Chapter 180: The Machine That Anyone Could Use
22 January 1975 — Bombay; San Francisco
The machine had been sitting on the desk for six hours when Aditya finally looked at it properly.
Not at the specifications. Not at the production numbers. At the machine itself.
It was eleven-thirty at night on January 21st, and the Shergill Electronics Pedder Road office was full of people who had been awake since six in the morning and who showed no particular sign of becoming less awake, which was the quality of teams in the last eighteen hours before something they had built for two years was about to be shown to the world. The packaging line in the Gorakhpur factory had completed its run at nine this evening — twelve thousand units, each in a box that was white with a single blue line and the word SIDDHARTH printed in the specific typeface that Rajiv Menon had spent three weeks arguing about and had ultimately won, which was Rajiv's characteristic relationship to things he cared about. The US inventory — four thousand units, shipped through the Bombay High commercial shipping network — had cleared San Francisco customs on January 20th and was in the Shergill Electronics US warehouse in Redwood City.
The demo unit on the desk was the twenty-third Siddharth-1 to come off the production line, which meant it was the one that had been through the most testing, the one that the software team had run their full suite against, the one that Sunita Rao had spent two days confirming contained the final ROM with the final BASIC interpreter and not any of the seventeen previous versions that contained the seventeen categories of bugs that the eighteen months of development had progressively eliminated.
Aditya looked at it.
It was a rectangular box, approximately the dimensions of a large hardback book — not the book itself, the dimensions of the box the book came in. The chassis was injection-moulded ABS plastic in a warm beige-grey that the industrial designer had chosen because it was the colour of the machines that would be in offices rather than in laboratories, the colour of something that expected to be used by people rather than by engineers. The front face had a speaker grille in the lower left, a power switch and a cassette port on the right edge, and the ISMC logo in the lower-right corner with the text: ISMC Proprietary Processor. Made in India.
Above the chassis, connected by a short cable, was the keyboard.
The keyboard was sixty-one keys — QWERTY standard layout, with a row of function keys across the top, the keys white with black legends, the travel distance calibrated to the specific twelve millimetres that Rajiv had insisted on after spending three weeks testing every keyboard mechanism available in Asia.
Beside the whole assembly, connected by the RF output cable, was a standard black-and-white television — the fourteen-inch Shergill Jawan, specifically — which was displaying the boot screen.
The boot screen said:
SHERGILL SIDDHARTH-1ISMC BASIC v1.032KB READY
>
The cursor was blinking.
Aditya looked at the blinking cursor.
He reached over and typed:
PRINT "HELLO"
He pressed RETURN.
The screen immediately displayed:
HELLO
>
He looked at it.
He thought about the Popular Electronics magazine that had arrived on his desk four days ago — the January 1975 issue, with the Altair 8800 on the cover. He had read the article carefully. He had read it as an engineer who understood what the Altair was and what it wasn't, and as a financial director who understood what the Altair cost and what the person who bought it received for the money.
The Altair 8800 cost $439 as a kit that you assembled yourself, which required soldering skills and an oscilloscope to test and several weeks of building and debugging. When you had built it, you had a blue metal box with a row of toggle switches and a row of blinking LEDs. There was no keyboard. There was no screen. There was no programming language. To put a programme into the machine, you flipped individual toggle switches to represent binary ones and zeros, one bit at a time, until you had entered the programme. Then you ran it and watched the blinking LEDs to understand the output.
The article in Popular Electronics had described this as a revolution.
Aditya had agreed with the word. He had disagreed with what the word was being applied to. The revolution was not the Altair 8800. The revolution was the idea that a personal computer was possible. The Altair was the proof of concept. The Siddharth-1 was the revolution.
He typed:
10 FOR I = 1 TO 1020 PRINT I, II* 30 NEXT I40 ENDRUN
The screen displayed:
1 12 43 94 165 256 367 498 649 8110 100
>
He looked at the screen.
The machine had calculated a multiplication table in response to four lines of BASIC typed on a keyboard, displayed the results on a television screen, and waited for the next instruction. The entire sequence, from the first keystroke to the final > prompt, had taken approximately two seconds.
He stood up.
He went to find Rajiv Menon.
Rajiv Menon was in the production manager's office adjacent to the demonstration area, on the phone.
He was thirty-two years old, from Bangalore, and had the specific quality of a product manager who had spent two years building something and was twelve hours from showing it to the world: the quality of someone who was holding the full complexity of everything that could still go wrong in one part of his attention and the full certainty that most of it was fine in the other part, and who was managing the relationship between those two parts with the exhausted precision of a man running on his second or third wind.
He had been the person who had written the original product specification for the Siddharth-1 eighteen months ago — the specification that had said: we are not building a hobbyist kit. We are building a machine that any person, with no technical training, can use on the day they bring it home. He had written this as a design principle and had then spent eighteen months defending it against every decision that would have compromised it, and the defence had been complete enough that the machine on the desk in the other room was exactly what the specification said it would be.
He looked up when Aditya came in.
"Four thousand units in Redwood City," Aditya said. "Customs cleared."
"Vikram called me an hour ago," Rajiv said. He ended his phone call. "The display units are set up at the venue. Five Siddharth machines on five separate tables. Each one connected to a monitor. Vikram's running a demo loop on three of them and leaving two open for the press to use themselves."
"The press in San Francisco," Aditya said.
"Eleven journalists confirmed," Rajiv said. "Three from trade publications — Byte, Creative Computing, Electronics. Two from mainstream newspapers — San Francisco Chronicle, New York Times technology desk. Four from general interest magazines. Two from radio."
"Radio," Aditya said.
"The Siddharth makes sound," Rajiv said. "The BASIC interpreter includes a BEEP command and a TONE command with frequency and duration parameters. Vikram suggested demonstrating the sound capability for radio journalists who couldn't transmit the visual." He paused. "Also, radio reaches people who don't read technology publications."
"Good thinking," Aditya said.
"It was Sunita's idea," Rajiv said.
"Tell Sunita," Aditya said.
He looked around the office. Through the glass, he could see the demonstration area — the space that the Pedder Road office had been reconfigured into for tomorrow's Indian launch event. Five Siddharth-1 machines on five tables. Connected to five Shergill Jawan televisions. The chairs arranged in a semicircle. The press area. The direct sales desk at the back, where orders would be taken.
"The price," Aditya said.
"₹4,800 in India," Rajiv said. "Includes the keyboard, the RF cable, the cassette interface, the power supply, and the BASIC Quick Reference Guide." He paused. "$395 in the United States. Same package."
"The Altair comparison," Aditya said.
"At $439, the Altair 8800 kit gives you a machine with no keyboard, no screen, no programming language, 256 bytes of RAM, and a construction project that requires an oscilloscope and soldering expertise." He paused. "At $395, the Siddharth-1 gives you a machine you plug into your television, turn on, and start typing in BASIC immediately."
"We're cheaper and we're ready to use," Aditya said.
