Miscellaneous
On the Virtual TR-DOS website, I found a text file...
The ZX topic is far from exhausted.
Perhaps, in the new century (and millennium), it makes sense to sum up what has been done, what hasn’t been done, and what should be done so that the ZX doesn’t become just one of many—and, even if it won’t be first in everything, at least it won’t be dead last.
This somewhat convoluted introduction may be unjustified; I just want whoever reads this not to rush to disprove it, shout, and stomp their feet, but to think, at least a little: WHAT HAVE I DONE TO KEEP THE SPEC FROM DYING?
In general, we should probably start at the very beginning and go over the history of our beloved ZX.
Somewhere around 1982, a computer called the ZX-Spectrum appeared. By the standards of the time, it had some pretty impressive specifications: it had 48 KB of memory, a fairly high-quality screen, and a very functional BIOS, with BASIC built in and procedures for using a tape recorder as a storage device. It also had printer support (at the OS level, not through a driver for operating it). In time, it became clear that the concept of unchangeable ROM made it possible to write very complex software packages (including the most popular games of the time), while the screen structure allowed for colorful (by the standards of the time, of course) loading screens and game scenes.
This was the first stage in the ZX’s life.
Years passed. The Spectrum acquired a huge amount of software (games and otherwise) and various peripherals. As it turned out, its spread depended not only on factors such as the amount of software, peripherals, support, and so on. Its population grew significantly in the wake of certain political events (in particular, the opening of the “fence”), thanks to which it made its way to us in the former Union.
So, people who wanted not only themselves but others to benefit set about creating their own Spectrum-like computers (that is, ones that matched the branded ZX-Spectrum in many respects). This led to the ZX-Spectrum-compatible computer market in Russia growing very rapidly. Since it (the ZX) wasn’t just a gaming station, there were people who wanted to write their own programs for it—and not only wanted to, but actually did. These could be programs for calculating a resistor’s value from its color bands, programs for calculating the product of matrices, finding inverse matrices, and so on. This second stage in its life can be characterized by the ZX’s arrival in Russia (and neighboring republics), and the beginning of the growth of the programming community as such.
Various press, periodicals, and so on appeared. The Spectrum was no longer just—or even primarily—a gaming station, but a platform for beginning the training of future qualified programmers. ZX gatherings became more active, and the first commercial projects and programs appeared—programs that you could say had no Western equivalent. ZX-compatible computers appeared that really were compatible with the ZX, rather than merely being called that. New ZX machines—with more memory, peripherals, and so on.
This stage can be characterized as the ZX’s rebirth as a platform for training programmers, and simply as a computer that could teach you about computers in general.
The fourth stage is happening now.
The ZX hasn’t been a gaming station for a long time. These days, only true fans are still obsessed with the Spectrum. There has been some decline in activity among those who can…
to be recognized as true ZX fans.
How many more stages there will be is unknown. Most likely, sad as it is to admit, the ZX will vanish into oblivion, becoming just one of many that existed and left nothing behind except a few lines in newspapers and magazines.
So why did it turn out this way? Why didn’t the Spec become one of the first?
I’ll try to analyze everything I know about the ZX and draw some conclusions about how to make its life last as long as possible. I don’t think I’m the one who started the topic I’m developing, nor do I think I have the right to consider myself the last to develop it. Anyone can carry this topic forward. In fact, it would be even better if everyone tried to add their own little brick to this project—the development of the ZX Spectrum.
Perhaps this whole text should be put up as an Internet page, so that every Speccy user is SURE to read what’s here. It would also be desirable for an electronic magazine to publish this letter.
1. Hardware
Looking at the history, the first thing worth discussing is the amount of RAM currently available on the ZX. The average amount of that RAM is around 190 KB (the average of 128 and 256). There are machines with a hefty 1 MB of RAM, and even variants with up to 2 MB of RAM for the Scorpion. At the same time, as with everything else, there is a huge amount of confusion (in my view) about the ports used to access this memory. Procedures have of course appeared that can identify 90 types of memory in use, and as a rule they are used where this is really necessary: file managers, disk copiers, and so on.
There is a certain amount of confusion with ROM, too: its size ranges from 16 KB to several MB (for ROM disks). And, of course, how this ROM works depends on the ports, which were generally made to suit individual needs.
The screen. This is a whole saga in itself. It’s probably pointless to count all the headaches caused by the Spectrum screen; one telling measure of the problem is that not a single issue of the Russian-language ZX magazine ZX-ревю came out without a question about how the screen worked, or a solution to one. In the end, no storage formats beyond the basic ones *1] and the most obvious ones (well, the screen, its thirds) were developed. On the other hand, things like Multicolor and Gigascreen appeared, but again at the basic dimensions (24*32 characters). I’ve heard that extended screens were made on some computers that weren’t entirely ZX-compatible, but I suspect they never caught on, since the base was still ZX-48 (or ZX-128, if you prefer).
Sound. Here there’s plenty to be happy about.
It all started with the simple Beeper. The sounds it produced, although generally not especially beautiful, were in particular cases good enough to be called works of art (Golden axe, dizzy, etc.). With the advent of the sound coprocessor, which could generate harmonic oscillations on its own, leaving the processor only to specify what should sound and how, you could say that the ZX found its voice. The same melodies, played on the coprocessor, sounded much better. The arrival of programs for working with actual digital samples, even on the AY, brought about a revolution. Now it was already difficult for the hated PC users to make fun of…
…: “And why does your Spectrum sound like a game console?”
After that, everyone caught digital-audio fever—and luckily, working with it doesn’t require much CPU power.
That was when the first digital-audio compos appeared, showcasing full-fledged digital music. Everyone started making add-ons for digital music on the ZX—there were more than ten different types. And here, once again, in my opinion, there’s confusion both over the ports and over the devices themselves. Many of these digital-audio add-ons have considerable capabilities. Of course, I’m not talking about quite such complex technologies as EAX, A3D, and so on. In terms of the quality the ZX can deliver when playing samples or musical works, it’s certainly no slouch. One example is the DMA USC: besides playing music, it can also perform DMA, and even in a way that helps the main program run. In this respect, the ZX has shown that you don’t always need huge, unreasonable amounts of CPU power to get good results.[2]
Information storage. Just as there was something to celebrate with sound, there’s nothing to celebrate here. Yes, nobody uses a tape recorder to work with the ZX anymore (well, perhaps just for learning). But what replaced the tape recorder is generally no better. These are various ROM disks, DOSes, and so on. Once again, there’s confusion over which ports are used, and over the protocols for working with SOS (Spectrum OS or Standart OS or whatever you want to call it). By the way, as far as I know, nobody has tried to solve this problem. More precisely, there have been discussions about it (for example, in ZX-Review in 1996), but no concrete solutions have been worked out.
The truth is, it really is difficult: to get a disk (or some other) OS to work with the standard SOS, you have to break so much that very little of SOS itself will remain. That means the set of programs designed to work with the SOS ROM won’t run. On the other hand, if you use tricks such as shadow ROM and so on, you run into problems with the OS working in shadow ROM—or, more precisely, with how it interacts with SOS.[3]
The root of all these troubles is that the Spectrum’s point of pride—its multifunctional ROM (a combined BASIC interpreter and BIOS)—has now become its clumsiness. That is, even a not especially experienced programmer now knows that any program must first be structured properly, with the code separated from the data. But the 48K ROM has the BIOS, BASIC, and even a little bit of something else all mixed together.
Yes, now we know (more than 20 years later) what’s where in the original 48K ROM. But these days, it would actually be easier to write a new ROM with a proper restart system, proper BIOS error handling, and so on. The situation with DOSes is similar.
Summing up the BIOSes, the situation here is as bad as it gets.
Taking the four basic points of computer architecture together, we can give an overall assessment:
- RAM — good (4)
- SOUND — excellent (5)
- SCREEN — poor (2)
- BIOS — poor (2)
- STORAGE — poor (2)
So, the average score is 3 :(((
2. Regarding software
The main (and most commercially advantageous) feature of software on any computer or console is its use:
- In business
- In the home
- For entertainment
Miscellaneous
Of all the areas mentioned, by the end of the 1980s the Spectrum was best equipped to meet gaming needs—which means most of its software was focused on entertainment. As a result, the Spectrum also developed in the business sector, but that was a much less important factor. And given that in our country people were only just beginning to appreciate brainpower (in the sense of intellectual work), all the programs that reached the Russian market in the late 1980s and early 1990s were pirated. That meant their authors were not paid for writing them, so they did not continue working in this field, since they had no financial incentive. Thus, over time, the number of programs being written in the West declined (if measured in units per year).
