A-Level Computing Key - Terms & Concepts Cheat Sheet by 0llieC

Mother­board

The central interface for all the components of a PC. Everything connects to the mother­board via slots, wires, readouts and connec­tors.

Secondary Storage

Used to store programs and data. It can be partit­ioned to allow for dual-b­ooting multiple operating systems.

Arithmetic Logic Unit (ALU)

The ALU carries out arithmetic calcul­ations and compar­isons. The result of any calcul­ation is sent to the Accumu­lator

Clock

The clock is the part that regulates the cycle of the CPU. It provides a regular pulse of high voltage then low voltage. This high-low transition is a cycle, each cycle is an instru­ction

Memory Buffer Register (MBR)

Values fetched from memory are sent to MBR.

Current Instru­ction Register (CIR)

The instru­ction currently being execut­ed/­decoded

Address Bus

Carries the memory location of the instru­cti­ons­/data being received.

Cache Size

Cache is a small amount of very fast memory. Repeatedly used instru­ctions and data is stored in the cache for quick access. The bigger the cache, the more can be stored on it thus reducing processing time

Pipelining

Where the stages of the F-D-E cycle are 'stacked' so that they can be processed at the same time. While one instru­ction is being fetched, the previous is being decoded. This may not necess­arily increase processing time but throughput is increased.

Issues
• If an instru­ction requires the result of a previous instru­ction, the CPU will remain dormant  leading to 'bubbl­es'­/pi­peline stalls in the pipeline.
• Jumps  lead to the pipeline having to be flushed due to the change in instructions

Hyper-­Thr­eading  Where the CPU is intell­igent enough to fill the bubbles caused by pipeline stalls with other non-de­pendant instru­ctions from separate threads.

Processor Archit­ecture

Von-Neuman

The Von-Ne­umann archit­ecture is commonly used in most PCs. It stores both programs and data in the same memory. Using the F-D-E cycle, it carries out a single instru­ction at a time.

Pros
+ More Robust than Harvard (older)
+ Single Storage

Cons
- Each Instru­ction takes 2 cycles (fetch­/decode and execute)
- Cannot implement pipelining

RISC

RISC (Reduced Instru­ction Set Computing)

• Only one value fetche­d/s­tored per instru­ction cycle
• Less Power Required
• Used in smaller devices (Smartphones)
• Generates less heat - requires less cooling methods

SISD

Single Instru­ction, Single Data-s­tream
Single Core CPUs

NB:
• No parallelism
• Single CU, fetches single instru­ction

MIMD

Multiple Instru­ctions, Multiple Data-s­treams
Multi-core CPUs

NB:
• Multiple autonomous processors simult­ane­ously executing different instru­ctions on different data
• Uses either one shared memory space or a distri­buted memory space.

Specific Applic­ation

A piece of software that has a specific purpose, such as order entry, payroll, stock management etc. It may be Bespoke (made to order) or Off-th­e-shelf (designed to be used in a variety of situat­ions).

Open Source vs. Closed Source

Open Source  Source code is readable to anybody and freely modifiable.
Closed Source  Executable only, source code is kept hidden.

Pros of OS
• Free (usually)
• Community Coding/Bug Fixing - Usually faster than any closed-source
• Customisable
• Freedom to do what you like

Pros of CS
• Profes­sional Develo­pment
• Lower security risks
• Well documented and customer support

Firstly the code is striped of anything uneeded such as comments and redundant whitespace

The code is then divided into Lexemes (the smallest 'unit' of code).

Tokens are then assigned to each lexeme indicating what it is. Some token examples:
• Identi­fiers - for variables, subrou­tines, classes etc.
• Keywords - new, if, for, while etc.
• Operators - +, - , / , == etc.
• Literals - fixed numbers and strings
• Symbols - {}, (), ; etc.


Errors are caused when a lexeme cannot be assigned a token

The stream of tokens generated is then analysed to check they match the rules of the language. Tree data structures are often used in this process.

An example of valid syntax would be:

String word = "­Hello, World!";
Datatype Literal Operator String­Literal Symbol


Errors occur when a series of tokens cannot be matched to a rule, such as multiple datatypes.