"We're cheaper and we're a completely different category," Rajiv said. "The Altair is a hobbyist kit. The Siddharth is a computer."
Aditya looked at him.
"That is the sentence for tomorrow," Aditya said.
"I know," Rajiv said. "I've been trying to find a way to say it that doesn't sound like we're attacking the Altair."
"Say it exactly like that," Aditya said. "We're not attacking the Altair. We're defining what we are. The Altair is what it is. We're something else."
Rajiv wrote it down.
"Get some sleep," Aditya said.
"After this call," Rajiv said.
"Who's the call?" Aditya said.
"Sunita," Rajiv said. "She found a bug."
Aditya looked at him.
"In the ROM?" he said.
"In the cassette save routine," Rajiv said. "It's not a crash bug. It's a performance issue — the save is slower than the load, which is inconsistent. It's not on the demo units. It's in the revision that's in the shipping inventory."
"The twelve thousand units in Gorakhpur," Aditya said.
"Yes," Rajiv said.
"How slow?" Aditya said.
"The save is 32 bytes per second slower than the load," Rajiv said. "At maximum cassette storage — about 16KB — the save takes approximately four minutes and the load takes approximately two minutes."
Aditya was quiet.
"A user who saves a large programme and then loads it will notice that the load is twice as fast as the save," Rajiv said. "Which is confusing and technically inelegant. It's not wrong — the data is saved and loaded correctly — but it's inconsistent."
"Is this fixable in the ROM?" Aditya said.
"Not without pulling twelve thousand units back from inventory and re-burning the ROMs," Rajiv said. "Which we can't do overnight."
"Then it ships as is," Aditya said.
"The cassette performance difference," Rajiv said. He was looking for the right frame.
"Is a known issue that will be corrected in a future ROM revision," Aditya said. "Which we will make available to all Siddharth-1 owners through a ROM replacement programme at no charge." He paused. "That's the answer. Not hiding the issue. Acknowledging it and committing to the fix."
Rajiv looked at him.
"If we announce a known bug at launch," Rajiv said.
"We announce a minor known performance inconsistency," Aditya said. "In the cassette timing. Which doesn't affect functionality. And we announce the fix programme at the same time we announce the product." He paused. "A company that acknowledges a minor issue and commits to fixing it is more trustworthy than a company that hides it and leaves the customer to find it."
"This is unprecedented," Rajiv said.
"So is the product," Aditya said.
He walked out.
He went back to the machine on the desk.
He typed:
PRINT 1975
The screen said: 1975
He looked at it.
The year was four days old and the thing on the table was about to change what it contained.
At six in the morning on January 22nd, Sunita Rao was still in the building.
She was twenty-eight years old, from Madras, and she had written every line of the ISMC BASIC v1.0 interpreter that was burned into the Siddharth-1's ROM chips. She had written it in fourteen months, which was the specific timeline that Rajiv's product schedule required, and she had written it in the way that good software was written: not quickly, but efficiently, which was different — code that was precise and minimal and fast because precision and minimalism and speed were the same virtue in three aspects.
The BASIC interpreter was 8,192 bytes long.
This was a significant achievement. The Altair BASIC, written by a young programmer named Bill Gates and his colleague Paul Allen who had read the same Popular Electronics article and had seen the business opportunity, was 4,096 bytes. Sunita's BASIC was 8,192 bytes and did considerably more — it included floating-point arithmetic, string handling functions, file input/output via the cassette interface, a GRAPHICS mode that allowed basic block drawing, the TONE and BEEP commands, and a built-in screen editor that allowed you to move the cursor around the screen and modify lines of code rather than having to retype them entirely.
The 4,096 extra bytes had been a constant negotiation with Dr. Arvind Bose, the hardware architect, who had wanted to fit the ROM into a single 8KB chip and had eventually agreed to a 16KB ROM arrangement that gave Sunita the space she needed. The negotiation had lasted three weeks and had produced, in its resolution, the specific quality of a technical compromise that served both parties: Bose got a clean hardware design, Sunita got a complete interpreter.
She was sitting at one of the demonstration machines at six in the morning running the cassette timing test.
Not the bug test — she had confirmed the bug the previous evening and had accepted Aditya's decision on the disclosure approach with the specific acceptance of a programmer who recognised that the decision was correct. She was running the timing test because she had been running the timing test every morning for six months, which was the specific habit of someone who understood that a software system that was working correctly today had to be tested again tomorrow because correctly-working systems had a way of finding new failure modes under conditions that hadn't been tested.
The machine was working.
She saved a 16KB file to cassette.
She loaded it back.
The save took four minutes and seventeen seconds.
The load took two minutes and three seconds.
She wrote these numbers in her test log.
She looked at the machine.
She thought about what it was.
She had been building it for fourteen months and she had, in those fourteen months, developed the specific relationship to the Siddharth-1 that engineers developed with things they built over long periods: the knowledge of it that went beyond specification and into the specific character of the system, the way it behaved under different conditions, the specific quirks that were not bugs but were the personality of the implementation.
The BASIC interpreter had a personality. She had given it one, not deliberately but inevitably — because the person who wrote the code was expressed in the code, and Sunita's code was precise and efficient and structured with the specific clarity of someone who had come to programming from mathematics and who therefore thought of code as applied logic rather than as craft.
When you wrote 10 FOR I = 1 TO 100 on the Siddharth-1 and pressed RETURN, the machine didn't pause. It registered the line instantly, formatted it correctly, and returned the cursor ready for the next input. That instantness was deliberate — every response cycle in the interpreter had been tightened to the minimum because Sunita believed that a machine's responsiveness communicated to the user whether the machine respected them. A slow response said: your input is an inconvenience. A fast response said: I was waiting for you.
The Siddharth-1 waited for you.
She typed one more programme. Not a test programme — a programme she had written for herself, months ago, when she was debugging the floating-point routines and needed a test case.
10 LET PI = 3.1415926520 PRINT "CIRCUMFERENCE OF CIRCLES:"30 FOR R = 1 TO 1040 LET C = 2 * PI * R50 PRINT "RADIUS"; R; "= "; C60 NEXT R70 ENDRUN
The machine displayed:
CIRCUMFERENCE OF CIRCLES:RADIUS 1 = 6.28318RADIUS 2 = 12.5664RADIUS 3 = 18.8496RADIUS 4 = 25.1327RADIUS 5 = 31.4159RADIUS 6 = 37.6991RADIUS 7 = 43.9823RADIUS 8 = 50.2655RADIUS 9 = 56.5487RADIUS 10 = 62.8318
>
She looked at the screen.
She thought about every mathematics student in India who could use this machine to verify their calculations. Every small business owner who could use it to track inventory. Every schoolteacher who could use it to demonstrate the relationship between equations and their outputs.
She thought: today this starts.
She closed her test log.
She went home to change before the launch.
Dr. Arvind Bose arrived at the Pedder Road office at seven in the morning.
