WebMaster icon, Lecture 2

Review of Browsing Techniques

"Trust Browser" Philosophy

Issues caused by Different Browsers

Netscape Communicator
Internet Explorer
Lynx (all text)

View > Page Source...

Save As... (source)

Plain text (ASCII text)



Prev | Next | The 'Old' Way to Make Your Own Home Page


The World Wide Web (WWW or just "the web") is composed of many web server computers that are constantly on waiting for requests. Browser programs, such as Netscape or Explorer, are the clients and they ask for certain pages. Browsers and web servers communicate with one another using the HyperText Transport Protocol (http). Most information sent from server to browser is plain text with special HyperText Markup Language (html) codes (called tags) embedded in the text. The browser takes the html page and converts the codes into actions the viewer sees on the screen. The browser is most familiar with the type of computer and operating system that it is running on. For instance, the browser program knows what font looks best on the screen, how wide the window is, what font size to use to get large headline text, etc. Thus, the server sends pages that say, in essence, "Display this text in your largest font size, centered and boldface". The browser determines the actual font used, how to center and boldface the text. Therefore, there is less information sent over the net to describe a page.

"Trust the browser" is the overriding philosophy of the web. Previous attempts to distribute information on the net failed because they were tied into precise descriptions of terminal types (i.e., "VT100", etc.). By letting the browser make decisions about how to format the text for this particular computer, the web is a cross-platform application. As long as there is a browser written for your computer, you can access any server on the web.

The disadvantage of such a philosophy is a loss of control by authors of web pages. You no longer have control over exactly how your pages will look. In fact, a user can override the browser's choices by changing preferences. For instance, a user could decide the basic type size she would like to view is 14 point rather than the default 12 point. That would cause every page sent to her to be wrapped differently than other browsers receiving the same web page. There are even differences in how different browsers running on the same computer handle the same html command. If a browser does not support a certain html command, it will just ignore the command. For example, Netscape extended the html language by adding a <center> command. This allows it to center images and text. Mosaic did not support the center command, so ignored it if it received such a command in a page. Lynx is a text-based browser, cannot display images at all and also did not support the center command. As a result, many sites have gone to two duplicate sets of web pages - one set for graphical browsers and one set for text browsers. Or, even worse, some web authors are warning people that their pages are "Netscape friendly" (using the latest extensions that only Netscape has) and, thus, are almost unviewable by other browsers. A major factor affecting an author's decision is that 60% of people on the net used Netscape as their browser. But now almost 90% use Explorer (as a result of some tactics mentioned in the anti-trust suit filed against Microsoft by the Justice Department).

A web page is constructed out of plain text. The web page author then embeds html commands within the text. It is a process very similar to the way some of the original word processors added style to plain text. You would insert a special code to "turn boldface on" and, later in the document, you would insert a special code to "turn boldface off". In html, you insert <b> to turn boldface on and </b> to turn it off.

Most html commands are containers. You have a command to turn something on and the same command with an added slash ("/") to turn it off. (But there are also some "container-less" commands!) All commands are found between angle brackets - < >.

A good way to begin learning html is by using the View > Page Source menu item in most browsers. This launches a text editor and shows the html code in another window. You can place the windows side-by-side to compare what effect each html command is having on the page shown in the browser window.

A similar option is to Save As and choose the source option. This saves an html text file on your disk that you can examine later. You can also do an Open File to have the browser format the text in a local mode. So you can do your comparisons while you are disconnected from the net.

The main item to remember is that all the fancy web formatting is being done in plain text. Plain text is a universal way of passing information between different computers and is the basis of email. It was chosen so all computers connected to the net would have access to information on the web.


Prev | Next | The 'Old' Way to Make Your Own Home Page


'The WebMaster Speaks....' 2-1

Webmaster Lectures
Welcome to a series of lectures on the World Wide Web
Home | Lectures overview | Lecture 1 | Lecture 2 | Lecture 3 | Lecture 4 | Lecture 5 | Lecture 6 | Lecture 7 | Lecture 8



Membership

Login





Last update: Monday, July 10, 2000 at 9:42:14 PM.
Go to John Koch's home page or send him a message.