The stage where code is checked for logical errors. For example:
• Datatype mismatch - assigning a String to an int
• Undeclared Variables, or out of scope variables
• Multiple variable declarations
• Array out of bounds with an integer literal.

Errors occur when one of the rules are broken. NB Not all semantic errors can be caught during compil­ation. For example accessing an array with a integer variable is logically fine, but the integer value may be out of bounds causing a run-time error.

The code is then converted into interm­ediate code (Java to Java bytecode, where it remains until use - not all languages do this)

Interm­ediate code is machine indepe­ndent

The code is then optimised, so that it runs faster and requires less resources, but still having the same output.

The final stage of compil­ation is the conversion to machine code. This process has to be repeated for each processor as it is machine dependant. Specific optimi­sations are also done on the separate processors as code that works well with one instru­ction set may not work as well with another.

Black Box (Alpha)

Testing that examines the functi­onality of a applic­ation, without looking at its internal code/s­tru­ctures.

Top Down

Testing of modules and sections of code that aren't yet implement. Testing the behaviour between modules.

Bottom Up

Testing each part of the applic­ation indivi­dually then testing the parts that rely on the sectio­n/m­odule.

Test Data

A range of test data must be used to properly test a system. It should include:
• Normal Data
• Boundary Data
• Standard Incorrect Data - incorrect data that could easily be entered
• Standard invalid data - e.g. text into numeric fields
• Extreme data - data that would never be entered normally, used to test the limits of a system

Class

A 'bluep­rint', a combin­ation of attributes and methods that create an object.

Encaps­ulation

Where attributes and methods are wrapped in their objects. Access modifiers control how the methods and the attributes can be accessed, whether that be by any class, or only within its own class.

Abstract Class

A class which contains attributes and methods like a normal class, but the class itself cannot be instan­tiated. An example is an Animal, you can write an abstract class, but you cannot create just an Animal.

HTML (Hypertext Markup Language)

The standard language for displaying webpages. A HTML document starts with
<!DOCTYPE html>
<html>

It consists of tags which are opened and closed <ta­g><­/ta­g>. Each document has a head and a body.

<a href="h­ttp­://­www.go­ogl­e.c­om">Link Text</­a>

<p>­Normal Paragraph text</­p>

JS (JavaS­cript)

JS is an interp­reted code that adds intera­ctivity to websites. It works on virtually all hardware and is used on nearly all websites. For example there are currently 300,000 JS repos on Github, Java has 200k

PageRank Algorithm

Google's algorithm that calculates the weighting of webpages. All pages have an initial rank, but for each link, it gives a certain amount to the webpage linked. Other algorithms are also used to give pages different rankings depending on what the user is searching for.

Server­-Side Processing

Processing performed on the server. The code is only viewable to people with access to the server­-files. It processes requests and serves a webpage based on the requests

Data Protection Act

8 Principles:
1. Personal data must be obtained lawfully and fairly
2. P.D. must be held for a specified purpose
3. P.D. must be adequate, relevent and not excessive
4. P.D. must be kept up-to-date and accurate
5. P.D. must not be kept longer than necessary
6. P.D. must be processed in accordance with data subjects rights
7. P.D. must be kept securely
8. P.D. must not be transf­erred outside the EU without permission

Copyright, Designs & Patents Act

Protects indivi­dua­ls/­org­ani­sations intell­ectual data. It protects:
• Income for the authors - Allows the author to license the data.
• Cost of creating the product - some products can cost thousands to produce
• Quality of Produce - pirates often alter products to bypass security
• Alteration Protection - Altering programs can have unintended aftere­ffects.

Computers in the Workplace

• Big Brother concern - an employer could watch over employees
• Reduced Produc­tivity - employees can do multiple things at once, which may reduce produc­tivity as employees may 'waste' time

AI

This is the one of the biggest issues, as AI use is rising among recent years. The issues are the same as automated decision making, but more issues arise when you consider cognit­ive­/when is a computer considered alive?

Censorship

Moral concerns are raised at whether the internet should be censored, would it be restri­cting the freedom of inform­ation. The issues arise when consid­ering adult content, and piracy.