He was forty-five years old and had been designing electronic systems for twenty years. He had worked for Bharat Electronics, for the Atomic Energy Commission's instrumentation division, and for the past two years for Shergill Electronics on the Siddharth-1 hardware. He was the person who had translated the product specification into silicon — who had decided how many chips, of which types, in which arrangement, connected by which bus, powered by which supply.
He was also the person who had spent the past two years dealing with the specific challenge that any hardware designer faced when building a system based on semiconductor technology that was more advanced than what the software team expected: managing their surprise.
When Sunita had first received the ISMC chip samples — the 3-micron custom processor at 4.5 MHz — she had run her benchmark suite and had come to Bose's office and said: the machine is faster than I planned for.
He had said: yes.
She had said: I need to rewrite the timing loops in the interpreter.
He had said: yes.
She had said: do you know how long that takes?
He had said: I know how long it takes. That's why I told you six months ago to build the timing loops as parameters rather than constants.
She had looked at him for a moment. Then she had gone back to her desk and spent three days confirming that he was right and that the parameterised timing loops he had specified in the system design document would adapt to the processor's actual speed without requiring a complete interpreter rewrite.
He had been correct.
He was usually correct about hardware. This was his value.
He arrived at seven and went directly to the five demonstration machines in the demonstration area. He had tested each of them personally the previous afternoon. He tested them again now, methodically, the specific hardware test sequence he had developed: power on, check the boot screen, check the keyboard scan (every key, in order), check the video output signal with a small instrument he kept in his briefcase, check the cassette interface with a standard tape he carried, check the expansion connector.
Machine one: pass.
Machine two: pass.
Machine three: the power indicator LED was slightly dimmer than nominal. He opened the chassis, found the resistor value was at the edge of its tolerance, replaced it from the small component kit he also kept in his briefcase, closed the chassis, retested. Pass.
Machine four: pass.
Machine five: pass.
He moved the dim-LED machine to the back position of the five — not the first machine a journalist would approach, not the most prominent. Not because the fix was uncertain — it was fine — but because a perfectly working machine should be at the front.
He sat down.
He looked at the five machines.
Each one was connected to a Shergill Jawan television. Each one was displaying the boot screen.
SHERGILL SIDDHARTH-1ISMC BASIC v1.032KB READY
>
He had spent two years building the system that produced those five lines of text.
The twenty people who came into this room in the next hour would look at those five lines and not understand what they meant. They would look at a television screen with text on it and a keyboard beside it and they would understand that these things were connected, and they would understand that the text was inviting them to do something, and they would not know quite what to do.
Then Rajiv's demonstration team would show them.
And then they would understand.
He sat with the five machines in the quiet of the seven o'clock morning, which was already warm in Bombay in January, and he thought about a different machine.
He thought about the first computer he had ever touched. TIFR's IBM 7090, in 1964. A room full of equipment that required specialist operators and a submission queue and punch cards and a wait of several hours between submitting your programme and receiving the output. He had been a graduate student and the access had been rationed and he had spent weeks designing his programmes to be as efficient as possible so they used the minimum amount of machine time.
He was forty-five years old, and the thing on the table in front of him was something that an eleven-year-old could turn on and use.
He thought about his daughter.
She was eleven years old. She was in school in Bombay and was good at mathematics. He had brought home a prototype Siddharth-1 three months ago for testing and she had sat in front of it for ten minutes and had then started typing BASIC programmes from the Quick Reference Guide he had also brought home.
She had spent four hours that evening writing a programme that generated multiplication tables and displayed them formatted in columns. She had found the formatting output functions by trial and error from the commands list. She had debugged the spacing herself. She had shown him the result with the satisfaction of someone who has made something work.
He had looked at the columns of numbers on the television screen.
He had thought: this is what the machine is for.
Not for the people who understood how it worked. For the people who could use it without knowing how it worked. For his eleven-year-old daughter who had never thought about transistors and didn't need to.
He looked at the five machines on the tables.
He had built the thing that his daughter could use.
Aditya arrived at eight.
He had gone home at two, slept for four hours, showered, and returned. He was twenty-one years old and he had the quality of his brother in one specific respect: the ability to operate at full capacity on a reduced sleep budget because the importance of the day was doing the work that the sleep would ordinarily do.
He walked the demonstration area. He talked to Rajiv about the flow. He talked to Dr. Bose about machine three. He received the San Francisco weather report from Vikram Khanna by phone — overcast, 58 degrees, the specific San Francisco January that was cold but not inhospitable.
He received the press list for the Bombay event. Seventeen journalists confirmed. Six from technology and engineering publications. Four from business publications. Three from general newspapers including Subroto Bhattacharya, who had been the person to break the LED story and who had been given first call on technology announcements from Shergill on the implicit understanding that he understood what he was being shown.
Four journalists from radio.
He reviewed the demonstration script that Rajiv had prepared. He suggested two changes. Rajiv accepted one and argued about the other for twelve minutes before accepting the second.
At nine, the first journalists arrived.
The first journalist through the door was Subroto Bhattacharya.
He was thirty years old now — four years since the first LED story, which had made his career in the sense that it had established his reputation for being the person who identified significant technology stories before they were obviously significant. He had a way of arriving at technology events that was different from most of the journalists he competed with: he arrived before he knew what he was looking for and he paid attention until he found it.
He came through the door and saw the five machines on the five tables.
He stopped.
He looked at them.
He walked toward the nearest one. He looked at the boot screen. He looked at the keyboard. He looked at the television — the Shergill Jawan, which he recognised, which he had covered at the Gorakhpur mall launch the previous year.
He sat down.
He looked at the cursor.
He typed: HELLO
The machine displayed: ?SYNTAX ERROR IN IMMEDIATE MODE
He looked at the error message.
The demonstration assistant, a young engineer named Kavya, came over. She said: "You need to put it in a PRINT statement. Try: PRINT 'HELLO'"
Subroto typed: PRINT "HELLO"
The machine said: HELLO
He looked at the screen.
He typed: PRINT 2+2
The machine said: 4
He typed: PRINT 100100*
The machine said: 10000
He looked at the screen for a moment.
Then he turned to Kavya.
"This machine," he said, "just answered a question I typed, in English, in less than one second."
"Yes," she said.
"On a television screen," he said.
"Yes," she said.
"Using a keyboard," he said.
"Yes," she said.
He looked at the machine.
"How much?" he said.
"₹4,800 complete," she said. "Keyboard, cable, BASIC guide. You provide the television."
He was quiet for a moment.
He said: "The Popular Electronics magazine this month. The American computer — the Altair."
"Yes," she said.
"That costs more than this," he said. "And it has—"
"Toggle switches and blinking LEDs," she said.
"And no keyboard," he said.
"And no screen," she said.
"And no programming language," he said.
"And no ability to actually use it without building it yourself first," she said.
Subroto Bhattacharya looked at the Siddharth-1.
He opened his notebook.
He wrote: The machine anyone can use. India is first.
He underlined India is first.
The other journalists arrived in clusters between nine and nine-thirty.
