Google never had any problems with indexing sites that use frames (when you know how to work with them), but the last two weeks or so Google is really messing things up. They now show framed pages with their own url, but with the description of the index (frameset) and with the title of the index or a mixture between the frameset title and the h1-text from the framed page.
This is the situation: Google crawls a page, let's say green-widgets.html. This page contains an external javascript that checks whether the frameset is present. If not, the user is redirected to index.html?green-widgets.html. A frame with navigation buttons appears en green-widgets.html is loaded in the main content frame.
Until recent the only thing you had to do to avoid duplicate content was to prevent url's with query-strings to be indexed, because they would all get the title, description etc from index.html. All content pages were indexed with their own url, title, description and content.
But now Google started indexing green-widgets.html like it used to index index.html?green-widgets.html because the description is now that of index.html. Sometimes the title is that of index.html (let's say: widget shop), sometimes it's a mixture of the h1-text of green-widgets.html and the title of index.html (let's say: green widgets - widget shop).
Besides the title all pages now start to look like duplicate content and the number of visitors is dropping fast.
Has anyone else noticed this new behavore of Google and managed to tackle this problem? (Besides not using frames.)