The Old Joel on Software Forum: Part 4 (of 5)

What is a "standard"?

Interesting CNET interview with Robert Glushko on "standards"

"Q: Why have so many standards emerged for electronic commerce?
A: One of the issues here is what a standard is. That is one of the most abused words in the language and people like you (in the media) do not help by calling things standard that are not standards. Very few things are really standard. Standards come out of standards organizations, and there are very few of those in the world.

There is ANSI (American National Standards Institute), there is ISO (International Organization for Standardization), the United Nations. Things like OASIS and the W3C (World Wide Web Consortium) and WS-I (Web Services Interoperability Organization) are not standards organizations. They create specifications that occasionally have some amount of consensus. But it is the marketing term to call things standard these days. "

Full interview at
http://news.com.com/2008-1013-5200672.html?tag=nefd.acpro
by way of Dare Obasanjo
http://www.25hoursaday.com/weblog/PermaLink.aspx?guid=72515234-1f57-475d-8374-ad1f0965dde9
who adds some comments of his own

"Every once in a while someone at Microsoft asks me “Is RSS a standard?“ and I always ask “What Does That Mean?“ because as shown by the articles linked above it is an ambiguous question. People ask the question for various reasons; they want to know about the quality of the specification, the openness of the process for modifying or extending the specification, where to seek clarifications, or whether the technology is controlled by a single vendor. All of these are valid questions but few [if any] of them are answered by the question “Is <technology foo> a standard“. "

Just me (Sir to you)
Friday, April 30, 2004

As they say: The wonderful thing about standards is that there are so many to choose from. :)

sgf
Friday, April 30, 2004

J2EE is often called a standard, when it isn't. It's a series of specifications from Sun Microsystems that lots of vendors and open-source projects happen to support.

John Topley (www.johntopley.com)
Friday, April 30, 2004

There are committee standards and de facto standards and never the twain shall meet.

Alyosha`
Friday, April 30, 2004

I like the 2nd article:
Standard means the specification contains conformance requirements.

Without conformance (and actual testing against conformance) a standard is just a specification (according to the 2nd article author). A very valid point that I think I will use in the future.

hoser
Friday, April 30, 2004