Personal Inform­ation

Rises a privacy concern. Computers can now monitor peoples inform­ation and collect it. When does this become a breach of privacy.

Offensive Material

Computers are general purpose, what people do with them can be considered offens­ive­/mo­rally wrong e.g. cyberb­ull­ying, which can have drastic effects

Array

A data structure used to whole elements of the same data type.

2D Array

An array with 2 dimens­ions. It is commonly used to represent coordi­nates or a table, with the indexes relating to rows/c­olumns

Linked List

A data structure where each element in the list points to the next one. This makes is very easy to add/re­mov­e/r­eorder elements as only the pointer needs to change each time.

Circular Queue

The end of a queue linking back to the beginning.

Graph

A set of nodes/­ver­tices connected by edges.

Bi-Dir­ect­ional Graph

A graph where the edges have 2 way direct­ions. 

Binary Tree

A tree where each node has a maximum of 2 sub-nodes. Nodes with no child nodes are called leafs, and the edges, branches.

Input Device

A device (piece of computer hardware equipment) that is used to provide data and control signals to an inform­ation processing system such as a computer or inform­ation appliance. Examples of input devices include keyboards, mouse, scanners, digital cameras and joysticks.

SDD/Flash

A storage medium that has no moving parts. It uses a data controller to control the read/write of data. 2 rules of the data controller:
1. You can combine pages to form a block, but a block cannot overwrite individual pages
2. Before writing to a memory location, the page previously allocated must be erased.

Pros
• Low Latency Time
• Fast Transfer speed

Cons
• More Expensive

Capacity vs Cost
• Largest: ~4TB
• Roughly 30p per GB (£0.00­000­000­0291625 per Byte)

Speeds

Solid State: 200 to 2500 MB/s
Hard-Drive: 1030 MB/s
Optical (x1 Speeds):
- Blue-ray: 4.29 MB/s
- DVD: 1.32 MB/s
- CD: 0.15 MB/s

RAM

The 'working' area of the computer. Programs and data currently in use is stored in the RAM. On startup the BIOS loads the OS into the RAM.

Charac­ter­istics
• Random Access - allows data items to be read or written in almost the same amount of time irresp­ective of the physical location of data inside the memory.
• Volatile  emptied on power down
• ~1-16GB

BIOS/UEFI

Basic Input Output System­/Un­ified Extensible Firmware Interface The BIOS is preforms the hardware initia­lis­ation during the bootup, and provides runtime services between the OS and hardware. UEFI was designed to be the successor to the BIOS

Operating System

Software that provides:
• Process Management
• Memory Management
• Device Management
• User Interface
• File Management


Fundam­entally its software that manage­s/i­nte­rfaces computer hardware and software

Memory Management

A OS must manage the computers memory including adding­/re­moving programs and data from RAM, allowing multiple programs to be run at the same time. The OS also reallo­cates memory when it is no-longer in use (i.e. when a program is closed)

Paging vs Segmen­tation & Virtual Memory
• Segmen­tation.  Memory is split into variable sized blocks, and programs are segmented, with each segment being a logical divider. A segment table then maps segments onto memory blocks. Generally slower than paging due to the placement algorithm
• Paging.  RAM is split into fixed sized blocks - frames. Programs are split into same-sized blocks - pages. Any page can be placed in any frame, easy to allocate as all equal size.
• If the RAM is full. Pages are transf­erred to the secondary storage acting as memory - Virtual Memory. Pages are moved in/out as needed.
• Thrashing  is when pages are being constantly swapped between RAM and V.Mem. It can cause speed issues as the secondary storage's speed << RAM's speed.

Process Management

Involves the scheduling and switching of programs and threads. Modern PCs have 'multi­tas­king' but it is just clever schedu­ling.

