‘In Hungary, they told us that the workers
would never own the means of production.’”
Dr. Charles Simonyi
We have learnt in the last few weeks that :
- technology drives sales
- availability drives sales
- restriction reduces sales
- lack of digital availability increases file sharing activity by 39%
Today we are going to change tack a lttle……
Presenting an alternative Viewpoint on how file sharing might HELP the software industry and the Computer Industry.
For the purposes of disambiguation, this article is about file-sharing.
I smile every-time I read how software file-sharing is damaging the computer business, the software business, the game business.
This is a business where the entire industry was built on the basis of sharing, swapping exchanging of ideas, code, programs and operating systems.
Microsoft’s famous PC Dos came from code that was thrown in a rubbish bin, recovered, copied and then later “licensed”. Had someone…. Not “stolen” the code from the rubbish bin, where would we all be today?
The idea that file-sharing could kill the software industry is quite laughable.
In the seventies, software code usually traveled on punch-tape and later floppies.
Unfortunately only universities and banks had punch code printers and readers therefore unless your room-mate was a punch code operator…. The only alternative was to join or form a Computer club. Which started popping up everywhere with the release of the fist hobbyists DIY kit, the Heathkit Altair computer. Operating system? BASIC. No, not a basic operating system, a programming environment called BASIC.
In these computer clubs, the currency was code, sometimes on punch tape, occasionally on a floppy disk but more usually, a “gosub 20, do something really cool, beep, run” written on a scrap of paper or the corner of a coffee shop napkin.
Coffee shops ? Of course we had coffee shops in the seventies. Where do you think all the computer clubs started?
So Koltai, is that when sharing become piracy? Not yet…. We need to sell a few computers first – there has to be something of value created before anyone wants to pirate it commercially.
But before we can create a “clone” we first have to hack (or clean room develop a copy) the IBM BIOS, which judging the amount of “Taiwanese
Clone” PC’s hitting the market in late ’81 was apparently achieved by several companies in Taiwan.
From Wikipedia
In the USA, Columbia Data Products (CDP) introduced the MPC 1600 “Multi Personal
Computer” in June 1982. It was an exact functional copy of the IBM PC model 5150 except for the BIOSwhich was clean roomed. IBM had published the bus and BIOS specifications, wrongly assuming that this would be enough to encourage the add-on market and prevent unlicensed copying of the design.
The trade in under the counter computer software began with the release of the first “clone” computer.
“Pssst, Buddy, wanna copy of PC Dos 1.0 ?”
There was no choice for early purchasers of Clone computers.
PC Dos was reserved for IBM.
The alternative was the (perceived) dumbed down version of Dos sold by Microsoft as MS Dos 1.0. (.exe instead of .com) and different
utilities. That’s if you were lucky and lived in a town with a computer store.
IBM Processor 8088 architecture = PC-Dos
Clone Computers 8086 architecture = MS-Dos
Of course in the early days, new operating systems were being released every six months.
QDos, MS-Xenix, Z-DOS, PICK, Beos, CPM/86, MSX Dos1.1, MS-DOS 1.25, PC Dos 1.1, ITT DOS, PC Dos 2.0.
Copies of each were distributed at each computer user club
meeting as we, the users struggled to work with incompatible offerings and restrictions that each operating system and processor platform attempted to force onto us.
Of course this meant that the executables created on one operating system, failed to work on another. Although this was in reality a “feature” of the chipset, in the early days, there were no X86 ports across all processors.
You selected your processor, your operating system and your compiler. Then you wrote your own spreadsheet, word processor and Database,
compiled it and started to be productive.
I must say, it was rather nice to be able to take a hunk of hardware and control how it interfaced with the user by the selection of operating system that came either on a tape or floppy.
The core memory required for MS Dos was only 23K. (For PC Dos it was only 16K).
Whilst users freely swapped programs (they had written themselves) and operating systems as if they were valueless. The reality was that the computer market was small. A PC retailed for over $2,000 dollars and the closest you could get to operational software was Wordstar and 20/20 spreadsheet.