The demonstration area filled with the specific noise of journalists at a technology event — the questions that were informed by partial knowledge, the note-taking, the specific photography of people documenting something they haven't fully understood yet for audiences that haven't understood it at all.
Rajiv stood at the front and ran the first demonstration.
He did it the way he had rehearsed it twelve times in the previous three days — not with performance, with clarity. He was not selling. He was showing.
"The Shergill Siddharth-1 is India's first personal computer," he said. "Not a kit you build. Not a machine you need an engineering degree to operate. A machine you bring home, plug into your television, and use."
He turned on the machine.
The boot screen appeared.
He said: "When you turn it on, it's ready in less than one second. There's no warm-up. There's no loading procedure. The machine is ready."
He typed:
PRINT "GOOD MORNING"
The machine said: GOOD MORNING
Someone in the room made a sound. Not amazed — the sound of someone who has just had something confirmed that they suspected but weren't certain of.
He said: "That instruction — PRINT — is part of the BASIC programming language. BASIC stands for Beginner's All-purpose Symbolic Instruction Code. It's a language designed to be readable by humans. If I want the machine to calculate something, I say PRINT, and then I write the calculation."
He typed:
PRINT 365 * 24 * 60 * 60
The machine said: 31536000
"That's the number of seconds in a year," he said. "Calculated instantly." He paused. "I didn't need to know how the machine does this. I just needed to know how to ask."
He typed a longer programme — the compound interest calculator he had prepared for the business journalists:
10 INPUT "PRINCIPAL"; P20 INPUT "ANNUAL RATE %"; R30 INPUT "YEARS"; N40 LET A = P * (1 + R/100) ^ N50 PRINT "FINAL AMOUNT: "; A60 ENDRUN
The machine asked: PRINCIPAL?
He typed: 10000
The machine asked: ANNUAL RATE %?
He typed: 8
The machine asked: YEARS?
He typed: 10
The machine said: FINAL AMOUNT: 21589.2
He looked at the journalists.
"₹10,000 at 8% annual interest for 10 years," he said. "The answer in under one second. And I can run it again with different numbers instantly." He typed RUN and entered different values. "Any bank manager. Any shopkeeper. Any accountant. Anyone who needs to calculate compound interest regularly — this machine does it in five lines of code that take two minutes to type once and then run forever."
A journalist from a business publication — Ramesh Iyer from the Economic Times — raised his hand.
"How do you save this programme?" he said.
Rajiv connected a cassette recorder to the machine's audio port — a standard Phillips cassette recorder.
"You press record on the cassette recorder," he said, doing this, "and you type SAVE 'INTEREST' in the BASIC prompt."
He typed SAVE "INTEREST"
The machine made a sound through the speaker — a distinctive tone as it wrote data to the tape. After thirty seconds it stopped and displayed: SAVED: INTEREST
"Now," Rajiv said, "I turn off the machine."
He turned it off.
He turned it back on.
The boot screen: 32KB READY
>
He pressed PLAY on the cassette recorder and typed LOAD "INTEREST"
After two minutes, the machine displayed: LOADED: INTEREST
>
He typed RUN.
The machine asked: PRINCIPAL?
The room made the sound again. This time louder.
"That is the programme I saved," Rajiv said, "loaded back from a cassette tape. Standard audio cassette. Available at any music shop for two rupees." He paused. "I want to note that the save is slower than the load — approximately four minutes to save a full programme versus two minutes to load it. This is a known timing inconsistency in the current ROM version. We will provide a free ROM replacement to all Siddharth-1 owners when the corrected version is available."
The journalists wrote this down.
The Economic Times journalist wrote: transparent about minor bug at launch — unusual.
He underlined unusual.
At ten o'clock in Bombay — which was one AM in San Francisco — Vikram Khanna was still awake.
He was thirty-five years old, from Hyderabad, and had been running Shergill Electronics' American operations for fourteen months. The American operation had been established primarily for the LED licensing transactions and the commercial sales of ISMC semiconductor products to GE and Philips, which required a US presence for contract management and delivery coordination. The Siddharth-1 US launch was the first time the American office had managed a consumer product launch, which was a different kind of work than component sales.
Vikram had been preparing for the San Francisco launch for six weeks.
The location: a technology demonstration space in the South of Market district of San Francisco — not the city's traditional business district, but the area where the technology hobbyist community was beginning to concentrate around small workshops and the Homebrew Computer Club meetings that had been happening in Menlo Park for the past year. He had chosen the location specifically because the people who would be most interested in the Siddharth-1 would be people who were already thinking about personal computers, and those people were concentrated in this specific geography.
The invited press was different from Bombay's press. The trade publications — Byte, Creative Computing, Popular Electronics — were the primary audience. These were publications read by people who already understood what a computer was and who would immediately understand what the Siddharth-1 was relative to what existed. The mainstream press — the Chronicle, the Times — were secondary, the audience who would need the story translated into general-interest terms.
He had also — on his own initiative, which he had confirmed with Aditya three weeks ago — arranged for five members of the Homebrew Computer Club to be at the event.
The Homebrew Computer Club was a gathering of technology enthusiasts in the Menlo Park area who met to discuss and demonstrate computing equipment. They had received their copy of the January Popular Electronics with the Altair cover story before Christmas and had been discussing it intensely for weeks. Several members were already building Altairs. Several others were waiting for the kit to ship.
Vikram had sent a letter to the Club's mailing list in early January. The letter said: We are introducing a new personal computer at a demonstration event in San Francisco on January 22nd. The Homebrew Computer Club is invited. Your technical expertise would be welcome.
Five members had confirmed.
The five members had not been told what the computer was or what it could do.
They arrived at the venue at nine AM — one hour before the mainstream journalists — and Vikram had set one machine aside specifically for them.
The Homebrew members were between twenty-two and thirty-five years old. They all knew what the Altair was. They all knew that "personal computer" in January 1975 meant a kit you built and programmed with toggle switches.
The first one to sit in front of the Siddharth-1 was a young man named Steve — twenty years old, wearing jeans and a workshirt, with the specific quality of an extremely technically capable person who had learned to assess technology by touching it before reading about it.
He looked at the machine.
He looked at the keyboard.
He looked at the television screen showing:
SHERGILL SIDDHARTH-1ISMC BASIC v1.032KB READY
>
He typed: HELLO
The machine said: ?SYNTAX ERROR IN IMMEDIATE MODE
He smiled. A smile of recognition — the recognition of someone who knew exactly what the error meant and why it happened and what it said about the machine's design.
He typed: PRINT "HELLO"
The machine said: HELLO
He typed: FOR I = 1 TO 10 : PRINT I : NEXT I
The machine printed the numbers 1 through 10 instantly.
He looked at the machine for a long time.
Then he said, to no one in particular: "How much memory?"
"32KB standard," Vikram said, from behind him. "On-board DRAM. You can add external RAM through the expansion connector up to 64KB."
"The processor," Steve said.
"Custom 8-bit/16-bit hybrid. 4.5 MHz clock. ISMC proprietary architecture."
"ISMC," Steve said. "The Indian semiconductor company."
