Credit: Samuel Zschokke
The relationship between client and server
is the fundamental paradigm that determines the characteristics
of computer languages on the web. The differences are played out in terms
of:
- How the languages are invoked
- What resources and information they can access
- What structures can be manipulated
- How they interact with the user and system
- The speed and predictability of their interactions
- How the languages are invoked: server side languages live on the
server and are typically invoked either by
- a user pressing a submit button in the client
browser HTML which transfers control to the server program along
with
any name-value pairs in the form associated
with the submit button.
Here is an
simple example of such an HTML program and here is the
simple server program that is executed by it
- a user typing in a URL with the name of the server program
optionally followed by a "?" and "name=value" parameters
separated by
"&"s. For example,
http://site.location.type/my.cgi?p1=abc&p2=10
- a user clicking on a #exec Server Side Include
element
in HTML which points to a server program
In contrast, client side programs are initially stored on the server
embedded in the HTML code but are transferred to the client browser
and operate there for the remainder of the session. They may be
activated just by the loading process, by "onLoad" commands,
or by user actions such as moving or clicking the mouse.
- What resources they can access: server side programs operate as
normal user programs unless web administrators have imposed
restrictions. Thus they can access all user programs, such as
graphics and databases, system programs, Unix system variables,
and a special set of variables associated with the web, including
a limited amount of browser information.
In contrast, client-side programs are almost entirely confined
to the browser, except for accessing images and other client
side scripts from the server, and cookies. However, they have
almost full access to the structure and operation of the browser.
- What structures can be manipulated: server side programs can
manipulate any resources on the server that are accessible to
any user program, and can also write (but not change) HTML
(including embedded scripts), plain text, and images to the
browser.
In contrast, client-side programs can not manipulate anything
on the server (they can read images and scripts but cannot write
or change anything), or the client (except for cookies), except
for the browser, where they have a great deal of control
- How they interact with the user and system: server side programs
cannot manipulate any information on the client browser: they
can rewrite a browser image, but they cannot modify an existing
one (although they can change the contents of graphic files)
In contrast, client programs interact easily with the user, and
in the case of DHTML, can move, modify, add, or replace virtually
anything in the browser
- The speed and predictability of their interactions: server side
programs can only rewrite a browser page, not change anything on it,
so interactions often require a good deal of modem traffic and
are subject to the varying loads placed on the server
In contrast, client programs can manipulate HTML elements down
to the smallest level and are affected only by the activity
on the client. Thus interaction is usally very swift, immediate,
and predictable.
|
|
 |
|
|
