Terminal Character Set Terminology and Mechanics

A terminal might be capable of displaying multiple character sets. These are chosen by the host application using mechanisms specified in ISO Standard 2022, which are summarized here.

A 7-bit character is one in which the 8th (high-order) bit is zero (0). An 8-bit character's 8th bit is one (1). A 7-bit character set has 128 7-bit characters. An "8-bit character set" is, in fact, composed of two 7-bit character sets, one of them, the "left half," usually being ASCII (ISO 646 International Reference Version), while the "right half" contains the "special" characters needed for a particular language, region, or writing system.

Standard-format character (such as ASCII and the ISO Latin Alphabets), are, in turn, divided into Control and Graphic sections, as specified in ISO Standard 4873. Nonstandard character sets do not necessarily observe this convention, the classic example being PC code pages. The following remarks apply only to standard character sets.

From the terminal's character-set repertoire, six sets are available at once: C0, C1, G0, G1, G2, and G3. C0 and C1 each contain 32 control characters; G0-G3 are graphic character sets of either 94 or 96 characters each. Specific graphic sets from the terminal's repertoire can be designated to G0-G3 by escape sequences from the host or by the Kermit 95 command:

  SET TERMINAL REMOTE-CHARACTER-SET name { G0, G1, G2, G3 }

The terminal's active character set is composed of C0 and C1 controls, plus GL (graphics left) and GR (graphics right). GL tells which of the G0-G3 sets is used if a 7-bit graphic character arrives; GR tells which set is used if an 8-bit graphic arrives. Similarly, the C0 set is used when a 7-bit control character arrives, and C1 is used when an 8-bit control character arrives.

Kermit 95's startup configuration for most ISO-2022 compliant terminals (such as VT220/320) is: ASCII controls designated to C0, ASCII graphics to G0, ISO 6429 controls to C1, and the 96 characters of ISO 8859-1 Latin Alphabet 1 to G1, G2, and G3. GL refers to G0 and GR refers to G2. This configuration can be changed with the SET TERMINAL REMOTE-CHARACTER-SET command or by character-set designation and invocation escape sequences from the host; at any moment, the current layout is revealed with the Kermit 95 command:

  SHOW CHARACTER-SETS

A common use for character-set switching is for displaying a combination of ASCII characters, accented or non-Roman letters, and line- and box-drawing and/or math symbol characters on the screen simultaneously. For example, in the DEC VT220/320, the host application might switch among ASCII, DEC Multinational (for accented Roman letters), DEC Special (for line- and box-drawing characters), and DEC Technical (for math symbols).

Eight-bit character sets can be used in the 7-bit communications environment if the host computer supports shifting. In the most common application of this technique, the Shift-Out (SO, Ctrl-N) control character is sent before any GR characters, then 8-bit characters are sent with their 8th bit removed, and then Shift-In (SI, Ctrl-O) is sent to change back to GL characters. In other words, SO invokes G1 into GL, SI invokes G0 back to GL. Kermit 95 obeys these shifting directives when they come in, a