However, every computer reseller offered purchasers a compiler suite at the point of sale.
At first it was the “upgrade” option to include the BasicA EEPROM.
Then it involved to include a variety of compilers, the most popular of which was Turbo Pascal from Borland.
A typical computer sale transaction…..
“Here you’ll need this to make some programs.”
“You mean there aren’t any programs that come with the computer?”
The computer salesman was starting to feel he was loosing the sale…..
“Look, I’ll tell you what, I’ll throw in a “copy” of IBM’s PC Dos. It’s much better than MS Dos. Let me just make you a copy…..”
So software copying in the computer world started for two reasons…
1. To enable the sale of hardware to more persons who would write more software to supply more users with more reasons to buy a computer.
2. A sharing environment where everyone felt sorry for their “computer buddy” who had just forked out so much money for the hardware.
Both reasons are understandable and arguably the basis of a technology in Genesis mode.
Everything changed with the release of (Digital Research) DR Dos in 1986. Dr Dos was seen by the market as a real alternative to MS Dos.
It would allow companies to run multiple “Dos” sessions on a single 80286 and provide users with just dumb (VT-220) terminals;
Suddenly, operating system piracy headed the news. Apparently, DR-Dos made the same kind of system calls to the hardware as MS-Dos and it’as interface “looked like the Parc Xerox GUI which had been copied by Apple”.
Of course, that particular foray failed to getup because all software was forced to address the hardware in a certain fashion if it expected software to be truly portable.
But the piracy meme was successful. The lawyers won the first round. Their efforts at smearing Digital Research worked. Suddenly a pefectly good competitor to Microsoft was wiped off the face of the computer world.
Dealers were told that they could stock EITHER Dos 2.11 OR DR Dos, but not both.
Of course, nothing was ever put in writing, so we only have my empirical memory retelling the annecdote.
This won’t mean anything to most people, so jump down five lines…. Geeks [over 45 - read on] but in January 1988, we had Rakon UNIX running virtual terminals with DR Dos sessions on OPUS VT220 terminals runing Dos porgrams like Q&A, Lotus 123, Clipper compiled DBASE III accounting software and not just the one terminal, the Espirit OPUS terminals supported two 19.2K serial connections (allowing us to toggle between two virtual server Dos sessions – 1988 !!!!). All data was on the central server, there were no floppy disks drives of CD-rom drives the user could access….. So in all respects it was a server model better than a stand alone PC. You could even play low res Dos games (Chess).
Was I angry that DR Dos was suddenly pulled off the market? Yes. The world lost the most potenitally successful challenger to the Microsoft empire.
Why? Because Microsoft paid some lawyers a lot of money.
However, the legal action came back and bit them where it hurts in April 2000 with the April 2000 ruling of Judge Jackson, U.S. District Judge who in his ruling said:
The facts of this case also prove the elements of the forced bundling requirement. Indeed, the Supreme Court has stated that the “essential characteristic” of an illegal tying arrangement is a seller’s decision to exploit its market power over the tying product “to force the buyer into the purchase of a tied product that the buyer either did not want at all, or might have preferred to purchase elsewhere on different terms.”
and….
The Court has already found, based on the evidence in this record, that there are currently no products – and that there are not likely to be any in the near future – that a significant percentage of computer users worldwide could substitute for Intel-compatible PC operating systems without incurring substantial costs. Findings ¶¶ 18-29. The Court has further found that no firm not currently marketing Intel-compatible PC operating systems could start doing so in a way that would, within a reasonably short period of time, present a significant percentage of such consumers with a viable alternative to existing Intel-compatible PC operating systems. Id. ¶¶ 18, 30-32. From these facts, the Court has inferred that if a single firm or cartel controlled the licensing of all Intel-compatible PC operating systems worldwide, it could set the price of a license substantially above that which would be charged in a competitive market – and leave the price there for a significant period of time – without losing so many customers as to make the action unprofitable. Id. ¶ 18. This inference, in turn, has led the Court to find that the licensing of all Intel-compatible PC operating systems worldwide does in fact constitute the relevant market in the context of the plaintiffs’ monopoly maintenance claim.
