Computer Fundamentals

Dual-series line charts are the SBI PO examiner's favourite playground. Two lines, four years, a dozen possible questions — and the candidate who pre-computes the right derived row finishes the whole set in the time others spend reading the first question.

Definition: A dual-series line chart plots two related quantities (here Revenue and Expenditure, in Rs cr) on the same axes across a common time-line, so that any vertical gap between the two lines is a derived quantity — in this case Profit = Revenue − Expenditure.

Reading the Chart First, Computing Later

The single biggest mistake aspirants make on Data Interpretation sets is to dive into Question 1 before they have understood the graph. A trained eye spends the first 20–25 seconds doing three things: identify the units (Rs cr, lakh, percentage?), spot the two series and their colour/legend, and mentally compute the invisible third row — Profit per year — for every data point shown.

For the lesson's data:

• 2021: Revenue 120, Expenditure 90, Profit = 30
• 2022: Revenue 150, Expenditure 100, Profit = 50

Writing these Profit values above each year on the rough sheet converts the chart from a two-line graph into a three-row table. Now "average profit", "highest profit year", "ratio of profits" and "percentage change in profit" all become single-step lookups.

Two Different "Percentages" the Setter Loves to Mix

This is the trap of the entire DI section in SBI PO Prelims and Mains. The same numbers (Profit 30 and 50) can be asked about in two completely different ways.

Percentage increase in profit from 2021 to 2022 — this is a growth question. The base is the earlier value:

Question: By what percent did profit grow from 2021 to 2022?
Solution:
Step 1: Change = 50 − 30 = 20.
Step 2: Base = profit of the earlier year = 30.
Step 3: % change = (20 / 30) × 100 = 66.67%.
Conclusion: Profit grew by 66.67%, often written as 66⅔% in options.

Profit as a percentage of Revenue for 2022 — this is a profit-margin question. The base is the same year's revenue:

Question: What was the overall profit % over revenue for 2022?
Solution:
Step 1: Profit (2022) = 150 − 100 = 50.
Step 2: Revenue (2022) = 150.
Step 3: Margin = (50 / 150) × 100 = 33.33%.
Conclusion: The profit margin for 2022 is 33.33%, i.e. 1/3.

Notice how the same numbers gave 66.67% and 33.33%. Choosing the wrong base is the number-one source of negative marks here.

Why the "Profit row" Trick Works

Most candidates re-read the chart for every sub-question, do a fresh subtraction, and then a fresh percentage. Across a five-question set that is roughly 20 unnecessary subtractions. Pre-computing four Profit values (one per year) costs you 15 seconds upfront and saves a full minute later. SBI PO Prelims gives you only ~25 minutes for 35 Quant questions, and a DI set carries 5 of them; that minute is the difference between attempting one extra set or not.

Why it matters

Banking exams (SBI PO, IBPS PO, RBI Grade B) treat DI as a speed topic. The questions themselves are arithmetic-easy — the difficulty is data-handling under pressure. Mastering the dual-series chart unlocks Tabular DI, Mixed-Graph DI and Caselet DI, because all of them rest on the same habit: derive the hidden row first, answer questions second.

Real-world example

Open any quarterly investor presentation by an Indian bank — SBI, HDFC, ICICI. The "Operating Profit" slide invariably shows Income (interest + other) and Expenditure (interest paid + operating expenses) as two lines or two bars across the last 4–8 quarters. Analysts immediately compute the gap, the YoY growth and the cost-to-income ratio. That is exactly what your SBI PO chart is rehearsing.

Common misconception

"Percentage increase" and "percentage of" sound similar but use different denominators. Many aspirants compute (50−30)/50 = 40% because they instinctively divide by the bigger number. Wrong. For change, the denominator is always the starting value (here 30), giving 66.67%. Train the reflex: "change-over-original".

