Help the Foundries But Reject an EOT Working Group

Thomas Lord (lord@emf.net)
Berkeley, CA — 22 Sept 2008
Copyright © 2008, Thomas Lord

Abstract

In a brief note we try to summarize our position of strong objections to calls to form an EOT working group in relation to an EOT submission to W3C . We also try to summarize the reasons we would very much like to help EOT-backers achieve their goals, even if their proposal is flawed. Finally, we link to a counter-proposal that exemplifies how to support the goals while rejecting the deeply flawed EOT proposal.

In Opposition to the EOT Submission

The EOT submission to W3C in effect proposes a standard which would specify that conforming Web browser software:

  1. Use all or part of the current URL to decide how to render with a particular font file. That is, the semantics of the font would depend, on all things, on the user's current URL.
  2. Do so for the express purpose of discouraging users from violationing copyright, patent, and trademark claims made by certain suppliers of font files.

On its face such a proposal is absurd:

  1. The proposal would change the protocol that currently exists between browser logic and font rendering logic, passing the current URL (a private piece of user data) across this barrier for the first time.
  2. The proposal violates separation of content, presentation, and interaction severely by tying the presentation of a font file to details of interaction.
  3. The proposal addresses a generic concern that applies to many media type representations yet advocates a solution that applies to only one media type representation. Thus, the proposal badly violates the principle of orthogonality
  4. The proposal aims to support legal claims which are controversial and which may not uniformly apply across jurisdictions
  5. In identifying those controversial legal claims as the rationale for details of a W3C Recommendation, the proposal would plausibly impose new forms of legal liability on programmers implementing W3C recommendations.
  6. The Submission invites needless strife in the manner it presents these controversial legal claims and uses them as rationale for features which restrict the user.
  7. The legally controversial, technically absurd details of the proposal have already generated strong objections by prominent browser implementors. Passage of a Recommendation would likely diminish the stature and effectiveness of W3C Recommendations. Formation of a working group would be little more than a public spectacle and distraction from real work (that benefits users).

Simply put, a working group formation around the EOT Submission would divide the W3C community, diminish W3C's stature, and fail to achieve a widely adopted Recommendation. It would be a harmful waste of time.

But Wait: EOT Proponents Have a Sane Goal

We understand the position of most backers of the Submission to be that they want a representation format for fonts on the web that:

  1. Includes meta-data sufficient to carry declarations of copyright, patent, and trademark status.
  2. Informs users of that status especially in situations where the font appears to have been used in unauthorized ways (regardless of whether or not there is a legitimate claim to require such authorization).
  3. Is not part of a DRM scheme that would activate provisions of the United State's "DMCA" law. Presumably they mean is not part of any scheme that would impose new kinds of liability on W3C implementors.
  4. Is perhaps easy enough, technically, to "work around" but that people are unlikely to "work around" accidentally without knowledge of what they are doing.

Those are not unreasonable goals. They describe a form of Digital Rights Expression (DRE) as contrasted with Digital Rights Management (DRM).

Digital Rights Mangementment seeks an arrangement whereby users are not free to even attempt to work around controversial legal claims -- claims enforced by software (or in other circumstances hardware). DRM is a very poor idea for many reasons not least that software can not accurately determine when a use of a font is legal or not.

Digital Rights Expression seeks an arrangement where ordinarilly operating software has features that tend to inform users of a font supplier's legal claims during interactions where those claims may tend to to be relevant to what the user is doing. In other words, DRE is simply a mechanism for communicating meta-data between (in this case) a font supplier and a font consumer.

DRM can prevent a user from using resources whose use she is entitled to, imposing legal liabilities on any effort to help the user or effort by the user to help herself.

DRE, on the other hand, simply informs the user of what the font publisher claims are her rights to use the font.

In this interpretation of it, the EOT Submission has reasonable goals in the abstract and is simply flawed in its actual construction. An alternative Submission should be substituted and we turn next to what that might be:

Anyone Still Remember Hypertext?

DRE is a special case of a general hypertext pattern -- the ability to include with an embedded resource meta-data describing the conditions of its use. What is called for is a uniform mechanism for representing such meta-data and a simple, user-friendly mechanism for presenting it.

We offer such a proposal separately, although admitedly in a less developed form than the current EOT Submission. We hope that readers will contemplate this alternative before agreeing to the current "scorched Earth" EOT Submission.