A search of literature (computer magazines) in the early eighties finds that it was the lawyers drumming up business that started talking about Piracy and Trademark infringement. It would appear the “Piracy” scare was so that Microsoft could implement per CPU licensing with PC manufacturers.
Interesting. I think Mr. Tom Chan was telling us that essentially there were two type of customers.
Those that wanted the DBASE II /DB-III with manuals and those that couldn’t afford to buy the product with manuals. (Remember, this was before the great internet rush of the mid nineties.) And……. Because he was referring to ASIA and the wages in the Asian countries in the early eighties was pretty abysmal, he has also told us that poorly paid persons in Asia could afford to buy PC’s. At $2500 per IBM PC this didn’t seem very likely.
Anecdotally, I had a computer shop in Darwin in the mid eighties (83-89).
Empirically, for every Toshiba, ITT, IBM, Olivetti, Adler, Altos, CCI, brand name system that we sold from the systems division, we sold at least 30 no name brands from the PC Division next door.
The Comsol Sales Manager John, with his two capable assistants, Linda and Fiona. Notice the Toshiba on the left next to the 455 QUME Graphic terminal and the el cheapo clone on the right.. The photo must have been after 1985, there are 12 function keys on the keyboard although the screen on the clone was a monochrome yellow cursor on black background. (I can’t remember the sales managers last name, but I
remember the monitor specs……very very sad)
We were mainly a Unix house which meant that we pushed high end systems. But in those days you could do an awful lot with an ITT 286, a copy of Xenix 2.11 and 32 dumb terminals. (Telecom 1100 faults system, so what does a $56,000 80286 with a small database on it look like?)
At first, the no name brands from Taiwan came with no BIOS chips installed, so we had to cook our own BIOS. (The easiest way to cook a bios was to take an EEPROM, Burner and…..).
Every computer store had an EEProm burner. Eventually Taiwan licensed the Phoenix BIOS and the trickle of clones became a flood.
When I was researching this article, I decided to total up the various numbers of computers on a global basis. From various sources, I discovered that the world passed one billion computers in 2006.
I was confused. I thought we passed one billion computers in about 2000; (Legard (2002) suggested that according to an IDG press release, we passed a billion units in 2002.
Then I saw the stats that the 2006 guestimates were based on…..
| | 1990 | 1995 | 2000 | 2003 | 2005 | 2008 |
| USA PC Sales (#M) | 9.54 | 21.4 | 46.0 | 48.3 | 55-57 | 62-63 |
| Dell
Computer: USA (#M) | 0.143 | 1.15 | 9.7 | 15.5 | 18-20 | 23-24 |
| HP &
Compaq: USA (#M) | 0.48 | 3.87 | 13.3 | 10.5 | 12-13 | 14-16 |
| WW PC
Sales (#M) | 24.2 | 58 | 129.6 | 149.3 | 180-182 | 221-225 |
| Dell
Computer: WW (#M) | 0.2 | 1.93 | 15.1 | 25.2 | 31-33 | 40-43 |
| HP & Compaq: WW (#M) | 1.04 | 8.2 | 28 | 24.4 | 27-29 | 34-37 |
…and
| | USA Million Units | WW Million Units | USA Value $B | WW Value $B |
| 1981-1985 | 3.8 | 5.7 | 10.5 | 16.9 |
| 1986-1990 | 28.1 | 60.3 | 76.4 | 181 |
| 1991-1995 | 64.3 | 172 | 153 | 447 |
| 1996-2000 | 162 | 444 | 335 | 1,010 |
| 2001-8/2006 | 267 | 855 | 424 | 1,440 |
| 8/1981-8/2006 | 580 | 1,540 | 998 | 3,100 |
| http://www.c-i-a.com/pr0806.htm
|
Totally ignoring the Chinese Mexican and Taiwanese contributions to the computer industry’s early days.