:::compare

:::keypoints

• Always derive Profit = Revenue − Expenditure for every year before attempting questions.
• For "percentage increase", divide by the earlier value; for "% of revenue", divide by the same-year revenue.
• Write the Profit row above the chart on your rough sheet — it converts the graph into a table.
• 2/3 of dual-line DI questions are about the gap, not the individual lines.
• A ratio like 30 : 50 simplifies to 3 : 5; learn to spot small ratios fast.
• Read units (Rs cr vs Rs lakh) before computing — a wrong unit can cost the whole set.
• The "average" question takes the simple arithmetic mean of the Profit row.
• Cross-check by reverse-multiplying: 33.33% of 150 = 50. ✓
  :::

:::memory
"PRE first, percent later."
P rofit row, R ead axes, E xtract values — before you touch any question.
For percentages, remember C-O-O: Change divided by Original (Old) value.
:::

:::recap

• Compute the Profit row for all years the moment you see the chart.
• "% increase" uses the older year as the base; "% of revenue" uses the same year's revenue.
• 2021→2022: profit grew 30→50, a 66.67% jump; 2022 margin was 33.33%.
• Pre-computation turns a 5-question DI set into five lookups, saving ~60 seconds.
  :::

By data handling: Analog (measures continuous data like temperature/speed), Digital (counts discrete 0s and 1s — most common), and Hybrid (combines both, used in hospitals/ICU monitors). By size, smallest to largest: Microcomputer (PC, laptop) < Minicomputer < Mainframe < Supercomputer. Supercomputers (e.g., India's PARAM, developed by C-DAC) are fastest and used for weather forecasting and research; speed is measured in FLOPS. Mainframes serve large organisations (banks, railways) handling thousands of users simultaneously. By purpose: General-purpose vs Special-purpose (designed for one task, e.g., ATM, traffic signal). Remember: a Supercomputer is NOT the same as a mainframe — supercomputer = speed, mainframe = many users.

Computer Awareness in IBPS Clerk almost never tests deep theory — it tests mappings. Which generation goes with which machine, which decade goes with which invention, and which Indian milestone goes with which year. Learn the matches and the marks fall into your basket.

Definition: A computer generation is a phase of computing history defined by the hardware technology used in its central processor. The five generations are vacuum tubes (1st), transistors (2nd), integrated circuits (3rd), microprocessors / VLSI (4th), and artificial intelligence / ULSI (5th).

Definition: A supercomputer is a computer with extraordinarily high processing power, measured in teraflops or petaflops, designed for scientific simulations, weather forecasting, nuclear modelling and other compute-heavy tasks.

The Big Mapping Table

Memorise this and you have already won half the marks. ENIAC, EDVAC, UNIVAC, MARK-I all sit in the First Generation (vacuum tubes, 1940s–1956). ENIAC was built at the University of Pennsylvania in 1946, EDVAC followed using stored-program architecture proposed by von Neumann, UNIVAC-I in 1951 became the first commercially available digital computer. These were room-sized, heat-spewing monsters that needed special cooling and consumed enormous power.

The Second Generation (transistors, 1956–1963) brought the IBM 1401, IBM 7090 and CDC 1604. Transistors replaced vacuum tubes, machines shrank and reliability improved.

The Third Generation (integrated circuits, 1964–1971) introduced the IBM System/360 family and the PDP-8 minicomputer. ICs put many transistors on a single chip.

The Fourth Generation (microprocessors / VLSI, 1971–1980s) is the age of the Intel 4004, the IBM PC, the Apple Macintosh, and modern desktops. PARAM, Cray, SUMMIT — every modern supercomputer you have heard of — sits in this generation (and its modern extensions).

The Fifth Generation (AI, ULSI, parallel processing, 1980s onwards) is the era of natural-language processing, neural networks, voice assistants, and quantum-computing research.

The "Father" Names You Cannot Skip

The first digital computer concept was sketched by Charles Babbage in the 1830s — his Analytical Engine is the conceptual ancestor of every machine in your hand today. For this reason Babbage is called the "Father of the Computer." The ABC (Atanasoff-Berry Computer), built at Iowa State College between 1937 and 1942, is widely cited as the first electronic digital computer — a court ruling in 1973 even formally credited Atanasoff over ENIAC in a famous patent case. So in MCQs, "first electronic digital computer" → ABC; "first general-purpose electronic computer" → ENIAC. Both turn up.

For supercomputing, Seymour Cray is the "Father of Supercomputing," and the Cray-1 of 1976 is the textbook supercomputer. SUMMIT (built by IBM at Oak Ridge National Lab in 2018) was, for a while, the fastest supercomputer in the world.

