On Mon, Dec 20, 2010 at 11:01 AM, Michael Pfeiffer <michael.w.pfeiffer@xxxxxxxxx> wrote: > > Am 20.12.2010 um 19:08 schrieb Urias McCullough: > >> Run 7 has the new results - 10506 new "potential defects" derived from >> many new types of "checker" filters. > > The results would be more useful if they didn't contain so many false > positives. > For example it complains when you allocate in a single locale variable more > than 1024 bytes, e.g. CID 8615, 8520, 8519, ... False positives are indeed an annoyance, but there doesn't seem to be any way to configure the software up front. When I run the software, the results are stored in some kind of database which I then have to upload to the server for final processing. Fortunately, once you've marked one as false - the next time the code is scanned, it shouldn't create a new issue for a previously found issue. You can also bulk-update a set of issues to a single resolution if necessary. Thanks for looking :) - Urias