The C64 Across Borders

One machine, six national presses, and no two agreeing what it was

The Commodore 64's specification was identical everywhere it was sold. What differed was everything else — when it arrived, what it competed against, and whether it mattered at all. Measured across the general computer press of Italy, Brazil, Greece, Poland and Yugoslavia, the C64 is nowhere the leading machine, and the machine that leads is different in every country.

commodore-64commodore-64non-anglospherecomparisonmarketsmeasurement1982–present

The Commodore 64 was the same machine everywhere. Same processor, same 64K, same SID, same VIC-II. Any account of it built from one country’s press will describe that machine accurately and still be wrong about what it was, because everything that made owning one different from owning something else was local — and only the local press wrote it down.

This entry compares the C64 as it appears in six national presses held in this project’s library.

It did not arrive everywhere at once

Only two magazines here were publishing before the C64 existed, so only two can say when it turned up.

	Publishing since	First names the C64
MC-microcomputer (Italy)	September 1981	March 1982
Micro Sistemas (Brazil)	October 1981	July 1983

Italy has it before it shipped — March 1982 is the announcement, not the product. Brazil takes until July 1983, eleven months after buyers elsewhere could have one.

And the manner of the Brazilian arrival is the useful part. It is not a review, a preview or an advertisement. It is a reader’s letter about a Commodore users’ club, from a correspondent who belongs to the Toronto PET Users Group. The C64 enters Brazil’s leading computer magazine as something a reader has heard about from abroad.

That is what an import-substitution market looks like from the inside. A machine you cannot legally import is not news.

Nowhere the leading machine

Counting how often each machine is named across the text of each title, in the general (not single-platform) press:

Country	Title	Leading machine	C64 share
Italy	MC-microcomputer	PC 33%	16%
Brazil	Micro Sistemas	PC 39%, MSX 29%	outside top five
Greece	Pixel	Amstrad 40%, Spectrum 38%	6%
Poland	Bajtek	Spectrum 39%	13%
Yugoslavia	Računari	PC 40%, Amstrad 25%	5%
Yugoslavia	Svet Kompjutera	Spectrum 29%, PC 28%	19%
Hungary	Mikroszámítógép (1984–89)	PC 25%, C64 23%	23%
Czechoslovakia	Mikrobáze (1985–89)	Spectrum 33%, PC 24%	4%

The C64 leads in none of them — but look at the two Yugoslav rows before trusting any single one.

One magazine is not one country

Računari and Svet Kompjutera were both general Yugoslav computer magazines, both published in Belgrade, both measured over 1984–87. They do not agree.

	Računari (n=540)	Svet Kompjutera (n=1,433)
PC	40%	28%
Spectrum	17%	29%
C64	5%	19%
Amstrad	25%	15%

Nearly four times the C64 share in one than the other, from the same country in the same years. Računari leans technical and business; Svet Kompjutera leans toward games. Both are “the general press” and neither is wrong.

Poland now shows the same thing. Bajtek, a supplement to the official youth newspaper, gives the Spectrum 34% and the PC 15%. Mikroklan, which began as an inset in the professional journal Informatyka, gives the PC 58% and the Spectrum 15% — same country, same years. The divergence tracks the magazine’s parent institution, not its nation; see institutional computer press.

⚠ This is the most important caveat on this page, and it was found by accident. Yugoslavia was the first territory here represented by two titles, and Poland is now the second. Every other row — Italy, Brazil, Greece, Poland — rests on one magazine, and the Yugoslav pair shows how far two magazines from one country can diverge. The rows should be read as what this title covered, not what that country used.

The comparison survives that, and this has now been tested rather than assumed.

The two load-bearing results, checked against second titles

Italy, Brazil and Greece each hold further titles in this library. Measured over a wider 1982–90 window, they say:

Country	Title	Leading machines
Italy	MC-microcomputer	PC 32%, Amiga 23%, C64 14%
Italy	K	Amiga 46%, C64 21%, Spectrum 13%
Brazil	Micro Sistemas	PC 37%, MSX 31%, Spectrum 13%, ZX81 12%
Brazil	CPU	MSX 93%
Brazil	Input	ZX81 37%, Spectrum 36%, MSX 23%
Greece	Pixel	Amstrad 41%, Spectrum 25%
Greece	PC Master	Amstrad 65%

Both load-bearing results hold. The C64 appears in no Brazilian top four across three independent titles, and MSX is prominent in all three. Both Greek titles lead with the Amstrad, the second more emphatically than the first.

Italy’s two titles differ on ordering — K is an Amiga-era magazine — but agree the C64 sits third or lower.

⚠ These are not all general magazines: CPU is MSX-focused and Input Sinclair- focused, which is why their spread is so wide. What makes their agreement useful is that titles with opposite platform loyalties still fail to put the C64 near the top in Brazil. The window here is also wider than the table above, so these percentages are not directly comparable with it.

