Another year has passed, I have added one more step to my life as a software engineer. Four years into the industry, it is long time now. In this year, I have groomed myself up.

I am no more the rough, lesser jumping to conclusions, lesser assuming stuff, improving listening and notes taking. Nowadays, I spend more time attending meetings, take the points, propose design. I have grown, but I don’t have a way to measure.

One more step forward is done. Stepping into the fifth year. All the best to all my friends who did join on the same day as me.

Today I was given the task of finding the age of a person, provided the birth date. When I was given the task, I just said, just use the DateDiff function. But it is not that simple.

One thing I wanted was to keep the code simple and in a few lines of code. Here goes my first iteration.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

         static void CalculateAge()
        {
            DateTime dateOfBirth;
            DateTime.TryParse("02/18/2008", out dateOfBirth);
            DateTime currentDate = DateTime.Now;
 
            TimeSpan difference = currentDate.Subtract(dateOfBirth);            
 
            // This is to convert the timespan to datetime object
            DateTime age = DateTime.MinValue + difference;
 
            // Min value is 01/01/0001
            // Actual age is say 24 yrs, 9 months and 3 days represented as timespan
            // Min Valye + actual age = 25 yrs , 10 months and 4 days.
            // subtract our addition or 1 on all components to get the actual date.
 
            int ageInYears = age.Year - 1;
            int ageInMonths = age.Month - 1;
            int ageInDays = age.Day - 1;
 
            Console.WriteLine("{0}, {1}, {2}", ageInYears, ageInMonths, ageInDays);
        }

But then, there were problems with this method. If the current date is ’06/18/2009′ and the birth date was ’04/18/2000′, it returns, 9 yrs, 2 months and 2 days. The 2 days part is wrong. I didn’t have any clue as to why it appears.

Then I went down to the basics, using elementary mathematics of subtraction. Here goes the second iteration of the code.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

        static void CalculateAge2()
        {
            DateTime dateOfBirth = new DateTime(2000, 6, 18);
 
            int ageInYears = 0;
            int ageInMonths = 0;
            int ageInDays = 0;
 
            CalculateAge(dateOfBirth, out ageInYears, out ageInMonths, out ageInDays);
 
            Console.WriteLine("{0}, {1}, {2}", ageInYears, ageInMonths, ageInDays);
        }
 
        ///
        /// Calculate the Age of a person given the birthdate.
        ///
        static void CalculateAge(DateTime adtDateOfBirth, out int aintNoOfYears, out int aintNoOfMonths, out int aintNoOfDays)
        {
            // get current date.
            DateTime adtCurrentDate = DateTime.Now;
 
            // find the literal difference
            aintNoOfDays = adtCurrentDate.Day - adtDateOfBirth.Day;
            aintNoOfMonths = adtCurrentDate.Month - adtDateOfBirth.Month;
            aintNoOfYears = adtCurrentDate.Year - adtDateOfBirth.Year;
 
            if (aintNoOfDays < 0)
            {
                aintNoOfDays += DateTime.DaysInMonth(adtCurrentDate.Year, adtCurrentDate.Month);
                aintNoOfMonths--;
            }
 
            if (aintNoOfMonths < 0)
            {
                aintNoOfMonths += 12;
                aintNoOfYears--;
            }
        }

And it works like a charm for all scenarios I throw upon it.

Scenario 1:
Current Date : 15 – 09 – 2009
Birth Date : 09 – 03 – 2000

Just a difference gives the result
Age : 6 – 6 – 9 Result is 9 yrs, 6 months, 6 days.

Scenario 2:
Current Date : 15 – 09 – 2009
Birth Date : 28 – 07 – 2000

Here since 15 – 28 < 0: we borrow one from the month and then add the no of days in the month to the current date. 15 + 30 = 45: 45 – 28 = 17 days
Then since one month is borrowed: 08 – 07 = 01 months
Remaining is normal difference
Age : 17 – 1 – 9 Result is 9 yrs, 1 months, 17 days.

Scenario 3:
Current Date : 15 – 09 – 2009
Birth Date : 28 – 12 – 2000
Here since 15 – 28 < 0: we borrow one from the month and then add the no of days in the current month to the current date. 15 + 30 = 45: 45 – 28 = 17 days
Then since one month is borrowed: 08 – 12: So we are supposed to borrow a year and add no of months in current year to current month. 08 + 12 = 20: 20 – 12 = 8 months
Since a year was borrowed: 2008 – 2000: 8 yrs
Remaining is normal difference
Age : 17 – 8 – 8 Result is 8 yrs, 8 months, 17 days.

But I know there must be better ways to do it. Feel free to drop in code snippets. Comments too are most welcome.

Little lately, I am obsessed in finding the best font to use, when I do the coding. I try to keep the same font across all editors and IDE’s that I use. The first font that I loved was ‘Courier New’ set at 10pt.

Then with VS2008 came in a new attraction for Consolas a true type font from Microsoft. Consolas is really good. And it appears even better on TextPad, jEdit and VS2008 IDE.

Last month, I began using Anonymous font. I would have sworn by it, if not for Monaco.

Now I use only Monaco. Smooth and pleasing! So what fonts do you use.

After installing rails, to test it, I created a rails application using the following command at the command prompt.

1

rails -d mysql todolist

Pointing to http://localhost:3000 displayed the expected Welcome message.

Welcome Message

But then clicking on the About your application’s environment displayed a error message like below.

Error in Rails

That’s the least thing I expected. Digging into the development.log file located in the project log directory, displayed the actual error message, “Client does not support authentication protocol requested by server; consider upgrading MySQL client”.

Looking into the web for help, the problem with the way rails (as a client) was trying to access MySQL server. Rails uses a old password hashing used in MySQL 4.1, and the current version that I am running is MySQL 5.0. This had caused the error. The fix is running the following script in mysql prompt.

