Data format

While transferring data over the network, several data representations can be used. The two most common transfer modes are:

1. ASCII mode
2. Binary mode: In "Binary mode", the sending machine sends each file byte for byte and as such the recipient stores the bytestream as it receives it. (The FTP standard calls this "IMAGE" or "I" mode)

In "ASCII mode", any form of data that is not plain text will be corrupted. When a file is sent using an ASCII-type transfer, the individual letters, numbers, and characters are sent using their ASCII character codes. The receiving machine saves these in a text file in the appropriate format (for example, a Unix machine saves it in a Unix format, a Windows machine saves it in a Windows format). Hence if an ASCII transfer is used it can be assumed plain text is sent, which is stored by the receiving computer in its own format. Translating between text formats might entail substituting the end of line and end of file characters used on the source platform with those on the destination platform, e.g. a Windows machine receiving a file from a Unix machine will replace the line feeds with carriage return-line feed pairs. It might also involve translating characters; for example, when transferring from an IBM mainframe to a system using ASCII, EBCDIC characters used on the mainframe will be translated to their ASCII equivalents, and when transferring from the system using ASCII to the mainframe, ASCII characters will be translated to their EBCDIC equivalents.

By default, most FTP clients use ASCII mode. Some clients try to determine the required transfer-mode by inspecting the file's name or contents, or by determining whether the server is running an operating system with the same text file format.

The FTP specifications also list the following transfer modes:

1. EBCDIC mode - this transfers bytes, except they are encoded in EBCDIC rather than ASCII. Thus, for example, the ASCII mode server
2. Local mode - this is designed for use with systems that are word-oriented rather than byte-oriented. For example mode "L 36" can be used to transfer binary data between two 36-bit machines. In L mode, the words are packed into bytes rather than being padded. Given the predominance of byte-oriented hardware nowadays, this mode is rarely used. However, some FTP servers accept "L 8" as being equivalent to "I".

In practice, these additional transfer modes are rarely used. They are however still used by some legacy mainframe systems.

The text (ASCII/EBCDIC) modes can also be qualified with the type of carriage control used (e.g. TELNET NVT carriage control, ASA carriage control), although that is rarely used nowadays.

Note that the terminology "mode" is technically incorrect, although commonly used by FTP clients. "MODE" in RFC 959 refers to the format of the protocol data stream (STREAM, BLOCK or COMPRESSED), as opposed to the format of the underlying file. What is commonly called "mode" is actually the "TYPE", which specifies the format of the file rather than the data stream. FTP also supports specification of the file structure ("STRU"), which can be either FILE (stream-oriented files), RECORD (record-oriented files) or PAGE (special type designed for use with TENEX). PAGE STRU is not really useful for non-TENEX systems, and RFC1123 section 4.1.2.3 recommends that it not be implemented.