India's Supercomputing Pride

The single most important Indian-context question on this topic is: India's first supercomputer is PARAM 8000, developed in 1991 by C-DAC (Centre for Development of Advanced Computing, Pune) under the leadership of Dr Vijay Bhatkar. The political backdrop matters — the United States had refused to sell India a Cray supercomputer, citing dual-use technology controls, and PARAM 8000 was India's bold response. Since then C-DAC has built PARAM 10000, PARAM Yuva, PARAM Siddhi-AI and PARAM Shavak. PARAM Siddhi-AI was ranked 62nd in the world (TOP500 list, Nov 2020) at the time of release.

Types of Computer by Data Processed

There is also a small "types" question that examiners love. Computers are classified, by the kind of data they handle, into Analog, Digital, and Hybrid.

An analog computer measures continuous physical quantities — voltage, pressure, temperature, speed. It uses electrical signals as continuous waveforms rather than discrete 0s and 1s. The speedometer in your scooter, the petrol-pump display showing the rising litre count, and the old slide rule are analog devices.

A digital computer processes discrete data — 0s and 1s. Every desktop, laptop and smartphone is a digital computer.

A hybrid computer combines both. The classic exam example is the ICU patient monitor, which reads analog signals (heart rate, blood pressure, oxygen saturation) and converts them into digital readouts for display. The petrol-pump dispenser that shows continuous litres while computing a digital amount in rupees is also a hybrid. So whenever the question says "continuous reading + digital output," the answer is Hybrid.

Speed Unit Chain

A staple one-mark question. The progression — each step is 1,000 times faster than the previous one:

millisecond (ms) = 10⁻³ s > microsecond (μs) = 10⁻⁶ s > nanosecond (ns) = 10⁻⁹ s > picosecond (ps) = 10⁻¹² s

So 1 ms = 1,000 μs = 1,000,000 ns = 1,000,000,000 ps. Modern CPU clock cycles are measured in nanoseconds; transistor switching speed approaches the picosecond range.

Why it matters

The exam's "computer awareness" section gives easy marks if you have memorised these mappings. There is no formula to derive. It is recognition — see ENIAC, pick "first generation"; see PARAM, pick "supercomputer / 4th gen / 1991 / C-DAC."

Real-world example

When IIT Madras' AQUA group used PARAM Siddhi-AI to study urban flood modelling in Chennai, they were running simulations on India's own supercomputer — a direct descendant of PARAM 8000 from 1991. Every kilometre your phone signal travels through a fibre cable is processed by chips whose ancestors were vacuum tubes inside ENIAC. The same lineage that started with Babbage's gears is now folded into ULSI silicon in your pocket.

Worked Example

Question: Match the following machines with their generation/type.
(A) ENIAC, EDVAC, UNIVAC
(B) PARAM, Cray, SUMMIT
(C) India's first supercomputer
(D) ICU patient monitor / petrol-pump dispenser

Solution:
Step 1: ENIAC, EDVAC, UNIVAC used vacuum tubes → 1st Generation.
Step 2: PARAM, Cray, SUMMIT are large-scale parallel processing machines → Supercomputers (4th generation and beyond).
Step 3: India's first supercomputer is PARAM 8000, built in 1991 by C-DAC, Pune, led by Dr Vijay Bhatkar.
Step 4: Devices that read continuous quantities and display digital readouts are Hybrid computers — ICU patient monitor and petrol-pump dispenser both qualify.
Conclusion: The mapping is (A) → 1st Gen, (B) → Supercomputers, (C) → PARAM 8000 / 1991 / C-DAC, (D) → Hybrid / Analog-Digital.

Common misconception

Many candidates believe Babbage's Analytical Engine was built and worked in his lifetime. It was not — Babbage designed it, but it was never fully constructed during his life. It earned him the "Father of the Computer" title because of the concept, not a working prototype. A second common error is to confuse PARAM-8000 (India's first supercomputer, 1991) with PARAM Padma (2003) or PARAM Yuva (2008). The "first" is always PARAM 8000.

