The guidelines do say” the concepts and processes can be readily applied to other media types like photos, graphics, video, and audio content” but it is not quite so easy as that.

What could an excerpt of a photograph be? A crop of a certain percent? Downrezzing to a thumbnail, eg. by taking every 100 pixels and averaging them as is done in making jpegs?

The difficulty is that even a thumbnail of an image at 10% conveys the gestalt of the entire image in a qualitatively different way than the extracting of 10% of the words in a story.

One essential point has to be providing accurate attribution and preserving all metadata embedded in the image as IPTC, XMP or EXIF data. This includes caption, keywords and contact information for the copyright owner.

The guidelines do say” the concepts and processes can be readily applied to other media types like photos, graphics, video, and audio content” but it is not quite so easy as that.

What could an excerpt of a photograph be? A crop of a certain percent? Downrezzing to a thumbnail, eg. by taking every 100 pixels and averaging them as is done in making jpegs?

The difficulty is that even a thumbnail of an image at 10% conveys the gestalt of the entire image in a qualitatively different way than the extracting of 10% of the words in a story.

One essential point has to be providing accurate attribution and preserving all metadata embedded in the image as IPTC, XMP or EXIF data. This includes caption, keywords and contact information for the copyright owner.

I wanted to respond on the points you brought up:

1) Full Copy Definition: We intentionally aimed high with the 125 word Full Copy definition to avoid the Fair Use debate. That said, we understand that disregarding any articles or posts under 125 words eliminates many works from being considered. There was some 'science' behind this number as we analyzed the works submitted through fairshare.cc; however, the data is old and needs to be refreshed. We'll do this prior to our next update.

2) Ad Network and Search Engine focus: One of the core concepts we are introducing is a graduated response in which content is only taken down in systemic cases; instead, initial actions are focused on monetary impairment: Traffic and Ad Revenue. Search Engine takedowns are designed to deliver the former with ad network takedowns delivering the ad revenue impairment. We're certainly interested in learning about your results, but our search engine and takedown effectiveness rates have been very high.

3) Focus on individual cases: This is a miss by us as our intention was indeed to focus on recurring syndication. Our expectation is that content creators will set acceptable thresholds for each type of reuse (e.g. no more than 5 full copies per month, no more than 15 unlinked excerpts per month, no more than 25 unlinked headlines per month), providing a more scalable way to manage content syndication. In the next version of the guidelines, we will make sure this is more clearly stated.

Very interested in what others think and please send in feedback to guidelines at fairsyndication.org or to me directly rich at fairshare.cc. Either way, the entire company will read it!