Hello, In compliance with the large volume of source code requests DataET Research has been receiving, the undocumented Pascal source code of UDCM (containing the IMDMP encryption algorithm) will be released in approximately one month. Also, the shareware version of UDCM completely complies with ITAR since only 56-bit keys are allowed. The registered version allows 2048-bit keys. The registered version can only be sold within the U.S. and Canada. Also: IN RESPONSE TO THE FLAME MAIL DATAET RESEARCH HAS RECEIVED: A "comment:" is from a flame mail. A "reply:" is DataET Research's reply to the comment. Comment: "What do you mean by 'extensive research'." Reply: Benchmark tests. Source code comparisons. Documentations of other algorithms. "Applied Cryptography". Comment: "Yah, okay, since when is a DLL not royalty-free?!" Reply: Read some component catalogs. Go search the Internet. Comment: "That stupid line about 140 function bloats your ad." Reply: Is this constructive criticism or what? Comment: "What the hell does DLL size have to do with cost-effectiveness?" Reply: Upload time. Disk creation time. Minimal archive size increase. Comment: "I've never even heard of IMDMP before. Is this a joke?" Reply: UDCM (IMDMP) came out January 1st, 1997. How do you think successful companies start anyway, with millions of dollars upfront? DataET Research is a relatively new company. Comment: "What kind of industry standard is IMDMP above?" Reply: 128-bit encryption. Average DES is 128. PGP tops 1024. IDEA goes at 128. RSA the same. Full security IMDMP is 2048-bit. Any other questions? Comment: "Any bonehead can tell that 256 bytes is 2048 bits." Reply: Does anybody out there feel insulted? Comment: "Why does IMDMP need sub-algorithms? Is there something wrong?" Reply: Sub-algorithms of IMDMP are basically algorithm modes. There is nothing wrong with IMDMP's structure. Why don't you see what sub-algorithms are for yourself? Go try out UDCM: http://members.aol.com/dataetrsch/udcm.html. Comment: "What algorithm doesn't process both binary and ASCII?" Reply: Go search the Internet for one. You'll find many dinky ones. Comment: "What algorithm doesn't allow multiple encryption layers?" Reply: Go search the Internet for one. You'll find many dinky ones. Comment: "Who says IMDMP doesn't have any back-doors or magical keys?" Reply: What more do you want us to prove? The source code will be out in about a month. Comment: "Why is there a dumb 2,147,483,390 file size limit?" Reply: It has do with 2^32 DOS file pointer long integer variables. You developers should know that. Comment: "Why the hell is the file size limit 258 bytes off? What, are keys kept in file headers?" Reply: Keys are NOT stored in headers. The 258 (actually 256) byte header is used to store certain IMDMP special security option settings. The header is itself secured using a 128-bit encryption code. Comment: "What the do you mean by 'file specific unique encryption'." Reply: For the idiot who sent this message to us: Why don't you download UDCM V2.0 and view its documentation for yourself? Comment: "Isn't IMDMP illegal?!" Reply: The version of UDCM (IMDMP) being distributed on the Internet, the shareware version, allows only 56-bit keys. The registered version allows 2048-bit ones. The registered version is only sold within the U.S. and Canada. Comment: "I want the damn source code to IMDMP now. How do we know for sure that IMDMP is better than DES and all the others you mentioned?" Reply: Wait one month please. Jeremy K. Yu-Ramos President DataET Research Data Engineering Technologies