At 15:53 -0500 13/12/07, Jeff McAdams wrote:
Charles wrote:
It's a standard because it has a public spec and because an
organization issues those spec.
In my experience, an organization (non-profit or not) can't simply
publish their own specification and claim, "hey, this is a standard".
That would certainly be easier.
Sure it can. Its not like there's some magic voodoo that has to happen
in order for someone to be able to write a document up that specifies
how something happens. (that's obviously a way-oversimplified version
of what a standard is, but you get my drift, here).
Actually, I think you are debasing a useful and well-defined term, in
exactly the same way that large companies try to, and I resist it on
both counts. We have all seen large companies claim that what they
do is "standard", "because it's used in over X% of <something>".
They are wrong, and with respect, you are also, and you are playing
the same game.
I have every respect for open-source and collaborative efforts, but
they are not formal standards. Heck, they don't even have the
"implementation agreement" or "specification" status of documents
published by trade associations like ISMA or OMA.
Standards organizations have a formal status and a wealth of rules
etc. over such questions as "who can propose and contribute? how are
decisions made? what are the IPR rules and disclosures? how do you
claim conformance? how is the specification made available, and in
what languages?" and so on. If we get further down the road with the
open-source codecs, I expect that one of the questions that will come
up will be "should these be stabilized by having them published as
specifications by a recognized organization?" (like the W3C),
precisely to settle questions like these.
The ITU, for example, has international treaty status; ISO is also
formally recognized, and there are various national or international
standards organizations (e.g. INCITS in the USA, BSI in the UK).
Properly, documents issued by bodies that are not formally recognized
as standards bodies are not standards, but specifications,
implementation agreements, etc. I know this is not a hard-and-fast
distinction, but it is a useful one. Let's not blur it.
At 16:22 -0500 13/12/07, Manuel Amador (Rudd-O) wrote:
Just because AVC is not a standard doesn't mean it's not proprietary. We call
proprietary anything that can't be implemented by the whole world without
third-party permission. If it serves the interest of the discussion better,
let's eschew the world proprietary and use the words "non-free" or "unfree",
both of which describe AVC well.
The word you may be looking for is "encumbered", I think. (Though
non-free is also true). And many open-source efforts are also
encumbered, but various permissions are granted. Sometimes those
permissions are even conditional, by the way.
Proprietary is usually reserved to technologies that are *controlled*
by a single company (e.g. Dolby AC-3 is controlled by Dolby).
Questions of implementability and use are different again. AFAIK the
patent owners in H.264/AVC don't worry about implementations, they
worry about sale and use. In fact, I rather suspect that they like
more implementations to exist :-).
--
David Singer
Apple/QuickTime