In China, the Apple II had been cloned by 1981 (Denis 1989 P.355) with copies (called “The Venus” being sold for around $2500. IBM PC
knockoffs were available for about 7,000 with inventory quantities growing at around 20,000 per month. (This was the official machines.)
The problem was that the Government was in a quandary.
Computers appeared to be important to the west and in fact the seventh “Five Year Plan” was all about building a semi-conductor industry in China. However, wages were lower than the cost of training office workers, so in the early eighties, the computers were made and stored, unused for the most part, waiting
for the compsci graduates to come rolling off the university machine to take advantage of China’s new rapidly growing VLSI IT production capability.
Meanwhile… the unofficial machines were being built on 51 second hand manufacturing lines imported from France in 1983 and these appeared to add another 100,000 to the official number of PC’s in 1984 – yet these machines shipped out via Taiwan based industry contacts and were redistributed via Singapore. Few of these computers saw the USA, most were shipped without DOS, without a BIOS or even a floppy drive. They were intended for the budget PC Dealers market where everything was an add-on option…..
China developed it’s own Chinese character language set for DOS in 1985 however did not opt for Microsoft’s generous $9.50 per cpu licensing agreement, preferring to let the reseller vendors worry about
sourcing an operating system.
The reality was that that the average Chinese worker was paid around 300 US Dollars per annum. Therefore the home PC craze was never going to take off unless the price of PC’s including their operating systems came down dramatically.
In summary, the production of all computers in the USA, according to the Computer Industry Almanac;
PCs based on the original IBM PC architecture have sold 580M units in 25 years worth nearly $B 1,000 in the USA. Worldwide sales of MS-DOS and Windows PCs since August 1981 have surpassed 1.5B units with a value of $B 3,100. The revenue figures include initial hardware sales only and exclude PC software and services.
China wasn’t the only country playing the two loading docks for every manufacturing factory. Mexico appeared to be shipping a lot more PC’s than the officially announced Trade numbers. The Pan American Highway allowed the discreet shipments to head south whilst the official shipments headed North.
By our AP (don’t ask but it’s a pluck guestimate) we estimate over 3.9 billion Official PC’s have been shipped with almost 2.1 billion retired. Retired?
Anecdotally, since 1982, I personally have retired …..(i.e.: placed into storage or junked – not resold): Australia only.
| Koltai’s PC’s | Sourced | Retired |
| ATT&T 3B2, | 1982 | 1986 |
| IBM 5120 | 1982 | 1987 |
| IBM AT | 1983 | 1988 |
| Ferranti Challenger | 1983 | 1983 |
| Olivetti 80186, | 1983 | 1983 |
| STAR AT, | 1984 | 1987 |
| UNI-x Clone, | 1983 | 1986 |
| Adler 80186 | 1984 | 1987 |
| ITT 286 XL | 1983 | 1991 |
| DRS 30, | 1991 | 1994 |
| DRS 50, | 1992 | 1994 |
| DRS 300, | 1989 | 1994 |
| Fujitsu Pearl Laptop | 1994 | 1997 |
| DRS 3000, | 1991 | 1995 |
| Toshiba 4800 | 1987 | 1991 |
| Compaq Laptop | 1997 | |
| Sun 5 Netra | 1995 | 1999 |
| Sun 20 | 1994 | 2000 |
| DRS 400 | 1989 | 1995 |
| ACER 30 | 1991 | 1993 |
| AST 210 | 1998 | 2002 |
| Sony FS….. | 1998 | 2002 |
| Sony C1V | 2001 | 2006 |
| Sony F505 | 2004 | 2005 |
| Pentium 2 clones (3) | 2002 | 2003 |
| Pentium 3 clones (4) | 2002 | 2003 |
| NEC Versa | 2004 | 2007 |
| Sony VGNA-190 | 2003 | |
| Dell GX-620 | 2007 | |
| EEPC (2) | 2009 | |
| Unlisted Computer 1 | 2009 | |
| Unlisted Computer 2 | 2010 | |
Green highlighted items are still in my possession. (This includes only one of the PC’s from my office, the rest belonged to the company.)