"Yes," Vikram said. "The same company that makes the LED."
"The 3-micron process," Steve said.
"Yes," Vikram said.
Steve was quiet.
He typed: PRINT PEEK(1234)
The machine printed a memory contents value.
He typed: POKE 1234, 65
He typed: PRINT PEEK(1234)
The machine printed: 65
He had direct memory access. He had read a memory location and written a value and read it back. This was the test that a programmer who wanted to understand a machine's architecture performed.
He turned to Vikram.
"How long has this been in development?" he said.
"Eighteen months," Vikram said.
"What's the price?"
"$395 complete. Keyboard, cable, BASIC guide."
"The Altair is $439 and it has no keyboard and no screen," he said. He was not asking a question.
"Yes," Vikram said.
Steve turned back to the machine.
He spent the next forty minutes typing and testing. He tested the floating-point arithmetic. He tested the string functions. He tested the TONE command and the BEEP command. He found the screen editor — the ability to move the cursor up to a previous line and modify it — and tested its limits. He tested the cassette save and load.
When the mainstream press arrived at ten, Steve was still at the machine.
One of the journalists — from Popular Electronics, which had just published the Altair cover story — sat beside him and watched.
"What's the verdict?" the journalist said.
Steve thought about this.
"The Altair," he said, "is a computer. It has the potential to do computing. To do actual computing with it — to solve problems, to write programmes — you need to add a keyboard, add a monitor, add a programming language, and spend months learning the hardware." He paused. "This machine—" He looked at the Siddharth-1 "—is a computer that already works. You don't add anything. You just use it."
The journalist wrote this down.
"What's the company?" he said.
"Shergill Electronics," Steve said. "Indian company. They make the LEDs in the Popular Electronics LED story from last year."
"The same company?" the journalist said.
"Same semiconductor facility," Steve said. "The chips in this machine are from the ISMC facility in India. 3-micron process."
The journalist looked at the machine.
He had written the Altair cover story. He had spent two months researching the Altair, understanding its architecture, explaining to readers why a hobbyist kit with no keyboard and no screen was a revolution.
He was now sitting next to a machine that had a keyboard and a screen and a programming language and 32KB of memory and cost less than the thing he had described as revolutionary.
He typed: PRINT "HELLO"
The machine said: HELLO
He looked at it.
He thought about how he was going to write this story.
By noon in Bombay, the demonstration had been running for three hours.
The journalists had each spent time at the machines — not the polite cursory engagement of people who were there to collect quotes and file, but the absorbed engagement of people who were trying to understand something that exceeded their prior category for what computers were. The BASIC interpreter was the specific thing that produced this engagement: the ability to type a question in something that was almost English and receive an answer was not what people expected computers to do.
Subroto was on his fourth programme.
He had started with the simple ones from the Quick Reference Guide. He had then started experimenting — modifying the programmes, adding conditions, trying things that weren't in the guide to see what happened. When something didn't work, he read the error message, figured out what it meant, and tried again. This was the process of learning to programme, and he was doing it naturally, without being taught, because the machine's error messages were in readable English and the corrections they implied were intelligible to anyone who could read.
He was writing a programme to calculate the number of days between two dates when Aditya came to stand beside him.
"The date calculation," Aditya said.
"I want to know how many days until my mother's birthday," Subroto said, without looking up. He was typing: LET D2 = 31 + 28 + 15 / (He stopped.) "Leap year. Do I need to account for leap year?"
"February 29th is 29 days," Aditya said. "Divide the year by four. If the remainder is zero, it's a leap year."
Subroto typed: IF Y / 4 = INT(Y/4) THEN LET FEB = 29 ELSE LET FEB = 28
He ran the programme.
He got a number.
"March 15th," he said. "53 days."
He looked at the screen.
"Aditya," he said.
"Yes," Aditya said.
"I learned to do this in the last twenty minutes," he said.
"Yes," Aditya said.
"I have no programming background," he said.
"I know," Aditya said.
"That means anyone can do this," he said.
"That is the point," Aditya said.
Subroto looked at his notebook. He looked at the machine. He looked at his notebook again.
"What are the specific uses?" he said. "The applications. A person buys this. What do they do with it?"
Aditya said: "What were you doing for the past twenty minutes?"
"Learning to write a programme that solved a problem I actually had," Subroto said.
"That's it," Aditya said.
"That's the answer?"
"The applications are whatever problems the person who buys it actually has," Aditya said. "A shopkeeper has an inventory problem. An accountant has a calculation problem. A student has a mathematics problem. A teacher has a curriculum problem. The machine doesn't solve those problems — it gives people the tool to solve their own problems." He paused. "In Bombay, in January 1975, the people who can afford ₹4,800 have specific problems that this machine makes easier to solve. That's the market."
"Who is the market?" Subroto said.
"Everyone who has a problem that involves calculation and information management and who currently manages it with paper and pen or mental arithmetic," Aditya said. "Which is everyone."
"At ₹4,800," Subroto said.
"At ₹4,800," Aditya said.
Subroto wrote: ₹4,800. The price of access to calculation.
He underlined access to calculation.
He thought about what that meant for a country where access to calculating machines had been restricted to institutions since independence — to universities and corporations and government offices, where the IBMs and the HCLs lived behind access barriers of cost and expertise and privilege.
He thought about a schoolteacher in Gorakhpur with a Siddharth-1 and a television set and what that teacher could do.
He wrote: This is not about computers. This is about who gets to compute.
In San Francisco, at eleven in the morning, Tom Wallace was having the specific experience of a journalist who has just been given the story of the year on the day that the story of the year was supposed to be someone else's story.
Tom Wallace was thirty-six years old and was the technology editor of Popular Electronics. He had commissioned and edited the Altair 8800 cover story. He had been involved in the decision to put the Altair on the January 1975 cover because he had believed — and still believed — that the Altair represented something genuinely significant: the first affordable computer that a person could own.
He had been right.
He had been right that the Altair was significant. He had not known, when he made that decision, that he was about to be shown something that made the Altair's significance look modest.
He was at a table with a Siddharth-1 in front of him.
He had been a technology journalist for twelve years. He had tested mainframes, minicomputers, programmable calculators, all categories of electronic equipment. He had the specific skill of a journalist who tested technology: the ability to find the relevant questions quickly, to identify what a thing was actually good for as distinct from what its manufacturer said it was good for, to understand the gap between specification and performance.
He typed: PRINT CHR$(65)
The machine printed: A
He typed: FOR I = 65 TO 90 : PRINT CHR$(I) : NEXT I
The machine printed: A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Character output. The machine could handle individual characters as numeric codes.
He typed: PRINT LEN("HELLO WORLD")
The machine printed: 11
String length. The machine had string functions.
He typed: PRINT MID$("HELLO WORLD", 7, 5)
The machine printed: WORLD
Substring extraction. The machine had full string manipulation.
He was writing a test programme — a programme to test the machine's speed on a computationally intensive task — when Vikram Khanna came to stand beside him.