Our Russian programmers, for their part, were not especially eager to fill the niche that Western programmers had naturally left open in our market. There were, of course, commercial projects that could be called successful, but by and large there were only a handful of them. Compared with the amount of standard software—that is, software that came from abroad—we can say that our gaming software never really took off. System programs were another matter. There were countless ones written, and in that respect, it has to be said, we far outdid Western programmers. Though it should also be acknowledged that, despite the functionality of many of our programs, it could sometimes take more than an hour to figure them out.
The ZX demo scene, on the other hand, belongs entirely to us, and our programmers never faced serious competition from the West in this field. Personally, I never quite understood why a team capable of writing a demo so good you could lick your fingers would not also take up developing software that could actually be sold—and more than once. In 1993–1997, you could say the groundwork was laid for programmers to make a go of it, but the results were modest. It is unclear whether this was because the same period coincided with the economic stagnation of the former Soviet Union, and thus: “...what kind of programmer am I if I can’t feed my family and I’m hungry myself?” Or did our programmers simply need more time to get going?
The field of operating systems on the Speccy is in complete disarray, too.
It seems that no small part was played by the fact that the BIOS—the thing that generally provides the basis for creating an OS—was highly inefficient, or, to put it more accurately, very weak. OSes conceived as serious projects mostly failed for the following reasons:
- Inconsistency on the part of the authors
- Unwillingness to work with people who use their OS
- Narrow outlook and lack of experience
Miscellaneous
As for the last point, I think everything is clear: few people in the former USSR could boast experience creating an OS, or even taking part in creating one. As for the second point: I think that’s clear too. PR (public relations) isn’t taught at all universities even now, let alone back then...
As for the first point: often an OS just came about by itself. Say, someone created a shell for working with files, and it turned out to be so functional that you could call it an OS (examples? Heh... iS-DOS, for one...). At the same time, what happened was that a so-called OS borrowed certain functional elements from other systems and applications that its author considered important, while completely omitting others which were, in fact, the important ones. So despite the huge selection of OSes for the ZX, there’s actually no alternative at all *4]. I saw about 10 Windows-like shell systems in some e-zine (a review of them)—what a thing to clone, honestly...
As for programming languages: well, if there are no OSes, what’s there to say about languages? Still, it’s worth pausing here. The Spectrum has found itself in a paradoxical situation: Assembly is the main accepted programming language, and accordingly there are a large number of Assembly platforms (about 20 different assemblers), while you can count the translators (let alone compilers) for high-level languages (C, Pascal, BASIC, etc.) on your fingers—and they date back to the last decade. And since ASM is the only programming language that’s alive, the result is that the only things programmers create are demos and system programs *5]. So, taking the weighting factors and ratings into account, we can conclude that the Spectrum’s software situation can be rated somewhere between poor and satisfactory (2.5).
In total, if we look at the combined score for hardware and software, we get 2.75, which doesn’t even qualify as satisfactory.
This score is objective, and yet it can explain quite clearly why even those who used to be its passionate fans are abandoning the Speccy.
P.S. We need to do something. At the end (below, I mean :)) I’ll present a few thoughts that are entirely my own, which are worth considering and, of course, discussing. They’re all my ideas, though I’m not sure I won’t accidentally turn out to be a plagiarist :)))).
P.S. I almost forgot to leave my contact details.
My name is Valera. Nick: GriV.
Email: value@kampi.bancorp.ru
I’m on ICQ, 178532526.
I think this is a lively topic, and the letter won’t go unanswered—maybe with an angry reply, maybe with a proper, constructive one.
P.P.S. If there’s no response at all (I don’t believe that, but still), I’ll consider the Spectrum dead and myself the only Speccy user amid a stormy sea of PC users, Mac users, and so on...
P.P.P.S. Spectrum forever for us.
Share your thoughts about the article