The AT&T, DRS400 (which by the way was the X.400 gateway to Australia for nine months, operating from my home garage converted office), and Sun 20 are really backup work machines that I had at my residence to duplicate and backup office procedures and data. So possibly they shouldn’t be counted; on the other hand, as I paid for them and not some corporation, they get included.
The Compaq laptop is used to resolve Windows compatibility issues (win95) about once a year.
The VGNA is now used purely as a Media centre.
I produce everything I do on the Dell and occasionally the EEPC’s. The unlisted computers just run iterations of numbers.
Therefore in twenty eight years, I personally have retired 34 computers. These are mainly personal machines and not business computers. (Seven replacements were due to coffee spills, lightning storms, electrical brownouts or physical shock (dropping)). The rest were retired either due to speed, or upgrade cost versus replacement cost analysis.
The average retention period by Koltai for a computer is therefore 3.423077 years.
Koltai’s idea of how a CEO should be connected to his work.
(Me, in my office at Norwich house (Level 22 overlooking the Sydney Harbour bridge) in 1995) with 6 PC’s visible and the laptop in my travel case on the floor under the desk. (There is also a vt100 terminal with Telnet access on the desk.)
Err… what about your home ….? OK – here’s home (1997)…. I was a bachelor at the time…. Can you tell? Please note the advanced use of BSD Post-it Notes on the Unix Terminal Screen (just couldn’t get that darn GNU X.11 desktop functioning…. OK – Tetris worked….)
and laptop on the coffee table.
If we use the same number on a global basis (not the number of PC’s, just the retention period as an average, then 1% of all PC users purchased a computer between 1983 and 1985 and likely exchanged/retired it at 3.42 years.
‘If we thereafter allow growth at the official rate of 40% (Gartner estimate from 1981-2000)
per annum, then the total number of PC’s manufactured worldwide is…… 7.8 billion.
So much for the official Gartner and IDG figures. But befiore we kill off the Gartner numbers, lets see if it fits with what we know about Australia.
Here is the Australian Version of the Gartner Formula with some Koltai Variations added in……
And….
Assumptions:
- 25% upgrade in 36 months
- 25% in 48 months
- 25% in 60 months
- 25% never upgrade
- 5% of all PC’s are Junked or placed in storage after 36,48 or 60 months
- 70% of all PC’s are hand me downs or sold after 36, 48 or 60 months.
- (The average age on eBay for secondhand PC’s has reduced from 52 months to 33 months in the last twelve months.)
- Mobile phones acquired PC abilities with Bluetooth, 802.11 and Spreadsheet capability (approx 2000)
- Accidental loss, burglary and damage account for 5.4% of pc replacement per year.
- Age of PC owners reduces in 2005 with Netbooks down to 12 from 15.
- 21% of Australians may never purchase a PC or not use it if there is one in the home.
The Average Australian will purchase a new PC or PC capable smart phone every three years giving an average mean of 19 devices per life expectancy cycle.
Yep, the Gartner formula works – a little bit on the high side, but – I like the numbers. (Incidentally, that’s 94 million PC’s that Australians have junked or retired since 1981.)
Of course, now everyone will redo their calculations and say – But we meant PC’s in actual operation.
Note of importance, number of Dos or Windows licences I transferred from one PC to another = zero.
Number of times I didn’t bother to buy Dos or windows with my computers (90%).
Now this is where it gets interesting.
The clones (the unofficial – second loading dock shipped clones) didn’t pay Microsoft a CPU licensing fee.
Out of the 34 computers, only ten had unix variant OS’es , one was CP/M and one was Pick leaving 22 computers of which I only purchased
Dos for about Ten of.