:::compare

:::keypoints

• ENIAC / EDVAC / UNIVAC → 1st Generation (vacuum tubes).
• PARAM / Cray / SUMMIT → Supercomputers, 4th Generation onwards.
• India's first supercomputer = PARAM 8000, 1991, by C-DAC, Pune (Dr Vijay Bhatkar).
• Charles Babbage = "Father of the Computer" (Analytical Engine concept).
• ABC (Atanasoff-Berry Computer) = first electronic digital computer.
• Continuous readings + digital display → Hybrid computer (ICU monitor, petrol pump).
• Speed: ms > μs > ns > ps, each 1,000× faster.
• Modern CPU clocks are in nanoseconds; transistor switching nears picoseconds.
  :::

:::memory
"57 EDVAC, 91 PARAM, Babbage the brain, ABC the first electric." And for speed: "Millie, Mike, Nano, Pico" — milli, micro, nano, pico, each 1,000× faster.
:::

:::recap

• Generations are decided by hardware: vacuum tubes → transistors → ICs → microprocessors → AI.
• India's supercomputing story begins with PARAM 8000 in 1991.
• Hybrid computers blend analog input with digital output.
• Speed units descend in factors of 1,000: ms → μs → ns → ps.
  :::

The CPU (Central Processing Unit) is the 'brain' of the computer. It has THREE parts: (1) ALU (Arithmetic Logic Unit) — performs all arithmetic (+, -, ×, ÷) and logical (compare, AND/OR) operations; (2) CU (Control Unit) — directs and coordinates all operations, acts like a traffic police but does NOT process data; (3) Registers / MU — small high-speed storage inside CPU. Memory hook: 'A Control Register' = ALU + CU + Registers. The CPU is also called the microprocessor in PCs. Clock speed is measured in Hertz (GHz). Examples of input→process→output flow always pass through the CPU. The motherboard is the main circuit board that connects CPU, memory, and all components.

INPUT devices send data INTO the computer: keyboard, mouse, scanner, joystick, light pen, microphone, webcam, barcode reader, OMR, OCR, MICR (used on cheques!). OUTPUT devices send data OUT to the user: monitor, printer, speaker, plotter, projector. Trick devices: a Touchscreen and a Modem are BOTH input and output. A printer is output; a scanner is input. MICR (Magnetic Ink Character Recognition) is heavily tested — it reads the special magnetic numbers at the bottom of bank cheques. Memory aid: if a human feeds it → input; if it shows/gives a result → output. Monitors are measured diagonally; resolution is in pixels.

When a clerk enters a cheque: the MICR reader (INPUT) reads the magnetic code → data travels to CPU via the motherboard → ALU verifies the amount, CU controls the sequence → result stored temporarily in RAM (primary memory) → final record saved to hard disk (secondary storage) → a receipt is produced on a printer (OUTPUT) and shown on the monitor (OUTPUT). Note the device categories: MICR = input, printer/monitor = output, RAM = volatile primary memory, hard disk = non-volatile secondary memory. Ports connect devices: USB (universal), HDMI/VGA (display), Ethernet (network). This single chain answers many 'which is input/output/storage' questions in one go.

Is India a federal country or a unitary one? The honest answer is "both" — and the founders meant it that way. They wanted a strong Centre that could hold the country together in a turbulent moment, while still giving states meaningful autonomy in their own affairs. UPSC examiners return to this tension every year.

Definition: A federal polity is one where governmental powers are constitutionally divided between a national (central) government and regional (state) governments, with each operating directly on citizens within its sphere and neither subordinate to the other.

Definition: A unitary polity is one where all governmental power is concentrated in a single central authority, which may delegate functions to lower units but can withdraw them at will.

Definition: Quasi-federal is the label K. C. Wheare gave India — federal in normal times, but with significant unitary tilts that the Centre can invoke under stress.

Federal features of the Indian Constitution

A genuine federation has several non-negotiables, and India has them.

Dual polity — There are two sets of government, Union and State, each with its own legislature, executive and (in matters of law) jurisdiction. The Constitution gives each its own areas.

Written Constitution — A clearly written, codified Constitution is essential because it lays down the boundary between Union and States. Customary federalism does not work.

Division of powers — The Seventh Schedule splits subjects into three lists: Union List (97 originally, now ~98), State List (66 originally, now ~59), and Concurrent List (47, now ~52). Each level legislates in its sphere.

Supremacy of the Constitution — Both Union and States derive their authority from the Constitution; neither can violate it.

Rigid amendment for federal provisions — Articles concerning federal structure (Schedule VII, representation of states in Parliament, election of the President, and others listed in Article 368) require ratification by at least half the State Legislatures in addition to the special majority in Parliament.

Independent judiciary — The Supreme Court is the constitutional umpire that decides disputes between the Union and the States, and between States.

