Here, I am done - I think this makes sense.
Read this while reviewing the flow chaeplin posted:
The transactions sent from the mixer (C) to the end address is of a set size (his example was 10). This is D.
He can then search the blockchain for possible candidates (value 10) via a script. Which gets us to block 28531 - this matches the 10.00 XC value. This can be linked at blocked 28533 with the mixer C - we already know C (output 9.99999).
This allows him to trace back to 28531 in the blockchain - to find the values that == 10.00 and match to a specific address - in this case: XQdBjeQtH1JGrkd2MWcXbtsRVeKHWZbnqa which is B.
You can then take all the transactions for this address B - review them and find two matching amounts that == 10 which belong to one single address. This address is A. Since this address B has never been used before this transaction - this is easy to do - and - even if it had multiple transactions - they would not all related back to one single point with one single value (10).
Just for the record, - I think this is right.
Again - I'm participating to participate - nothing else.
If my logic is flawed, let me know please, this is vexing me - because this looks right and all I get is nothing constructive back.
Thanks, this is a helpful description of what Chaeplin did.
However it does not adequately express what ATCsecure wanted tested. He wanted *proof* of a direct link, not just interestingly coincidental amounts sent and received. This is because he's testing the mixer and xnode functionality, not the multi-path feature which is yet to be implemented.
For more information, read page 356 of this thread, and also the following quotation:
sending 0.03 to both address's doesn't count as a LINK
but thanks for the XC's
So read page 356 and you will see that XC's implementation is successful and that Chaeplin was trying to test for the wrong thing. He returned later when ATCSECURE was gone and make his case again - yet concealed the fact that what he presents is not what's at stake here.