. But later, once the DEAD STOCK on the market ran out, the manufacturer offered us some. At such a crazy price that we had to throw away the printed circuit boards—it turned out to be cheaper.

6. Video controller.

Here I can agree with you: this really is needed. But how do we make it happen? I had a talk with some people from NIIРA—we have an institute like that here; they work on televisions. They moonlight by making encoders for video cameras. We could put them in the Spectrum, too. With a steady order for video cameras, they could make any quantity, however small, for the Spectrum as well. But the trouble is that the video processors in Spectrums differ from one another—that is, they DO NOT COMPLY WITH THE TELEVISION STANDARD to varying degrees. For example, there are machines—and plenty of them on the market—that have 320 lines instead of the required 312/313. That means a standard encoder will either work badly or not work at all. And then, who’s going to solder all this onto those machines—in other words, “make it fit” on site? As they say, “failure to comply with the standard is punishable by law.” In this case, with a lifetime ban on the use of an encoder. Still, all is not lost. There’s SCART. Essentially, the signal is fed straight to the video amplifiers, just like RGB. So you get excellent quality, and it’s not sensitive to minor deviations from the standard.

Everything said above applies to video studios as well.

(c) Nemo.

╔══════════════════════════════════════════════════════════════╗
║                           ° ° °                              ║
╚══════════════════════════════════════════════════════════════╝

Share your thoughts about the article