Bicameralism — Parliament has a Lok Sabha (people) and a Rajya Sabha (states), giving the units of the federation direct representation in central law-making.

Unitary features: the tilt toward the Centre

The Constituent Assembly knew that a young, partitioned India needed strong central muscle. So they wove in a long list of unitary provisions.

Single Constitution and single citizenship — Unlike the United States, no state has its own constitution (except Jammu and Kashmir before 2019), and every Indian is a single citizen of the Union, not separately of a state.

Appointment of Governors by the Centre — The Governor of each state is appointed by the President (effectively the Union Cabinet) and serves "during the pleasure of the President." This makes the Governor an agent of the Centre as much as the constitutional head of the state.

Emergency provisions (Articles 352, 356, 360) — During a national emergency, Parliament can legislate on State List subjects, and the federal structure essentially folds into a unitary one. President's Rule (Art. 356) lets the Centre take over a state government. Financial emergency (Art. 360) lets the Centre direct state finances.

Integrated judiciary — There is a single hierarchy of courts; High Courts and the Supreme Court enforce both Union and State law. Contrast this with the United States, which has separate federal and state court systems.

Flexibility of amendment — Most of the Constitution can be amended by Parliament with a special majority; only a small set needs state ratification.

All-India Services (IAS, IPS, IFoS) — Officers recruited centrally but serving in state cadres. This embeds a Centre-trained, Centre-controlled administrative spine inside each state.

Single election machinery — The Election Commission of India conducts both Parliamentary and State Assembly elections.

Governor's discretionary powers — Reserving bills for Presidential assent, recommending President's Rule, dismissing chief ministers in certain conditions — all let the Centre intervene in state affairs.

The constitutional language

Notice what the Constitution does not say. The word "Federation" appears nowhere in the document. Article 1 calls India a "Union of States." Dr. B. R. Ambedkar explained the choice in the Constituent Assembly: this was an "indestructible Union of destructible states" — meaning the Union itself cannot be broken up, although states can be reorganised by Parliament. Crucially, the Indian federation was not formed by an agreement among states (unlike the United States), and states have NO RIGHT TO SECEDE. This was an answer to the trauma of Partition.

The Supreme Court's verdict

In S. R. Bommai v. Union of India (1994) — a landmark 9-judge bench ruling — the Supreme Court declared federalism to be part of the basic structure of the Constitution. That puts federalism beyond the reach of constitutional amendment under Article 368. The Court also laid down strict guidelines on the misuse of Article 356, requiring that a President's Rule proclamation be testable in court — a major curb on the unitary bias.

Why it matters: UPSC Prelims often asks whether a particular feature is federal or unitary; Mains asks for analysis of "cooperative federalism," "competitive federalism," and the changing Centre-state balance under GST, NITI Aayog, and the 15th Finance Commission. Knowing both lists and the case law is the foundation for the entire Polity paper section on federalism.

Real-world example: When President's Rule was imposed in Arunachal Pradesh in 2016, the Supreme Court (Nabam Rebia case) struck down the proclamation, citing Bommai-era guidelines. The unitary tool (Art. 356) existed; the federal restraint (judicial review) reined it in. That balancing act IS Indian federalism in action.

Common misconception: Students sometimes equate "Union of States" with "Federation of States" and treat them as identical. They are not. "Union" was deliberately chosen because the Indian federation was not the result of a compact among pre-existing states; the Constitution created the federation top-down. Another error is to call India a "unitary state with federal features." The textbook answer, supported by both Wheare and Bommai, is federal with a unitary bias — or quasi-federal.

:::compare

:::keypoints

• Article 1: India is a "Union of States" — not a "Federation."
• K. C. Wheare: "federal in form, unitary in spirit" — quasi-federal.
• Federal features: dual polity, written Constitution, Schedule VII division, rigid amendments, judicial review, bicameralism.
• Unitary features (mnemonic in the body): single Constitution and citizenship, Governor as Centre's agent, emergencies, integrated judiciary, flexible amendments, All-India Services, single ECI.
• States CANNOT secede; Union is indestructible (Ambedkar).
• S. R. Bommai (1994): federalism is part of the BASIC STRUCTURE.
• Federalism in India is asymmetric: special status for some states under Articles 371 etc.
• The unitary tilt was a deliberate response to Partition and the need for national integration.
  :::

