I have some 50 years experience with computers, including writing my own, and testing routines for computerized card dealing. During this time I have come across several card dealing programs where the randomness was indeed unsatisfactory in some, or even many ways. When BIGDEAL was announced I took the liberty to verify also this program. All my tests were successful. However I discovered some disturbing risk for a C+ library failure which (at least in theory) could compromize the program with internal addressing exception or even undetcted failure. This risk was associated with the need to handle integer numbers within the full range [0 - 53644737765488792839237440000] (the number of possible different deals). So I 'translated' BIGDEAL to Delphi where I have my own library routines able to handle integers of essentially unlimited size without any risk of overflow or other failure. My tests confirmed that the outcome from both versions (started with identical parameters) were identical, but I was not able to eliminate the risk for malfunction in the C+ library. I understand that WBF require all deals for tournaments under their responsibility to be created using BIGDEAL, and I fully sustain this decision. It is in my opinion immaterial whether the deals are to be used in F2F bridge or any version of computerized bridge.