29 January 2008

Open Media Definitions

Definitions matter. If you want to see why, compare the worlds of open source and open access. The very specific definition of what is open source - having an OSI-approved licence - means that it is relatively easy to police. Open access, by contrast, does not have anything like a tight, "official" definition, with the result that less scrupulous publishers try to pass off their wares as open access if it's vaguely open or vaguely accessible. (That, BTW, seems to me a very strong argument for something to be set up along the lines of the OSI, to give its stamp of approval to open access resources. Are you listening, Peter?)

For this reason, trying to define open media is an excellent move:

With the change in the media climate and distribution experiments such as Radiohead’s In Rainbows (in music), and Four Eyed Monsters (in film) which have open qualities (temporarily available to watch or listen/download for free, for example) but are not truly open content, it is getting harder to tell what you can and cannot officially do with your media.

These are three proposed states for open media, each building on the next:

Open (O-):
The baseline, concerned with freely consuming and sharing the content (1-3)

Open source (O):
Being able to view and remix the source files (1-5).

Open Plus (O+):
The ability to participate in a transparent, documented process (1-8).

* 1. Freely accessible
Available to stream, or download without a fee. Should be available via direct download and P2P media, so it is not behind a gateway.
* 2. Freely available.
Permanently available without DRM, or release windows. The end user able to share the work without restriction.
* 3. Freely viewable
Available in multiple formats, and to be converted freely (in the case of video works, for example, as dvd, xvid/divx, mp4, and HD formats).

The above qualities are essential for open content. Open source content adds to the cultural commons by making creation of new content from the work.

* 4. Giving source files
Source media, such as rushes and raw graphics files should be archived and available for other creators to work with.
* 5. Allowing remixing
Materials should be licensed explicitly to allow derivative work (eg. other works based on the script, or video mashups, and remix edits) for at least non-commercial/artistic purposes. Creative Commons and other licenses are available for flexible copyrighting.

Open Plus adds more opportunities for participation and involvement in the work whether as a creator, or as part of what used to be called ‘the audience’.

* 6. Reveal the process
Allowing access to not only the final source media, but work-in-progress material and software files, adding another layer of transparency and documentation.
* 7. Open contribution
Adding ways to influence and participate in the creation of the original work through various types of community/audience involvement (opportunities such as open crewing, direct feedback or contribution mechanisms).

I think this is really important work: let's hope it can be built on. Open Media Initiative, anyone? (Via P2P Foundation.)


Bill Hooker said...

Open access, by contrast, does not have anything like a tight, "official" definition

Eh? The BBB's are internally compatible and consistent, and together form a very tight and useful definition that is in every sense "official".

You're spot-on about the dangers of dilution of the term, though, especially with bad actors doing it deliberately. I hadn't thought about an OSI-type certification process... not a big fan of top-down methods in general, and particularly in respect of Open Whatever...

Glyn Moody said...

OK, let's take that from the bottom.

We agree that dilution is danger. The question is, how can we stop it? From my experience in the open source world over the last ten years, it's clear to me that the OSI definitions have proved extremely effective against even Microsoft's worst efforts in this respect.

Why? Because there is a page with an extremely clear set of conditions that must be met for code to called open source.

This means that even *journalists* can understand when someone is falsely claiming to be open.

Now, the BBBs may be internally compatible and jolly nice, but no journo is going to slog through them trying to understand whether publisher X is compliant.

If you want to nip this dangerous trend in the bud, as I think OA must if it is to reach the next level, you must provide this crutch to the media.

A formally-constituted OA organisation, recognised by all the main players, with a very precise, easy-to-understand defintion of what OA is would be a huge boon.

I really hope people like you and Peter will think about this, because I do worry that OA is going to start hitting serious problems with this issue.

Peter Suber said...

I welcome the attempt to define openness for video. But I have to take issue with the claim in your opening paragraph that "Open access [for research literature]...does not have anything like a tight, "official" definition, with the result that less scrupulous publishers try to pass off their wares as open access if it's vaguely open or vaguely accessible." The three public definitions from Budapest, Bethesda, and Berlin are well-known and reasonably "tight". (Thanks to Bill for jumping in to make the same point.) It's true that many publishers are claiming to provide OA without conforming to these definitions, but (as I put it in my recent Richard Poynder interview, p. 30), "the problem is not that our definition is unclear or inadequate. The problem is that the term has spread much further and faster than the definition." Open source suffers from the same problem, which is why some people still mix up free as in speech with free as in beer. But having said that, I agree that OA can learn from OS that a very good way to solve this problem is by connecting the definition to one or more precise licenses. However, we don't need a top-down organization to do this, and we are working on a more open or bottom-up way to accomplish the same goal.

Glyn Moody said...

I hesitate to disagree with you, so I'll just offer some comments on your own words.

You say that the “three public definitions” are well-known and reasonably "tight": but there are three of them, so their tightness leaves something to be desired. And as I mentioned in my comment above, the problem is not with people within the OA community, but people outside it. They find it hard to bring three public definitions together to form a coherent, tight whole. That's why I advocate a reframing in terms of something similar to the Open Source Definition: because it makes it easy for those people to understand what is and what is not open access.

And I'm not advocating a huge organisational carapace for OA: simply something visible that can be pointed to as the keeper of the flame, and which can adjudicate in cases of doubt. Maybe the DOAJ/SPARC Europe seal of approval programme will do that: we shall see.

But the virtues of that approach seem to me inarguable, and were shown most clearly when Microsoft – Microsoft! - came cap in hand to the OSI, asking ever so 'umbly that it would approve some of Microsoft's licences as open source. Imagine Elsevier doing the same....

Matt Hanson said...

Thanks for posting, and discussing the 7 rules for open source media blog post I wrote recently. It was designed to start a discussion so I'm glad you have!

The idea was to create a simple ratings system for an end user perspective (rather than a rigid legal framework, CC, FCD, and others already do this).

It still needs refining, and I already think the labels need to be redone, but currently I'll stand with 3 simple 'classes' with their rules.

Please trackback or add a comment at the original post, and I'll try a follow up post in February ;)

Oh, and I think the OSD for software is fundamentally different to how it should be applied to 'open source' content. Its like comparing apples and oranges.

Glyn Moody said...

Why do you say "OSD for software is fundamentally different to how it should be applied to 'open source' content"?