Well, the source is L'Equipe, so we shall see where this goes, but here's what we know today:
CN does a clip job on today's report by L'Equipe that seven urine samples extracted from Floyd Landis during the 2006 Tour have shown traces of exogenous testosterone. These are the B samples of tests whose A-samples came back negative. It's not clear whether "traces" equals a positive result; more likely, the samples could be considered negative tests but still be used in USADA's case against Landis as circumstantial evidence regarding his Stage 17 sample.
As for other sources, "French anti-doping agency president Pierre Bordry told Agence France Presse that although the analyses had been carried out, the results were not yet known." So it remains to be seen if L'Equipe is jumping the gun, or getting a real scoop.
Landis' response is covered well at VeloNews. Landis called the press today to complain that the leaks prejudice his case and demonstrate that the Lab can't be trusted. He is also contesting the results because they were done without a Landis Team rep present, something they'd been arguing for since the B-sample testing matter came up.
Read Landis' whole response on his website.
IMHO we are on the verge of some damning information, so I don't want to react until we have more reliable verification as to what the lab actually found. But it doesn't look good. L'Equipe, for all their conspiratorial behavior, hasn't announced wrong test results IIRC.