Koltai are you advocating piracy? Nope… please remember the Microsoft CPU licensing up until about version 6, I think from memory which was around 1997.
However, lets just examine what we have read here today.
About 5-6 billion PC’s made.
About 1.5-2 billion in use.
About 1.5 – 2 billion Microsoft licences purchased (200 million in 2008) (not all manufacturers signed up for the CPU licensing option).
An anecdote on the subject of why people file share….
Comment
by Sorefeet August 10, 2008
You guys have it all wrong. Price? It’s a trophy thing. If I
create a hack for a game and make it run better/longer/for less or no money, it
allows to impress the circle. If you trade said game with your peers, you’re
considered an interesting source. Social networking and all that. All kids
collect. Used to be stamps, used to be games, used to be p2p l/ps, bet it’s
zombies soon. It’s a power trip for those “in”.
If you want a more pedagogical answer, the warez and demoscene
worlds were once interlinked, and well known breeding places of todays “game
design talent”. You don’t learn this stuff at any school, not even at the
recent hot ‘games academy’ fad.
So, what I’m saying is.. if you want to build a kingdom, you
have to dispense a bit of power to your entourage..
Microsoft has 88.14% of Operating System market share
| 1. Windows XP | 55.34% |
| 2. Windows 2000 | 17.23% |
| 3. Windows 98 | 8.23% |
| 4. Macintosh | 8.18% |
| 5. Windows ME | 3.67% |
| 6. Linux | 3.67% |
| 7. Windows NT | 3.67% |
| 8. Macintosh Power PC | 3.67% |
So what I want to know is quite simple.
If file sharing is supposedly damaging to a companies sales
volumes.
And it would appear that Microsoft MS-DOS and Windows are
the most pirated software in the world.
Then…. if three billion plus copies of DOS and Windows were
pirated. Why is Microsoft the number one Software company in the world. How is that even possile with file sharing apparentely representing according to the numbers we have analysed, guestimated, AP’ed and plucked about 270% of total Microsoft OS sales.
As an item of interest, we estimate that 720 million license have been retired in the western world.
| Company | Software Revenues
(M)US$ | Software revenue
growth (%yoy) | Total Revenues
(M)US$ | Software Revenue
share | File Sharing
Likelihood (Koltai) | Likely at 100% |
| | | | | | | |
| Microsoft | 49,453 | 10% | 61,900 | 80% | Likely | 49,453 |
| IBM | 22,089 | 11% | 103,630 | 21% | Unlikely | |
| Oracle | 17,560 | 17% | 22,102 | 79% | Unlikely | |
| HP | 11,604 | 8% | 16,111 | 72% | Unlikely | |
| Symantec | 5,692 | 8% | 6,152 | 93% | Likely | 5,692 |
| Activision Blizzard | 4,622 | 73% | 5,032 | 92% | Likely | 4,622 |
Ignore the chart (it’s from something we’ll be posting in the future, however it’s an extract from the list of the top 100 software
company’s in the world showing revenues from software.)
Resources:
Old, used, and out-of-print Microsoft® MS-DOS® for Sale
References:
Microsoft sales fall for first time in 23 years
Software giant revenue drops 6% and earnings plummet 32% as PC sales continue to weaken. http://money.cnn.com/2009/04/23/technology/microsoft_earnings/index.htm
Cliffski’s Blog » Genuine call for emails from pirates
Dedrick, Jason, Kraemer, Kenneth L., & Palacios, Juan J.(1999). Impacts of Economic Integration on the Computer
Sector in Mexico and the United States. UC Irvine: Center for Research on Information Technology and Organizations. Retrieved from:
http://escholarship.org/uc/item/1tx1b7×7 Microsoft defends DOS licensing
Denis Fred Simon, Merle Goldman (1989) “Science and technology in post-Mao China
Volume 5 of Harvard contemporary China Harvard Univ Asia Center,
1989ISBN0674794753, 9780674794757