Types of OS

Embedded
• Mostly hidden in devices, generally within the hardware themselves.
• Built into objects
• Have a dedicated purpose
• Little/no user interface
• Fully Autonomous
• Use limited resources - only whats required

Multi-­Tasking
• Several progra­ms/­pro­cesses at the same time (concurrent).
• Can either be process management or through parallel processing
• Most General Purpose OS' are now Multitasking

Multi-User
• Must be a multi-­tasking OS too
• Several users accessing the proces­sor­/pr­ogr­ams­/re­sources at the same time.
• Usually a round robin approach.
• Shared processing.

Real-Time
• Inputs being processed under strict time limits. For requirements:
1. Support Non-Se­que­ntial programs
2. Handle parallel and unpred­ictable events
3. Produce responses within the time limit
4. Have fail-safes to guarantee response time

Distri­buted
A collection of indepe­ndent nodes, each with its own hardware. The OS presents the systems as an indivi­dual. For example: AI; Weather Foreca­sting; Online Shopping. Each may have the main system on one server, and other things processed on another. The pros of this are that it reduces the load on one computer, and if one fails, it may be able to continue.

Object Orientated

Code is divided into objects which possess state and behaviour. Follows the principles of encaps­ula­tion, abstra­ction, inheri­tance and polymo­rphism.

Data Query Languages

Queries to a database or other data structure are specified by what is wanted rather than how to get it.

Procedural

Allows structured progra­mming with sequence, selection, iteration and recursion. Code can be made modular with the use of proced­ures.

Assembly Languages

One to one corres­pon­dence between lines of code and processor instru­ctions. Unlike raw machine code however, you can have variable names and labels.

Lossless

Compre­ssing a file without the loss of data

Run Length Encoding (RLE)

RLE identifies repeating patterns of data and stores a copy of the inform­ation and how many times it occurs in succes­sion.

Symmetric Encryption

Encryption that uses the same key to both encrypt and decrypt.
Uses: Encrypted Harddrives

Client­-Server Commun­ication

The client generates a session key and uses the public key to encrypt it

The server decrypts the session key with the private key

Client-Server now commun­icate using symmetric encryption with the session key

Public Key

The public key is the product of the 2 prime numbers making the private key. As no efficient non-qu­antum integer factor­isation algorithm exists, it is practi­cally impossible to crack the private key by brute force.

Database

A structured system to hold data

Flat File Database

A flat file database is a database that stores data in a plain text file. Each line of the text file holds one record, with fields separated by delimi­ters, such as commas or tabs. While it uses a simple structure, a flat file database cannot contain multiple tables like a relational database can.

Attribute

A feature of the entity

Composite Primary Key

A combin­ation of 2+ fields that act as a primary key

Secondary Key

A key that is indexed to allow for faster searching. There can be multiple secondary keys and they don't have to be unique.

2NF

• 1NF
• If the primary key is a composite of attributes (contains multiple columns), the non-key attributes (columns) must depend on the whole key.

READ

SELECT fieldNames FROM tableName WHERE fieldName = value ORDER BY fieldName

DESTROY

DELETE FROM tableName where fieldName = value

DROP tableName

Atomicity

Transa­ctions are either done, or not done. Never partially applied

Isolation

Transa­ctions preformed simult­ane­ously must have the same result as if they were preformed sequen­tially

Record Locking

Making a file read-only to anybody else who opens the file while changes are being made.

Serial­isation

Create a clone of the data item, so the user can make changes, then upload a copy of the clone to the database. This will ensure that no updates or changes can be lost due to uploading a copy of the local version.

Standard

A definition or a format that has been approved by a recognised standards organi­sation.
de jeur (by force of law) or de facto a standard that has just been accepted over time

LAN

Local Area Network

• Geogra­phi­cally Small (build­ings/a site)
• Equipment is generally owned by the compan­y/p­eople using it
• Generally Faster
• Uses layers 1 and 2 devices - hubs/switches

Network Topologies

Bus Topology
• All devices connected to a central cable (backbone)
• Devices have equal rights
• Collisions can occur if multiple devices send data at once

Star Topology
• A hub at the centre of the network. Requests are sent to all other devices connected to is
• The hub reads the packets and determines the MAC address of the recipient

Ring Topology
• A token is passed around the ring until one of the devices requests to use it.
• The token is filled with the frame of data
• It is passed around the network to each device until it reaches the recipient
• Recipient acknow­ledges the data has arrived.

