15-JUN-90 NMODEM is offered as an alternative to protocols that think huge blocks of data and bells/whistles is the way to go. My feeling is that the file transfer must be as fast as possible! I do NOT profess to replace ZMODEM with this driver in difficult conditions. However in good to medium conditions tests indicate that NMODEM will save you time and telephone bills. If you are trying to transfer huge files via a difficult connection then ZMODEM should be your choice (mine as well!). ZMODEM is a master at recovering from and handling these situations. I also feel the protocol driver should give the user information about the transfer that is pretty accurate. Large blocks look good but the penalty for an error is huge. My tests have indicated a 2048 Byte Block (2k) is where the efficiency limit is reached. A hit only requires resending 2k NOT 8k! NMODEM does NOT include any "undocumented" requirements as do some other protocol drivers that I have used. I have tried to let you know what it needs to function as far as I know. I was frankly surprised at how well the early BETA versions of NMODEM did in terms of speed. But undaunted I said "Well that means I don't have to play assembly tricks to get this thing up to speed." Now that I have found the flaw in the original CPS calc scheme, I may have to goto a small Inline routine for the send and receive byte collection loops. I intend to service and improve NMODEM to make it a valid and usefull protocol driver. All I ask is that if it saves you time and/or $ then let me know by registering! If NOT let me know as well! Shareware is great but can break you and me if the fees are out of line with the average users return for the use of the program. "cheapware" is my attempt at an answer to the problem. The question is how much does $5.00 or $10.00 really compare to the return of the use of this program? If it saves you 1 hour of time in one year it must be worth $5.00. If it saves you 2 hours in 1 year then it must be worth $10.00.