Due to the "snippet match" nature of the scans, we found that it was too much effort to properly validate and catalog each open source component with every new project/product. Incremental results were also difficult to achieve even after consulting with the vendor. We found there were too many false positives and the code-snippet validator had bugs and presented too many false positives. My experience with this tool has turned me away from "snippet"-focused composition analysis. We have switched to one that uses more complete code signatures that do not require validation and review of findings in most cases.
