- From: derek lane <derekglane@yahoo.com>
- Date: Sun, 7 Oct 2001 12:34:31 -0700 (PDT)
- To: www-patentpolicy-comment@w3.org
The PPG proposal might well be clarified by the following scenarios. Perhaps their use might clarify the alternatives. 1. GIF submarine. W3C ignores patents. The W3C promulgates a standard which by some stretch of legal imagination will require infringement of an existing or future patent. Much open source/littleCo work on the patented standard stops soon after the first successful "cease and desist letter" appears. An alternate standard appears but is slowly implemented because of de facto market share built by widespread adoption by open source/littlecos. Discussion. This seems the status quo. Unisys has asserted patent rights to various implementations of the GIF (non W3C de facto standard). It has convinced at least Boutell to make his drawing utilities non-gif compliant, with follow-on effects in various dependent projects. An alternate implementation (libungif) appears still to be in use although I think Unisys could send a similar nastygram to libungif's author(s). PNG which is a better technical solution was created by the Internet community and became a W3C standard but has not been usable until very recently in widely used browsers. Probably W3C spends at least some time avoiding famous patents (Chris Lilley was intimately involved with PNG), and has (Intermind on P3P) actively worked to kill submarine patents. The work that went into gzip's work-around for the LZW patent hints at the cost and complexity for a successful patent work-around. Note that the gzip research, which was successful, was apparently done by *one person*. 2. Disclosure required, with flexibility. W3C goes thru some legal machinery to require members to disclose patents required to implement one of its standards. There are more clearly defined categories (RF) and less (RAND) of license possibilities. A standard comes with a collection of statements from w3C members about how they will license patents. Some groups remain chartered as RF thru promulgation, others start or become RAND, and some patent claims exist at promulgation. Discussion. This is roughly the current PPG proposal. After the MP3 fiasco, Open source/littleCo developers run away from anything they imagine has patent encumbrance. The "RAND" license on MP3 famously describes just how bad RAND licenses can be. W3C can improve this by requiring members to describe their licensing terms at promulgation (Chris Lilley's proposal). License terms like "free for non-commercial use" run into the same issues (aggreration, third-party distributor liability) that lead to Linux' GNU licensing. LittleCo's run away if the cost is higher than they can extract from their users. The ITU experience from chuck weinberger suggests that RAND has potentially very bad consequences for the standards making process. 3. RF licensing only. W3C attempts to find existing or future patents and avoid them. W3C only promulgates RF standards. RF is defined to handle the requirements of GNU and other Open Source licenses. Discussion. W3C as it moves into areas with existing patent holders wi(VoiceXML, streaming media) runs the risk of defining unimplemented standards since the proprietary ones are better. Open source/littleCo's will follow their existing policy of implementing W3C based on technical merit or their interest. Small patent-related fud. W3C's standards have been ignored in the past (browser wars/DOM) and can take some time to be implemented (XML,SVG) or implemented well (CSS, DOM, CSS) but have generally improved interoperability. Some percentage of web users do not care about interop. The web splits into non-W3C standards with possibly company-controlled licenses and broadly implemented RF standards from W3C and others. Proprietary interests with their own standards are less likely to participate in W3C standards groups since doing so may damage their business plan. Experience from PDF and Flash has shown so far that large proprietary standards often broadly license their technology in the interest of ubiquity. Lacking uniform RF licensing from a known enitity, they suffer from a level of FUD and typically do not become part of the code for mainstream browsers. Existing non-W3C solutions (Flash, PDF,MP3 or 4) have no W3C sanction but may become de facto standards anyway. Their probability of being widely implemented gets determined by fit to various interests and published business model. To make clear my own stance, I prefer aggressive RF enforcement on W3C membership (dr_lenz) and at least extra clarity for RAND (Lilley's proposal) with careful drafting of RF so as to be compatible with common Open Source licenses. --Derek Lane Just some guy Marketdriven __________________________________________________ Do You Yahoo!? NEW from Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month. http://geocities.yahoo.com/ps/info1
Received on Sunday, 7 October 2001 15:34:33 UTC