Weekly Assignments

Week 1 | Week 2 | Week 3 | Week 4 | Week 5

Week One

Week of 10/30/2017

1. Video:

• Basics of computer memory and architecture:

2. Reading:

• HTTP Basics
• Premature Optimisation
• Good article on (numerical) data representation

3. Redis is a very popular in-memory database. This means it stores all data in RAM instead of most databases that store data on the hard disk. What are the advantages and dangers of this 100% in-memory approach?

4. Define the term "protocol". Provide 3 examples of protocols of human interaction.

5. Summarize the discussion on software engineering article. Provide your opinions on how much time you should spend making code optimal as opposed to moving on to the next task.

---

Answers:

Redis

Advantages:

• faster performance since it doesn't have to access the file system/disk
• smaller footprint since it doesn't need to carry the overhead of RDBMS

Disadvantages

• size of data set is limited to amount of available primary memory
• data is not persisted once the session ends

Protocol

Definition:

A system of rules that governs interaction between entities.

Examples:

• Military - It is customary for a soldier to stand at attention when a superior officer walks by.
• Diplomatic - Heads of State do not enter the country without the President's knowledge.
• Social - It is considered good manners to say "Thank you" when someone offers you a gift.

"Premature Optimization is the Root of All Evil"

Summary:

The gist of this article is that the statement by Sir Tony Hoare has been frequently taken out of context and, consequently, misunderstood by many software engineers. These individuals have mistakenly interpreted these words to mean that writing efficient code is not important. However, when taken in the proper context, one understands that Hoare was actually saying that being concerned about small efficiencies at too granular a level before you know what the actual performance bottlenecks are is not useful. This does not mean that optimization itself is not useful, rather, it should be part of the design process.

Key Points

• When designing software at a system level, performance issues should always be considered from the beginning. A good software developer will do this automatically, having developed a feel for where performance issues will cause problems.

• Software engineers should worry about other issues (such as good algorithm design and good implementations of those algorithms) before they worry about micro-optimizations such as how many CPU cycles a particular statement consumes.

• Efficiency is an important criterion when developing software. Software engineers must accept optimization as a necessary step and not reject it as evil.

• The only way to gain the experience needed to properly optimize code is by constant practice:

1. profile everything (timings)
2. be willing to make some mistakes

• Premature optimization is always bad, but the truth is that some concern about small efficiencies during program development is not premature.

• Design performance into an application from the very start. Design first, code second.

Week Two

Week of 11/6/2017

Video:

IEEE floating point representation

Reading:

Concerns with choosing data type sizes
C Data Types

Key Points

• Embedded Systems require more precision in choosing the best type of integer for efficient code.
• Some compilers allow for different integer sizes (not C99 compliant)
• There is a difference beween the natural word length of the CPU and that of the external memory bus. In this case, the optimal integer depends on where it is stored.
• Most of the embedded world uses unsigned integers, which are much faster and much safer than unsigned integers. ✨ ✨
• For embedded systems, use 'least' and 'fast' data types for best results:

Fixed width unsigned 8 bit integer: uint8_t
Minimum width unsigned 8 bit integer: uint_least8_t Fastest minimum width unsigned 8 bit integer: uint_fast8_t

Functions in C
Signed Binary Numbers
Slide Deck
Floating Point Numbers

Fun Extra Content:

Javascript based DDOS attacks

Assignment:

1. Consider the following 32 bit sequence

     11000100100000000010000000000000
   

What is the value of these 32 bits when interpreted as:

• a signed 32-bit integer
• an unsigned 32-bit integer
• A 32 bit IEEE floating point value

Answer:

• for a signed 32-bit integer, the first bit is the sign bit. Since it is 1, this value is between -2,147,483,648 and -1.
• (3,229,622,272) for an unsigned 32-bit integer, this value would be between 0 and 4,294,967,295.
• for a 32-bit IEEE floating point number, the first bit is the sign bit, followed by the exponent bit, followed by the mantissa (fraction bits).

2. When you declare a variable in C, storage for that variable is set aside in memory. This storage has a numeric address in memory. This address might look like 0x7ffeef8f3aa8. The "address-of" operator is the ampersand, &. This operator allows you to get the address of a variable. Observe the following program:

	#include <stdio.h>

	int main(){
	    int x = 10;
	    int y = 15;
	    printf("The address of x: %p\n", &x);
	    printf("The address of y: %p\n", &y);
	    return 0;
	}

The result of running the program on my computer (the values will differ for you!)

The address of x: 0x7ffeed495aa8
The address of y: 0x7ffeed495aa4

Create a program that uses print statements along with the address of operator to either prove or disprove the following claim:

When a variable is passed to a function, a copy of the variable is passed and not the original variable.

Answer:

A copy of the variable is passed to a function. The address of the original value was different from that of the value inside the function. (Refer to "value_or_ref" program.)

Week Three

Reading:

C Primer Plus - Chapters 1-3, 5-7
Complete exercises at the end of each chapter, except those exercises that are related to formatting text.

Week Four

to be continued...

Week Five

to be continued...