Technical Note: Web Standards
There seems to be a good deal of confusion these days regarding
standards: what they are, who makes (or has the right to make) them,
how they come about, or even why they are needed. When one speaks
of current web standards and practices, what does that mean?
This document attempts to address that question.
So what are standards?
Standards are a formally-agreed-upon set of rules for how
things work. Standards of many types are critical for
everyday life: standards are the reason you can buy a nut at
one hardware store and be sure it will fit a bolt you got
somewhere else, or what tells you that the plug from your
toaster will fit into the wall, and that the current it
receives won't cause it to explode. Standards are more
important in highly-technical and technology-dependent areas
than in any other type of endeavor.
There are two main types of standards: open and proprietary. The
difference is in who benefits the most from the standard. (Any
standard benefits everyone to some extent; it is an improvement
over chaos and uncertainty.) An open standard is one whose creation
is a process in which any interested party may participate: the
resulting standard benefits the greatest portion of the people
who will use it and whose lives are affected by it. A proprietary
standard, in contrast, is one which is designed specifically to
provide advantages to the group which created and which owns the
standard.
Who makes them?
The Internet, of which the Web is a part, exists and flourishes
because of its reliance on open standards. The Internet is an
international concern, not owned or controlled by any entity,
either national or corporate. It is a self-governing system, which
creates its own regulatory entities as needed: the membership of
such entities is drawn from the people and organizations who use
and provide services to the Internet, and the entities exist to
serve the needs of the entire Internet community, rather than just
one company or group.
The current structure has evolved over time, as has the Internet
itself. New Internet Standards originate as
Requests For Comments
documents (RFCs). Standards-track RFCs are developed by the
Internet Engineering Task Force
(IETF). The proposed standards are then considered by the
Internet Engineering
Steering Group (IESG), with appeal to the
Internet Architecture Board
(IAB). During the review process, Proposed Standards mature
into Draft Standards, and then eventually into Internet Standards,
which are promulgated by the
Internet Society (ISOC) as
international standards. A complete description of the standards
process is available as
RFC 2026.
A complete list of the current standards and their status is
available as
STD 1,
the current version of which is RFC 2000.
Other organizations involved in the standards process include
the Internet Assigned Numbers
Authority (IANA), and the
Internet Network Information
Center (InterNIC).
Internet standards define the infrastructure upon which the Web
is built. At the application level, the organization which is
responsible for initiating and overseeing the development of
web standards is the
World Wide Web Consortium (W3C).
Recommendations from the W3C are passed along to the IETF, but
due to the rapidly changing nature of web technology generally,
most W3C recommendations are considered to constitute best
current practice, even before they are formally accepted as
standards.
So what is the standard for the Web?
Actually, there are quite a number of standards that are relevant
to the web. Grafman Productions has provided an excellent
list of web-relevant specifications and standards
which should serve as a starting point. The
InterNIC
Archive is also quite useful.
The most important part of this question is "What is the
HTML standard?" The current HTML standard as of this writing
(21 May 1997) is
HTML 2.0.
This is officially still a
"Proposed Standard", and has not been given a standard
number: the specification document is RFC 1866. The best definition
of current practice is
HTML 3.2,
which is officially a W3C
Recommendation; it has not yet been issued as an RFC.
Many people are frustrated by the slowness of the standards
process, and advocate simply using whatever HTML the "most
popular" browser supports, and defining the behavior of
that browser as a "de facto" standard. The fact that
such usage is neither universally-applicable nor in any real
sense agreed-upon means that this practice defeats the very
purpose of having standards in the first place. The additional
fact that such a "pseudo-standard" is by definition a
proprietary, rather than an open standard, means that this practice
also places you at the mercy of whichever browser-maker you
choose as your "standard," and makes you wholly dependent
upon whatever whim their commercial interests dictate. For any
independent organization (such as Lehigh) that intends to maintain
a long-term web presence, such a position is clearly untenable.
|
Table of Contents:
