Building Dynamic Sites– The Database Way

I’m sure you are quite familiar with HTML pages or web-pages, as they are

popularly called. These are pages stored on the web server, which you can

download into your browser by typing in the website address or by clicking on a

hyperlink that points to that page. This seems rather easy. So we’re not

looking at simple web pages or websites. In this article, we would be looking at

something far more complex–namely, dynamic web pages and websites!

The Need for Databases



Let me start off with a story.

Whichever city or town you live in, I’m sure you have a favorite sabzi-wala,

who comes around to your house every day with his cart piled up with greenish

potatoes, stale onions, sorry-looking peas, and beans that may have been fresh a

week ago. However, that is not the point. The point is that he sells you

something every day. And for example’s sake let’s call him Haribhai. Thanks

to the online medium, your good old honest sabzi-wala need not visit you any

more. Why? Because



he’s launched his own website, www.sabziwala.com.
And, his young nephew, who is under training to become a proud sabziwala in his

own right some day, delivers whatever you order on the Internet.

Haribhai has engaged a leading American multinational (headquartered in

Boston, Massachusetts) to create his e-commerce software, which is naturally

full of HTML pages, each page containing vegetable listings, package sizes,

prices for each, time to deliver, and so on and so forth.

Suddenly one day there is a rumor that the transporters of the area may go on

strike. And Haribhai realizes with that he can now charge twice as much for the

same old stale vegetables. But when he goes to the MNC, to ask if they could

modify the rates in his website, he is in for a big shock. "Sorry, sir, but

we’ll have to change the HTML pages. And that means you need our programmers

all over again. And that will mean an additional sum of about $10,000. But since

you are a valued customer and a major account for us, we’ll cut it down to

just $8,000," came the reply.

Haribhai is shattered. How can he possibly make up this money even if he

charges three times the normal rate. And then he gets another terrible thought–tomorrow,

when the rumor is squashed by the appropriate government official, he would have

to bring the prices down again. So he would again need to go back to the

multinational - $8,000 once again. And indeed every time he changed the price of

his vegetables; the gainer would only be the software company!

Apart from feeling sorry for Haribhai, can we really offer a solution?

Indeed, there has to be a solution, because the data that you put out on your

HTML page needs to be changed at regular intervals. In fact, there are many

software companies that have moved from HTML websites to dynamic websites, which

use relational databases.

Dynamic and Static Sites



A database in the basic sense can be defined as a collection of data–obviously

organized in some manner. And you can query this database to get whatever data

you want. Databases are an essential component of any website today, and in fact

of any computer-based application, whether on the Internet or otherwise. So let’s

take a look at Haribhai’s vegetable database. Assuming he needs to store three

pieces of information–name of the vegetable, size of packing, and price. The

database would then be in the form of a table (See Table 1).

Notice that this table defines a "relationship" between name of

vegetable, size of pack, and price of pack. Consequently it doesn’t take much

imagination to see why such a database is called a relational database. In fact

the table itself is often called a relation!

Now let’s see how we can query such a database. If we wanted to get the

price of a half kilo pack of onions, we would give a command to the database

which would look something like this: "Get the price of pack for the row

where name of vegetable is ‘onion’, and size of pack is ‘0.5 Kilos’

And naturally, the result will be another table–this time a smaller one.

And the result of the query is price of pack–Rs 20. Notice that the result of

the query doesn’t need to give the first two entries (name of vegetable and

size of pack). Because you know that anyway–after all you were the one who

asked the question. It’s like an exam where the student doesn’t have to

write down the entire question along with the answer–just the answer will do,

as long as he writes down the question number.

This type of querying is usually done in a standardized language called SQL,

which stands for structured query language. Obviously it is more formal than the

simplified stuff I have been showing you, but the essence is the same. Now,

Haribhai’s problem can be finally solved

Because now, whenever Haribhai needs to change prices, he doesn’t need to

change the entire HTML page. He doesn’t need a programmer. Most of all, he

certainly doesn’t need the Boston-based multinational. All he needs is someone

who understands simple English, who can enter this data into the database every

time it changes. In fact, his 12-year old niece who goes to an English Medium

school could do the job.

Another advantage of having an improved website is that it gives you one more

set of technical jargon that you can brag about, namely dynamic sites and static

sites. A static website is one, which consists of HTML pages that are permanent

and cannot be easilty modified or updated. Any changes in such a site require an

entire re—programming of the HTML coding. Therefore such sites are useful in

cased where infomation does not have to be changed very often.

On the other hand, dynamic sites are those where the web page is created

dynamically–as in our example–through a database, instead of being

permanently coded in HTML. As discussed, these sites are useful when some of the

HTML pages need the flexible option of editing and upgrading a part or the

entire data. In other words, these kind of websites can be called that

database-driven sites as they are more maintainable than plain HTML websites,

which are simply based on static HTML pages.

So, where does all this information that is constantly being updated get

saved and stored? This data is resting on a separate server called the data

server. The data server is generally connected to the web server through a LAN

or local area network link (See Fig. 1).

Fig 1: The data server and web server are connected through a Lan link

At end of the day, Haribhai got a solution to his problem, and without

burning a hole in his pocket! Needless to say, his website is a roaring sucess.

DHRUV NATH is a Professor at MDI,

Gurgaon.