:::memory
"SAGE-IF AAG" — Single Constitution and citizenship, Appointment of Governors, Governor as agent, Emergency provisions, Integrated judiciary, Flexibility of amendment, All-India Services, single election machinery, Governor's discretionary powers. Nine unitary tilts in one phrase.
:::

:::recap

• India is federal in form, unitary in spirit — quasi-federal.
• Article 1 calls it a Union, not a Federation; states cannot secede.
• Federal core: dual polity, Schedule VII, judicial umpire, bicameralism.
• Unitary tilt: single Constitution and citizenship, central Governors, emergencies, All-India Services.
• Bommai (1994) locks federalism into the basic structure — Parliament cannot amend it away.
  :::

Smallest unit = Bit (binary digit, 0 or 1). 4 bits = 1 Nibble. 8 bits = 1 Byte. The ascending ladder, each step ×1024: Byte → KB (Kilobyte) → MB (Megabyte) → GB (Gigabyte) → TB (Terabyte) → PB (Petabyte) → EB (Exabyte) → ZB → YB. Memory aid for order: 'Kind Men Give Tasty Pizza Every... ' (K-M-G-T-P-E). 1 KB = 1024 bytes; 1 MB = 1024 KB; 1 GB = 1024 MB. A single character (like 'A') typically takes 1 byte (8 bits) in ASCII. Speed tip: to convert GB to MB, multiply by 1024; to go up a level, divide by 1024. These conversions appear directly in IBPS Clerk MCQs.

RAM has two types: DRAM (Dynamic) — needs constant refreshing, cheaper, used as main memory; SRAM (Static) — faster, no refresh needed, used in cache, more expensive. ROM variants: PROM (Programmable, write once), EPROM (Erasable using ultraviolet light), EEPROM (Electrically Erasable — basis of flash memory/BIOS updates). Flash memory (used in pen drives and SSDs) is a type of EEPROM. Example recall: cache uses SRAM; your laptop's 8 GB main memory is DRAM. CD-ROM = optical, read-only; CD-R = write once; CD-RW = rewritable. A typical CD holds ~700 MB, a single-layer DVD ~4.7 GB, and a Blu-ray ~25 GB. SSDs are faster than HDDs because they have no moving parts.

Computers use four number systems. Binary (Base 2): digits 0,1 — the language of computers. Octal (Base 8): digits 0-7. Decimal (Base 10): digits 0-9 — human system. Hexadecimal (Base 16): digits 0-9 then A,B,C,D,E,F (A=10, B=11... F=15). Memory aid for bases: B-O-D-H = 2-8-10-16. Each binary digit is a 'bit'. A group of 4 bits maps neatly to ONE hex digit (since 2^4=16) and is also a nibble — this is why hex is used as shorthand for binary. Knowing place values (...8,4,2,1 for binary) lets you convert fast. IBPS Clerk keeps conversions small (numbers under 16), so master 0-15 in all four systems.

To convert BINARY to DECIMAL, write place values from right to left as powers of 2: 1, 2, 4, 8, 16, 32... Then add the place values wherever there is a 1. Example: binary 1011 → positions (8 4 2 1) → 1×8 + 0×4 + 1×2 + 1×1 = 8+2+1 = 11. Reverse (DECIMAL to BINARY): repeatedly divide by 2 and read remainders BOTTOM to TOP. Example: 13 → 13/2=6 r1, 6/2=3 r0, 3/2=1 r1, 1/2=0 r1 → read up = 1101. Speed tip: memorise 8-4-2-1 columns; any number 0-15 fits in 4 bits. Quick checks: 1111=15, 1000=8, 1010=10.

Computers store characters using coding schemes. ASCII (American Standard Code for Information Interchange) uses 7 or 8 bits, representing 128 (or 256 extended) characters — capital 'A' = 65, 'a' = 97, '0' = 48. EBCDIC (8-bit) was used on IBM mainframes. UNICODE uses up to 16/32 bits and can represent characters of almost ALL world languages (including Hindi, Chinese, emojis) — over 1 lakh characters. Memory hook: ASCII = English/basic, Unicode = Universal/all languages. BCD (Binary Coded Decimal) represents each decimal digit by 4 bits. For IBPS Clerk: remember ASCII 'A'=65 and that Unicode is the universal standard supporting Indian languages.