"The BASIC is comprehensive," Tom said, without looking up. "String handling, floating point, character codes. More complete than Gates' Altair BASIC."
"Sunita added the string functions in month ten of the development," Vikram said. "The specification originally didn't include them. She added them because—" He paused. "She said that a language that can only handle numbers can only help people who only need to manipulate numbers. Most people's problems involve words."
Tom looked up.
"The programmer is a woman," he said.
"Yes," Vikram said. "Sunita Rao. She wrote the entire interpreter."
Tom wrote this down.
"The price," he said. "$395."
"$395 complete," Vikram said. "Keyboard, RF cable, BASIC Quick Reference Guide, power supply."
"The Altair is $439," Tom said.
"Yes," Vikram said.
"And requires assembly," Tom said.
"The Altair kit contains approximately 170 parts that the buyer assembles and solders," Vikram said. "The assembly typically takes several days and requires an oscilloscope to test. After assembly, the machine has no keyboard, no screen, and no programming language. The Siddharth-1 arrives assembled, connects to any television via the included RF cable, and runs BASIC from the moment it's powered on."
Tom was quiet.
"The memory," he said. "32KB."
"Standard," Vikram said. "Expandable to 64KB through the expansion connector."
"The Altair ships with 256 bytes," Tom said.
"Yes," Vikram said.
"256 bytes versus 32 kilobytes," Tom said. "That's 128 times more memory."
"The ISMC DRAM chips are produced at the 3-micron fabrication facility in Gorakhpur," Vikram said. "The same facility that produces the LED devices."
Tom looked at the machine.
"The processor," he said.
"Custom ISMC design," Vikram said. "4.5 MHz. 8-bit/16-bit hybrid architecture."
"The Intel 8080 in the Altair runs at 2 MHz," Tom said.
"The ISMC processor is faster in the same 8-bit operations and handles 16-bit arithmetic natively," Vikram said.
Tom was quiet.
He looked at the machine.
He was writing notes rapidly.
"How many units are you bringing to market?" he said.
"The initial production run for the US market is four thousand units," Vikram said. "Orders above that will ship from the January production run, which is underway now at twelve thousand units per month."
"Twelve thousand units per month," Tom said.
"Yes," Vikram said.
Tom looked at the machine for a long moment.
He had committed to the Altair cover story. The issue was already printed and in distribution. He could not un-publish it. But he could write the next issue's cover story, which he was now going to do, and the next issue's cover story was going to be about the Siddharth-1.
He typed one more thing:
10 PRINT "HELLO WORLD"20 GOTO 10RUN
The machine began printing HELLO WORLD in an infinite loop.
He held down BREAK to stop it.
The machine stopped instantly and returned the cursor.
BREAK IN 10
>
He looked at the cursor.
He had been using computers for twelve years. He had never had a computer of his own. He had never had a computer that he could sit in front of alone and use without a specialist and without an institutional framework.
He was looking at the first machine he could have owned.
"I want to buy one," he said.
"The sales desk is at the back of the room," Vikram said.
Tom Wallace walked to the sales desk.
He bought a Siddharth-1.
He paid $395 out of his own wallet.
He carried the box to his rental car.
He drove back to his hotel.
He set the machine up connected to the hotel room's television.
He spent the next five hours learning BASIC.
At midnight, he called the Popular Electronics office and told the editor-in-chief that they needed to clear the cover for the February issue.
By three in the afternoon in Bombay, orders were being taken at the desk at the back.
The orders came from the journalists who had stayed — not all of them, not half, but the specific seven who had been at the machines long enough to understand what they were looking at and who had made the calculation and decided that ₹4,800 was the price of access to something they needed.
Subroto Bhattacharya ordered one.
He ordered it for his employer — not for himself personally, for the Hindustan Times editorial office. He put it in writing on the order form: Hindustan Times, Bombay. One unit. For editorial department use.
He had already decided what he would use it for. He would write a programme to calculate the word count of his articles. He would write a programme to manage his source list. He would write a programme to convert currency figures from one year to another using historical exchange rates, which was a calculation he made repeatedly in economics reporting and which currently required either a calculator and manual conversion tables or a phone call to the paper's research department.
He had spent four hours with the machine and he already knew what he was going to do with it.
The Economic Times journalist ordered one for his personal use.
Two engineering journalists ordered one each.
One of the radio journalists — a man who had not understood the technology when he arrived and who had been explained it by Kavya in terms that connected it to the specific problems of running a small radio programme — ordered one with the intention of using it to manage his programme scheduling.
At four, Aditya called Gorakhpur and spoke to the production manager.
"The first production run of twelve thousand units," he said. "What's the status?"
"11,847 units complete," the production manager said. "Expected completion of the run by January 25th."
"We're going to need a second run," Aditya said.
"How many?" the production manager said.
"Start with twelve thousand," Aditya said. "And talk to the line manager about whether we can increase throughput."
"To how much?" the production manager said.
"To as much as is possible without compromising quality," Aditya said.
He put the phone down.
He looked at the demonstration area. At the five machines, which were still running, still being used by various people who had arrived throughout the day — not just journalists now but people who had heard about the launch and had come to the Pedder Road office because technology events in Bombay in 1975 were not common and word had spread through the specific informal networks of the city's engineering and business community.
A man in his fifties — a shopkeeper, from his dress — was sitting at one of the machines with Kavya beside him. He was learning the INPUT command. He was building a programme that calculated the profit margin on items he sold. He was checking the calculation against his mental arithmetic and finding that they agreed, which was producing in him the specific satisfaction of a man who has found a tool that does what he has always done, faster and without error.
Aditya watched him.
He thought about what Karan had written in his notebook in October, after the LED programme: the feeling of being thought about. This is what the product is.
He thought: this is also what the computer is.
Someone in this building had thought about the shopkeeper. Had thought about what the shopkeeper's problem was and had built something that solved it. The shopkeeper didn't need to know that. The shopkeeper just needed to know that the machine solved the problem.
He sat down at the fifth machine.
He typed:
10 PRINT "SHERGILL ELECTRONICS"20 PRINT "SIDDHARTH-1"30 PRINT "22 JANUARY 1975"40 PRINT "INDIA'S FIRST PERSONAL COMPUTER"50 ENDRUN
The screen displayed:
SHERGILL ELECTRONICSSIDDHARTH-122 JANUARY 1975INDIA'S FIRST PERSONAL COMPUTER
He looked at it.
He thought: Karan is going to see this tomorrow and he's going to say that the product sold itself and I didn't need to do anything.
He thought: he would be right.
He deleted the programme.
He typed:
PRINT 1+1
The machine said: 2
He smiled.
This was the thing. Not the sophisticated demonstrations. Not the business applications. Not the date calculations or the compound interest or the inventory management. The thing was that any person, with no knowledge, could sit in front of this machine and ask it a question and get an answer.
PRINT 1+1
2
This was the beginning.
At eight in the evening in Bombay — which was nine-thirty in the morning in San Francisco — the Bombay event was winding down.
