Hi William, If you are selling the images, why not put crummy thumbnails onto the web so that people can't actually view the proper images in the first place? Perhaps you could use a more robust watermark, or place a black X in the middle of each picture?
Also, I'd be suspicious of statistics about "downloading": every web hit is a download, regardless of whether it is called a "hit", a "view" or a "download". Remember, the only way your web browser can ever display an image in the first place is by downloading it! Thus, you cannot grant people the ability to view your image via the web unless you implicitly grant them the ability to store the image in their computer. I'm not sure how copyright law deals with "buffering" (which is a technical description of what's going on), but you basically can't have a digital system that doesn't involve buffering anyway. As Craig mentioned, the only difference between a colloquial "download" and a "hit" is cosmetic: it's whether your web browser saved the picture to the Desktop or whether it saved it to a hidden folder. When someone asks their computer to "download" the image to disk, all the browser has to do is re-save the copy it had already downloaded. It doesn't even have to contact your web server to do this. I also agree with Craig that obfuscating the user interface acts as a deterrent to legitimate customers, or makes them find ways to work around it. For example, I find myself having to advise people how to work problems with images and it is easy in Safari: open the 'Activity' menu, find the image in the list, and double-click it. It then opens in its own window and can be dragged to the Desktop. It's possible for software to be designed to keep all your images in "memory" or "encrypted", so that plain copies are never stored on disk, but you don't really have any control over this unless you write a plug-in that people have to install to use your website. But even then, since your resource is available via the web, people can always bypass the restriction. PS. Perhaps the only way your web server *could* know about "downloads" versus "hits" is if it tries to make a guess (e.g., perhaps if people's web browsers disclose this via the presence or absence of information in Referer headers) but that depends on what web browser is being used, and may be misleading if the person uses products like 'accelerators' or 'proxies'. It may also underestimate the amount of "downloading", if people's browsers don't contact your web server.
