In Pro Choice, Sam Ruby says: I prefer simpler definitions.  A standard is one that has multiple, inter-operable, independent implementations.  An open standard, at least in the software world, is one where at least one of those implementations is open source. You can sign me up for that, too.



Contributions

Comment feed for ongoing:Comments feed

From: Alex (Jan 28 2007, at 18:37)

"An open standard, at least in the software world, is one where at least one of those implementations is open source."

This definition is simple, but wrong. With that definition a standard becomes "open" in the moment when its first open source implementation is finished. I dont think so.

A standard can be open even if there are no implementations available at all. Just as a mathematical statement can be true before its first proof found.

[link]

From: Asbjørn Ulsberg (Jan 29 2007, at 03:11)

I also agree that the definition of "open" is wrong. What I read into an "open" standard is not that any of the specification's implementations are open themself, but that the process around the specification and input to it, its editors et al are open. That the standardisation process is an open process.

[link]

From: Dan Davies Brackett (Jan 29 2007, at 07:27)

Alex's comment speaks to differences in Alex's and Sam's (and Tim's) interpretation of the word "standard". I tend to agree with Sam; the specification document for a standard is just that -- a spec doc, which can be wrong and is almost by definition ambiguous in critical areas. The reference implementation is the standard itself, in the sense that it is the standard to which other implementations of the idea behind the specification are held.

[link]

From: Ben Finney (Jan 29 2007, at 13:54)

I agree that an open standard means the *standards process itself* must be explicitly open to contributions from outside parties.

Here's an article on PDF moving from a closed standard (despite several open implementations) to an open one:

http://www.adobe.com/aboutadobe/pressroom/pressreleases/200701/012907OpenPDFAIIM.html

[link]

From: Joe Pallas (Jan 29 2007, at 14:06)

The problem is that Sam's definitions, however appealing, don't reflect the actual usage of these words. MS Word sorta kinda fits his definition of "standard" but the cold hard truth is that it would satisfy the commonly used meaning of "standard" even if there were no other program on earth that could read a .doc file.

Sun has its own definition of "open standard" (http://www.sun.com/software/standards/definition.xml) that's quite different from Sam's.

I guess there's no standard definition of "open standard."

[link]

From: Sam Ruby (Jan 30 2007, at 03:00)

A mathematical statement for which a proof has not yet been found is said to be unproven. There needs to be an analogous concept for standards which can only be said to be potentially implementable in an inter-operable way by a third party, but that openness is, as of yet, unproven.

PDF has a tremendous market share, and becoming an accredited international standard is certainly a good thing. But from a broad stroke picture, the process of becoming a de facto standard and then approaching a standards organization with the prospect that any change would create a tremendous incompatibility with the deployed base is reminiscent of what other, less open, organizations have tried to do.

At some point, one has to look at results and not just process. PDF has a proven track record of multiple, independent implementations. And in a very real way, Java became more open with the GPLing of Sun's implementation.

[link]

author · Dad · software · colophon · rights
picture of the day
January 28, 2007
· Technology (85 fragments)
· · Software (69 more)

By .

The opinions expressed here
are my own, and no other party
necessarily agrees with them.

A full disclosure of my
professional interests is
on the author page.