The machines were still running. The orders desk was closing after forty-one orders taken. Not forty-one machines — forty-one orders for delivery, with some orders for multiple units. The total unit count from the day's orders: sixty-three.
Sixty-three machines from one event on the first day.
Aditya looked at the number.
He called Vikram.
"San Francisco?" he said.
"The event is running," Vikram said. He was at the venue, managing the afternoon press. "Orders taken so far: forty-eight units."
"Forty-eight," Aditya said.
"We had a group from the Homebrew Computer Club," Vikram said. "Five people came. Three of them ordered immediately. The other two said they needed to think about it and then called back an hour later and ordered." He paused. "One of the journalists from Popular Electronics bought one out of his personal wallet."
"Tom Wallace," Aditya said. "I know who he is. He wrote the Altair cover story."
"He set up the demo unit in his hotel room and has been there for the past three hours," Vikram said.
Aditya was quiet for a moment.
"The Popular Electronics cover story," he said.
"He told me before he left the venue," Vikram said, "that he was going to propose the Siddharth-1 for the February cover."
Aditya looked at the ceiling.
The February cover of Popular Electronics.
The same magazine that had just put the Altair 8800 on its January cover as the revolution in personal computing.
February: INDIA'S SIDDHARTH-1: THE COMPUTER ANYONE CAN USE
He thought about what that meant for the orders.
"Production," he said.
"I know," Vikram said.
"How fast can we move the inventory from Gorakhpur to San Francisco?" he said.
"The current shipping route is seventeen days by sea," Vikram said. "We have four thousand units in Redwood City. If the February cover generates orders—"
"It will generate orders," Aditya said.
"How many?" Vikram said.
"More than four thousand," Aditya said. "Get me a freight quote for air shipping from Gorakhpur to San Francisco."
"Air freight," Vikram said. He was calculating. "For a twelve-thousand-unit run — the weight and volume—"
"Get me the quote," Aditya said. "Tonight."
"Yes," Vikram said.
Aditya put the phone down.
He sat in the demonstration area of the Pedder Road office at eight in the evening. The machines were still on. Kavya and two other engineers were still at them, demonstrating for the people who were still coming through the door — not journalists now, the public, people who had heard through the afternoon that something was happening at the Shergill Electronics office and had come to see.
A university student was learning the FOR-NEXT loop.
A woman who ran a small tailoring business was programming fabric measurements.
Two young men who were clearly engineering students were testing the limits of the expansion connector.
The shopkeeper from the afternoon was back, with his wife. He was showing her the profit margin programme.
She looked at the screen.
She said something to him in Marathi that Aditya didn't fully catch.
He said something back.
She nodded.
She said something to Kavya.
Kavya translated: "She wants to know if it can store recipes."
Kavya typed:
10 PRINT "RECIPE: DAL TADKA"20 PRINT "SERVES: 4"30 PRINT "LENTILS: 250 GRAMS"40 PRINT "GHEE: 2 TABLESPOONS"50 PRINT "ONION: 1 MEDIUM"60 END
She ran it.
The screen displayed the recipe.
The shopkeeper's wife looked at the screen.
She looked at Kavya.
She said something.
Kavya translated: "She says: 'and it remembers?'"
"We save it to cassette tape," Kavya said. "Then it always remembers."
The woman looked at the machine.
She looked at her husband.
She said something firm.
He looked at Aditya.
He said: "She says we need one."
Aditya thought about Sunita Rao at six in the morning running her cassette test. About Dr. Bose testing every key on every keyboard. About Rajiv arguing for three weeks about the typeface on the box.
He thought about a woman in Bombay who wanted to store her recipes in a machine that would always remember them.
He said: "The orders desk is at the back."
In San Francisco, the afternoon session continued.
Vikram was managing the room with the quality of a man who had prepared for one kind of event and was experiencing another kind of event that was larger than he had prepared for. The journalists who had come for the morning event had mostly left, but the people who had heard about the morning event were arriving throughout the afternoon — engineers who worked in the technology companies beginning to cluster in the South Bay, young people from the university, people who had been following the Altair story and who had come to see what the competition was.
The competition was not quite the right word.
One young man — maybe twenty-three, in a workshirt and jeans, who had been at the machine for forty-five minutes — said to Vikram:
"I've been building Altairs. Me and three other guys in a garage in Cupertino."
"Building them from the kit?" Vikram said.
"Yeah. We've been doing it for two months. We were going to modify them — add a keyboard, add a monitor interface, add a programming language." He paused. "We were doing all of that. Separately." He looked at the Siddharth-1. "It already does all of that."
"Yes," Vikram said.
"For less money," the young man said.
"Yes," Vikram said.
The young man was quiet for a moment.
"What's the expansion bus?" he said.
Vikram gave him the specification document.
He read it for five minutes.
"Clean," he said. "The edge connector is clean. No cross-talk issues."
"The hardware team spent three months on the bus design," Vikram said.
"The Altair S-100 bus has interference problems when you add more than two boards," the young man said. "It's a rushed design. They didn't think through the electrical characteristics."
"We had eighteen months," Vikram said.
The young man looked at the machine.
He sat down.
He typed: PRINT PEEK(65535)
The machine printed a value.
He typed: POKE 65535, 0
He typed: PRINT PEEK(65535)
The machine printed: 0
He typed: NEW
The machine said: OK
>
He typed a programme — the beginning of something, the first lines of an idea he had been thinking about for the past ten minutes, the idea that the Siddharth-1's clean expansion bus and complete software environment provided the foundation for the thing he and his friends had been trying to build in the garage in Cupertino.
He typed for twenty minutes without stopping.
When he stopped, he looked at what he had built.
He ran it.
It worked.
He turned to Vikram.
"I want six units," he said. "For the garage."
"I can take your order now," Vikram said.
"And I want to talk to whoever designed this," he said. "The hardware architect. I have questions about the expansion connector."
"I can put you in contact with Dr. Bose," Vikram said. "He's in Gorakhpur."
"I'll write him a letter," the young man said.
He gave his order information.
He wrote his name on the order form: Steve Jobs, Cupertino, California.
He paid $2,370 in cash.
He carried the order receipt to the door.
He stopped.
He turned around.
"One more question," he said.
"Yes," Vikram said.
"The BASIC interpreter," he said. "Is the source code available?"
"That's Dr. Rao's decision," Vikram said. "Sunita Rao. She wrote it. I can put you in contact with her."
He wrote something on a piece of paper.
He gave it to Vikram.
"That's my address in Cupertino," he said. "Tell her I'd like to correspond."
Vikram looked at the paper.
He put it in his pocket.
He went back to managing the room.
At ten in the evening in San Francisco, the event was officially closed. Unofficially, four people were still at the machines — Steve from the morning, who had come back after dinner, and three others who had arrived at different points in the afternoon and had not left.
Vikram let them stay.
At eleven, he counted the day's orders.
US orders: 127 units.
He called Aditya.
"127," he said.
Aditya was quiet for a moment.
