HTML and CSS Reference
There are only a few cases when it is acceptable to discontinue a web address, such as if the company or
organization has been closed. Unless there is a really good reason to do so, we addresses should not be changed.
Poorly designed URIs are responsible for a large share of dead links on the Web. While challenging, URIs can be
designed in a way that there will be no need to change them for the years to come.
URIs can be designed to be persistent by minimizing the information provided in them . The author of an updated
document can be different from that of the original one, and thus the author should not be included. The subject
should also be eliminated since it changes very fast. For example, a web technology blog should apply the directory
name markup instead of the names of exact technologies that are currently the most up-to-date one ( HTML5 ). Directory
names that indicate the status of documents such as draft or latest should not be used in URIs simply because
document status changes over time. A persistent URI is required for the latest version of each document.
Some parts of a web site might be restricted to members only. The access should also be eliminated from URIs
because documents might be moved from the private section to the public one, or vice versa.
The most frequently provided needless information in URIs is the file extension. Technologies and tools that are
currently considered as the most advanced will probably change in the near future, or the developer might change
the applied technologies. Changed URIs can affect not only the findability of web pages or web page components but
also your maintenance tasks. Carefully selected directory names are logical and seldom should be changed in the
references in the markup.
■ use the name script or scripts for the directory where you store the script files of a web site instead of php or
language rather than a common name. Similarly, the directory name style or styles is more practical than css or xsl ,
and the name news or feed is more fortunate than rss or atom . the name images is better than jpg , and a doc or docs
directory can hold a variety of documents from pdf to Word documents, not just a certain type.
Directory names that indicate software mechanisms such as cgi-bin should also be eliminated from URIs.
They might change. Topic names, company sections, access levels, or security levels are also inappropriate for URIs.
Classifications can also change. The creation date is constant so it can be included in web addresses, however, dates
are not the best strings for SEO. Keep in mind that multiple web servers can be hidden behind an apparent server with
proxying and redirection.
Beyond providing descriptive names for users and machines, URIs should be simple, stable, and manageable.
Properly designed URIs are fundamental parts of the Semantic Web .
Web resources can be located by unique IP addresses. However, they are hard to remember. Consequently, domain
names are used instead in most cases. Figure 4-1 shows the relationship between a domain name and a URL;
www.example.com is a subdomain of the node example.com , which is the subdomain of the com domain (stands for
commercial). The domain name syntax rules are defined by RFC 1035 , RFC 1123 , and RFC 2181 .