Peer-t­o-Peer

• All computers have equal rights and act as both a client and server
• Popular applic­ations include the BitTorrent Network, and BitCoin

Pros
+ Easy to set up
+ More reliable as central depend­encies are eliminated
+ No-need for a system admini­strator as every user is the admin of their machine
+ Cheaper to implement and maintain.

Cons
- Difficult to administor as there is no central dependency
- Less security therefore viruses and other malware can easily be transmitted
- Data recovery is difficult as there is no central storage, each computer requires ots own backup system
- "­(Lots of movies, tv shows and music are transf­erred using P2P, via torren­ts)­"

Circuit Switching

Commun­ication where a dedicated channel (or circuit) is establ­ished for the duration of a transmission.

Pros
+ Data arrives in order sent
+ No additional inform­ation has to be added - e.g. headers

Cons
- Portion of the network is unavai­lable while in use
- Data is easily interc­epted.

MAC Address

Unique 6-byte identifier that is given to NICs. Assigned to the NIC by the manufa­cturer.

IPv6

IPv6 is a 128-bit address, so there are 2¹²⁸ addresses available.

HTTPS (Secure Hypertext Transfer Protocol)

Connects via a different port and encrypts the data between the HTTP and the TCP protocols.

POP3 (Post Office Protocol 3)

Allows emails to be received from a server. The protocol connects to the email server, downloads a local copy then deletes them from the server.

SSH (Secure Shell)

Remote­-access protocol, allows secure commun­ication between a client and server

Worm

A virus but it is contained within its own program.

Ransom Ware

A trojan­/worm that encrypts data and then charges the owner to decrypt it.

Proxy Server

f a user requests a service from the network, it is first passed to the proxy, before the proxy server then performs the request on the behalf of the network user. If the resource is banned the request can be rejected, There is never any direct contact between user and resource, as the proxy acts as a "­middle man".

Switch

A switch has a small amount of internal memory, that allows it to generate a look-up table. When data is sent the switch finds the approp­riate node. Unlike a hub, it doesn't send the data to all the nodes, just the receiver.

Gateway

The entrance and exit of networks. The main use is to connect multiple networks with different archit­ectures

Overall Search engine

Crawling the web with 'spiders' TF-IDF (Term Frequency - Inverse Document Frequency) The PageRank algorithm Other factors, such as domain name, page age, mobile friend­liness

Page Rank

PageRank works by counting the number and quality of links to a page to determine a rough estimate of how important the website is. The underlying assumption is that more important websites are likely to receive more links from other websites.

Integer

A whole number

Boolean

A value with a True or False condition - can possibly use 0/1 instead

String

A set of charac­ters, used to store text.

Binary (Base 2)

Groups of bits in 1 and 0 (2 possible values). It uses positional numbering but with powers of 2, not 10 (denary numbers - normal)

Denary (Base 10)

Normal number formats with the positional numbering in powers of 10.

Two's Comp

A improved way of showing negative numbers. If the first bit is a 1, it is taken as the negative version, and all following numbers are added to it. If it is a zero, it behaves just like a negative number.

To change a normal +ve binary number to twos comp. Flip the bits, and add 1
E.g.
01101010 = 106
10010101 (Flip the bits)
10010110 = +1
10010110 = -106

Floating Point Binary

A method of showing binary composed of 2 parts: the mantissa, and the exponent. The mantissa is the actual number, and the exponent is the number of units to move the floating point up or down by.

10101011 | 0011
Mantissa | Exponent

1.0101011 x 2 ⁰⁰¹¹
1.0101011 x 2³
1.0101011

1010.1011 = 10.6875

The rule for powers:


(Decrease the power, per jump right)

Overflow

A calcul­ation that results in a number too large to be stored. For example, any numbers past x10¹⁰⁰ on most calcul­ators

Half Adder

Add two single bits, produces an output S, and a carry signal C. It consists of an AND gate (C) and a XOR gate (S) in parallel.

Basic Flip-Flop

Takes a set and a reset signal. The idea is that the FF stays in one state until the change signal is sent.