"Combined with Bombay?" he said.
"Bombay was 63," Vikram said.
"190 units on day one," Aditya said.
"The Popular Electronics contact," Vikram said. "Tom Wallace. He called me at nine tonight. He's proposing the Siddharth-1 for the February cover story. He wants to know if he can speak with Sunita Rao."
"Give him her contact through us," Aditya said. "Everything goes through the office, not directly."
"Yes," Vikram said.
"Vikram," Aditya said.
"Yes," Vikram said.
"The six units that were ordered for a garage in Cupertino," Aditya said. "Do you have the name?"
"Steve Jobs," Vikram said. "And he wants to correspond with Sunita about the BASIC source code."
Aditya was quiet for a moment.
"Sunita's decision," he said.
"Yes," Vikram said.
Aditya thought about this for a moment.
He thought about eighteen months of development. About Sunita Rao writing an interpreter alone. About Dr. Bose spending three months on the bus design. About Rajiv arguing about the typeface for three weeks.
He thought about 190 units on day one.
He thought about a garage in Cupertino with six Siddharth-1 machines.
He thought about a shopkeeper's wife who wanted to store her recipes.
He thought about Tom Wallace writing the February cover story.
He thought about what Karan was going to say when he told him.
Karan would say: good. What's next?
Aditya thought about what was next.
The second production run. The air freight quote. The distribution network in the US that didn't exist yet but would need to exist before the Popular Electronics cover came out. The retail relationships in India — the engineering supply shops in Bombay and Delhi and Madras and Calcutta that needed to carry the machine, because ₹4,800 was reachable for many more people than the ones who came to a demonstration event. The software library — the programmes that users would write and share, the cassette library that would develop as the user community grew.
The school system. The specific application that Sunita had mentioned in a conversation three months ago: if a school has one Siddharth-1 and a mathematics teacher who knows BASIC, every student in the school can learn what a computer is and what it can do.
There were 600,000 secondary schools in India.
Not all of them could afford ₹4,800. Not most of them. But the ones that could — the government secondary schools in the state capitals, the better-equipped district schools, the schools adjacent to the industrial towns where the factory workers' children were — there were fifty thousand of those. Maybe more.
He thought about the LED and Chandra's programme.
He thought about the SPEI methodology.
He thought about what it would mean to apply SPEI's approach to software — not the hardware, the software, the creation of an institutional library of useful BASIC programmes that any school or small business could use on the Siddharth-1 without needing to write their own.
He picked up a pen.
He opened his notebook.
He wrote: Siddharth Software Library. Institutional programmes. Schools, small business, government. Free to share. ISMC maintains the archive.
He looked at what he had written.
He wrote: This is what the machine is for. Not the technology. The access.
He closed the notebook.
He turned off his office light.
He walked through the demonstration area.
The five machines were still on.
The boot screens were showing.
SHERGILL SIDDHARTH-1ISMC BASIC v1.032KB READY
>
Five cursors, blinking in the quiet.
He turned them off one by one.
The screen of each machine went dark as the television's phosphor faded.
One after another.
Until the demonstration area was quiet and dark and the only sounds were the Bombay night outside the window and the specific quiet of a space that had held something significant all day and was now resting.
He walked out.
He went home.
He was twenty-one years old and he had just run the launch of India's first personal computer, simultaneously in Bombay and San Francisco, and the day had produced orders for 190 machines and a cover story for Popular Electronics and six machines for a garage in Cupertino and a woman who wanted to store recipes.
He thought: Karan would ask what was next.
He thought: the next is the software library and the school system and the retail distribution and the air freight and the second production run and the letter that Sunita Rao was going to receive from Steve Jobs in Cupertino and whatever came from that.
He thought: there is a lot of next.
He went home.
In the morning, Karan read the Hindustan Times.
Subroto Bhattacharya's story was on the front page, above the fold.
The headline: INDIA LAUNCHES FIRST PERSONAL COMPUTER — SIDDHARTH-1 MAKES COMPUTING ACCESSIBLE TO ALL
He read the story.
He read: The Siddharth-1, introduced yesterday by Shergill Electronics in simultaneous events in Bombay and San Francisco, is not a kit or a hobbyist project. It is a complete, ready-to-use computer that any person can operate from the moment it is connected to a television set. At ₹4,800, it is priced for the serious professional or the educational institution — not yet for every household, but within reach of those for whom calculation and information management are daily requirements.
He read: The American Altair 8800, featured on the January 1975 cover of Popular Electronics and widely described as a revolution in personal computing, is a kit that requires assembly, costs more than the Siddharth-1, and has no keyboard, no screen, and no programming language. The Siddharth-1 has all three. By the metric that matters — what a user can actually do with it on day one — there is no comparison.
He read: India has not followed the world into the age of personal computing. India has arrived first.
He set the newspaper down.
He looked at the factory outside his window.
He picked up his notebook.
He wrote: January 22, 1975. Siddharth-1 launched. Aditya and the team. I didn't touch it.
He looked at what he had written.
He added: They didn't need me.
He looked at that sentence.
He added: This is what the programme was always for.
He closed the notebook.
He went to the seven o'clock S-35 briefing.
There was work to do.
End of Chapter 180
Shergill Siddharth-1 — Product Record
Launch Date: 22 January 1975 (simultaneous, Bombay and San Francisco)
Specification:
Processor: ISMC Custom 8-bit/16-bit Hybrid, 4.5 MHz RAM: 32KB DRAM (expandable to 64KB via expansion connector) ROM: 16KB (contains ISMC BASIC v1.0 interpreter) Display: RF modulator and composite video, 40-column text, block graphics Input: 61-key QWERTY keyboard Storage: Cassette interface (audio in/out) Audio: Speaker, TONE/BEEP commands Expansion: Shielded edge-connector bus, clean electrical specification Boot time: <1 second
ISMC BASIC v1.0:
Written by: Sunita Rao (14 months) Size: 8,192 bytes Features: Floating-point arithmetic, string handling, file I/O (cassette), GRAPHICS mode, TONE/BEEP, line editor, full error messages
Pricing:
India: ₹4,800 (keyboard, RF cable, BASIC guide, power supply) USA: $395 (same package) Comparison: Altair 8800 = $439, no keyboard, no screen, no language, 256 bytes RAM
Day One Orders:
Bombay: 63 units San Francisco: 127 units Total day one: 190 units
Known Issue Disclosed at Launch: Cassette save timing inconsistency (32 bytes/second slower than load). Free ROM replacement programme committed.
Notable Day One Purchasers:
Subroto Bhattacharya, Hindustan Times (institutional) Tom Wallace, Popular Electronics (personal — proposing February cover story) Steve Jobs, Cupertino CA (6 units for garage workshop; requested contact with Sunita Rao re: BASIC source code)
Production: 12,000 units/month, Gorakhpur factory. Second production run authorised January 22.
Next phase (Aditya notebook, January 22 1975): Siddharth Software Library — institutional programmes for schools, small business, government. Free to share. ISMC maintains archive.