1

mysql> SET PASSWORD FOR 'railsuser'@'host' = OLD_PASSWORD('password');

That fixes it and you would get the following screen.

Success message

Sorting the result in a DataTable’s select command can be done as below.

1
2

DataTable table = dataSet.Tables[0]; // Get a datatable from the dataset.
DataRow[] row = table.Select("id = 500", "name desc, age asc"); // Get from table, where id = 500 order by name desc and age desc.

The first argument to the select is the filter condition, the second argument is the sort option.

Trying to use,

1
2

DataTable table = dataSet.Tables[0]; // Get a datatable from the dataset.
DataRow[] row = table.Select("id = 500 order by name desc, age asc"); // Get from table, where id = 500 order by name desc and age desc.

will result in a error “Missing operand before order”.

C# has a ?? operator, which is called the ‘Null coalescing operator’. The ?? operator is a infix operator used on nullable types or objects. If the operand on the left is null, it returns the value of the expression on the right, else it is the left operand itself.

Here is an example showing how it is used.

1
2
3
4
5
6
7

    int result;
    int? num = null; // num is null
    result = num ?? 10; // sets result to 10.
    Console.WriteLine(result);
    num = 5; // num is 5.
    result = num ?? 10; // sets result as 5.
    Console.WriteLine(result);

Math.Round has been improved in C#.
Consider the below piece of code:

1
2
3

Console.WriteLine(Math.Round(10.4)); // Rounds to 10.0
Console.WriteLine(Math.Round(10.7)); // Rounds to 11.0
Console.WriteLine(Math.Round(10.5)); // Rounds to 10.0

There is nothing surprising about the first two statements.
In the third statement however, 10.5 is rounded to 10 not 11. C# provides for a way to specify how the middle point has to be treated.
A enumeration ‘MidpointRounding’ that defines how mid points are treated.

1
2
3
4
5

Console.WriteLine(Math.Round(10.5, MidpointRounding.AwayFromZero)); // Rounds to 11.
Console.WriteLine(Math.Round(10.5, MidpointRounding.ToEven)); // Rounds to 10.
Console.WriteLine(Math.Round(11.5)); //Rounds to 11.
Console.WriteLine(Math.Round(11.5, MidpointRounding.AwayFromZero)); // Rounds to 12.
Console.WriteLine(Math.Round(11.5, MidpointRounding.ToEven)); // Rounds to 12.

‘AwayFromZero’ – Rounds the number to the next highest value.
‘ToEven’ – Rounds the number to Even number.

So for a odd fraction, the default round will round it to the lesser number and any of the above overloads will take it to the Next number.

C# 2008 allows for creating implicit variables using the ‘var’ keyword. But the usage of ‘var’ cannot be truly justified for just declaring a ‘int’ or ‘string’ in code as shown in example below.

1
2
3
4

var intNum = 5;
 
Console.WriteLine("intNum is a: {0}", intNum.GetType().Name);
Console.WriteLine("intNum is defined in: {0}", intNum.GetType().Namespace);

The output of the above program is

1
2

intNum is a: Int32
intNum is defined in: System

Here instead of using int in the declaration, we have used ‘var’ keyword. There is no difference between using ‘int’ or ‘var’ in the above example.
The real usage of ‘var’ comes in LINQ. Consider the following code.

1
2
3
4
5
6
7
8
9
10

int[] numbers = { 10, 20, 30, 40, 8, 7, 6, 2 };
var resultSet = from i in numbers where i < 10 select i;
 
foreach (var i in resultSet)
{
Console.Write("{0}", i);
}
 
Console.WriteLine("resultSet is a: {0}", resultSet.GetType().Name);
Console.WriteLine("resultSet is defined in: {0}", resultSet.GetType().Namespace);

Here resultSet is declared as a ‘var’. From the code, we understand that anytime, the resultSet will be an array of integers. But ‘resultSet.GetType().Name’ gives a surprising result.

1
2

resultSet is a: d__0`1
resultSet is defined in: System.Linq

So ‘var’ has its best usage in LINQ. So why use it in a normal program when the datatype can be used in itself.

Here is a tip shared by Kannan, a colleague of mine on how to avoid flickering when rendering controls dynamically in a Windows Application.

Enable double buffering, so that the flickering does not happen.

Add the following after the ‘InitializeComponent’ method.

1
2
3

      SetStyle(ControlStyles.UserPaint, true);
      SetStyle(ControlStyles.AllPaintingInWmPaint, true);
      SetStyle(ControlStyles.DoubleBuffer, true);

That’s it.

Yesterday I was working on the layout for a new site. I missed the DOCTYPE declaration in the HTML. The page had the main content included inside a div.

Following is the CSS that I used.

1
2
3
4
5
6
7
8
9

body {
      font-family: Georgia;
      font-size: 14px;
}
#main-content p{
      margin: 0;
      padding: 0;
      padding-bottom: 15px;
}

This worked perfectly and rendered well on Firefox and Chrome. But when I opened the same page in IE, the font was too large for all content inside the ‘main-content’ div. Also in order to center the page, I had used ‘margin: auto’ on the page container that encompasses all content in the body as below

1
2
3
4

#page-container {
      width: 760px;
      margin: auto;
}

This too didn’t seem to work in IE as the content was left aligned instead of centered, but rendered ‘centered’ in Firefox and Chrome.
Looking out for a solution on the net, I found that missing the DOCTYPE declaration was the cause for the issue. Here is the reason.

Without the DOCTYPE declaration, the browser has rendered the page in quirks mode. Adding the DOCTYPE made the browser render the page following the rules set by the standard.

So never forget your doctype declarations.