No single percentage on this page should be quoted as a national figure. The patterns that survive multiple titles can be.

Two of those results are worth sitting with.

Brazil’s second machine is the MSX, at 29% — ahead of the Spectrum and the ZX81, and far ahead of the Commodore. MSX was built in Brazil under licence by Gradiente and Sharp. The ZX81 and Apple II, the next most prominent, were the machines Brazilian firms cloned. The pattern is not a preference; it is a list of what could legally be manufactured.

Greece leads with the Amstrad, at 40%, ahead of the Spectrum — an ordering that does not hold in Britain, where the Amstrad ran third behind Spectrum and Commodore.

One machine, three prices, one country

The clearest single artefact of the whole comparison is a price column in Bajtek. Its February 1989 table carries three headings over the same list of machines:

GIEŁDA „BAJTKA“	PEWEX / BALTONA	RFN
thousands of złoty	US dollars	Deutschmarks

The open market, the hard-currency state shops, and West Germany. A British magazine printed one price because there was one way to buy a computer; a Polish one printed three because which price applied to you depended on what money you could get.

A C-64C is listed at 650 on the market that week. The same table puts a Spectrum 48K at 90–190 DM across the western border.

Machines confined to one market

An earlier version of this section claimed Britain, Italy, Greece and Spain had “no domestic machines” in their press. That is plainly false — the Spectrum, the BBC Micro and the Amstrad CPC are British. The real distinction is not domestic versus foreign but traded versus confined, and it is testable.

Counting where each machine’s name appears across the territories held:

Machine	Mentions	Where they are
Galaksija	276	Yugoslavia 100%
PMD-85	91	Czechoslovakia 97%
Didaktik	36	Czechoslovakia 94%
Enterprise	33	Hungary 78%

Each appears in one country’s press and essentially nowhere else. Britain’s machines are domestic too, but they were exported: the Spectrum appears in every territory in this comparison.

So the pattern is that market-confined machines exist only where importing was restricted. Britain, Italy, Greece and Spain have none — not because they built nothing, but because what they built was traded. Hungary, Czechoslovakia and Yugoslavia each have machines that never left.

⚠ The counts for the confined machines are small — 276, 91, 36, 33 — because a machine sold in one country is written about in one country’s press. Small numbers are the finding, not a weakness in it, but they are too small to support any claim about relative popularity.

⚠ Hungary’s Primo is excluded from this table although it belongs in it. Primo is an ordinary Italian word, and a corpus-wide count returns 20,475 mentions of which 98% are Italian prose. The machine cannot be separated from the adjective by name-matching, so it is left out rather than reported wrongly.

Seven routes to the same silicon

Set the territories side by side and what varies is not the machine but the arrangement by which a person came to have one.

Territory	How the machine reached people
Britain	Imported and sold retail. One price, one channel.
Italy, Greece, Spain	Imported through a commercial market, covered by a commercial press.
Argentina	Built locally under licence. The product is la Drean Commodore 64 — the licensee’s name is in the machine’s name.
Brazil	Imports restricted, so no licence and barely a C64. The domestic industry built MSX and cloned the ZX81 and Apple II instead.
Poland	Three parallel prices — open market in złoty, hard-currency shops in dollars, West Germany in Deutschmarks.
Yugoslavia	A magazine printed a machine you built yourself.
East Germany	A state-licensed journal teaching the domestic stack — the Soviet 8086 clone, the GDR’s own dBASE.

The specification is a constant across that table. Everything else — price, legality, whose name is on the case, whether you bought it or soldered it — is local, and none of it is visible from the hardware.

⚠ These are seven cases this library can document, not a claim that there were seven kinds. Countries whose press is not held here will have had arrangements of their own.

What this does not show

⚠ Mention share is press attention, not market share. A magazine names a machine when it reviews it, advertises it, or argues about it. Nothing here counts a single unit sold.

⚠ Editorial focus dominates the number. A single-platform magazine reports its own platform by construction: Commodore Computer Club in Italy gives the C64 67%, and that says nothing about Italy. Only general magazines are compared above, and the one single-platform British title is shown alongside as a control — ZZAP!64, a C64 magazine, still names the Spectrum exactly as often as the C64.

⚠ The windows are not identical. Italy, Brazil, Greece, Poland and Yugoslavia are measured 1982–87 or as much of it as each title covers, but the titles begin in different years and no adjustment is made for that.

⚠ This library is not the world. The territories compared are those whose magazines have been scanned and uploaded, which reflects the interests of present-day preservation communities rather than the size of past markets. Countries absent from this table may have had a thriving press; their absence here is evidence about archives, not about computing.