It starts from basic testing procedure (A and B samples, etc) and widens to implications of "innocent until proven guilty", "burden of proof", and due process as it pertains to the anti-doping law and specifically the case of Floyd Landis.
I learned some interesting things from the article:
There were clear violations of UCI procedure, but it appears that there is little official recourse for Floyd to take.
The leaks from the lab to the press may not be stricly against the laboratory code of ethics. Leaking that there "is a positive" without naming who is positive is not prohibited. Nor is leaking results of tests (like the IRMS test) attached to a name if permission is granted by the testing agency and race organizers. Whether permission was granted, I do not know. It still makes the testing organization look biased.
What was not clear from the article is if the testing lab knew they were testing Floyd landis's sample for the IRMS test and the B test. The article claims there is a possibility that this occured. This was a huge issue with the Tyler Hamilton case. Testing a sample non-anonymously is bad. Testing a sample with an expected result (based on a previous test, comments by race organizers, press, rider, etc.) is worse. It is easy to see how the lab could be pressured to test a sample until there was a positive, or interpret a borderline test as a positive in these cases. Tyler Hamilton claims that his samples were initially declared negative, but reinterpreted as positives due to UCI or WADA pressure. The fact that this doubt exists makes the Lab and the testing organizations look unethical and does not instill trust.
Anyhow, read the article yourself, again here, and see what you learn.
As a side note, I seem to remember that there was a paragraph in the article that covered a lesser sentence in exchange for testimony leading to conviction of doping suppliers, but it looks like that sentence is gone. Anyone else remember that? Or am I getting sources confused? Damn psuedo-journalism...
More Doping musings: