Programming Certain Complex Types of Typesetting

V. M. Lyukhanov, B. A. Ponyshev, V. E. Domashchenko

When typesetting certain complex kinds of text, especially scientific and technical information, within a single line it is necessary to take into account the widths of characters of various typefaces and the graphical bases of typographic type, and in phototypesetting also of various point sizes.

If \(G\) is the number of point-size/typeface/style variants on different graphical bases, then the relations determining the line length \(L_1\) with the minimum \((\Delta_1)\) and \(L_2\) with the maximum \((\Delta_2)\) value of the variable interword space take the form:

\[ L_1=\sum_{j=1}^{G}\sum_{i=1}^{n_j}\delta_i^j+\sum_{j=1}^{G} k_j \Delta_{1j} \]

\[ L_2=\sum_{j=1}^{G}\sum_{i=1}^{n_j}\delta_i^j+\sum_{j=1}^{G} k_j \Delta_{2j}, \tag{1} \]

where \(\delta_i^j\) is the width of the \(i\)-th character of the \(j\)-th point-size/typeface/base;

\(k\) is the number of variable interword spaces in the line;

\(n\) is the number of characters of the \(j\)-th point-size/typeface/base.

In the case when erroneous characters in the amount \(p_j\) and erroneous spaces in the amount \(l_j\) for the \(j\)-th point-size/typeface/base have been introduced into the setting, relation (1), taking into account correction of the widths of these characters and spaces, takes the form:

\[ L_1=\sum_{j=1}^{G}\sum_{i=1}^{n_j}\delta_i^j-\sum_{j=1}^{G}\sum_{i=1}^{l_j}\delta_i^j+\sum_{j=1}^{G} k_j \Delta_{1j}-\sum_{j=1}^{G} p_j \Delta_{1j} \]

\[ L_2=\sum_{j=1}^{G}\sum_{i=1}^{n_j}\delta_i^j-\sum_{j=1}^{G}\sum_{i=1}^{l_j}\delta_i^j+\sum_{j=1}^{G} k_j \Delta_{2j}-\sum_{j=1}^{G} p_j \Delta_{2j}. \tag{2} \]

In the typesetting process, a typographic line passes through three qualitative states, which are determined by the relations between \(L_1\), \(L_2\), and the value of the typesetting format \(F\).

Fig. 1.

Block-diagram labels:

• Input device
• Control unit
• IBM-72 typewriter
• Reprogramming device
• Control device
• Conversion field
• Expansion block for the line
• Hyphenation and line-fill block

For \(L_1(n_j, k_j, l_j, p_j, G) > F\), the line is considered tight;
for \(L_2(n_j, k_j, l_j, p_j, G) < F\), it is considered loose. In those cases when

\[ \begin{aligned} L_1(n_j, k_j, l_j, p_j, G) &< F, \quad \text{and} \\ L_2(n_j, k_j, l_j, p_j, G) &> F, \end{aligned} \]

the line can be “justified,” i.e., brought to format by varying the width of the variable spaces within a specified tolerance field corresponding to the approved standards.

For programming complex types of typesetting whose line length is governed by relation (2), an algorithm and a device implementing it were developed.

The block diagram of the device is shown in Fig. 1.

The use of the IBM-72 typewriter as an input device in typesetting mode and as an output device in printing mode makes it possible to obtain a typewritten original layout practically in any

faces, owing to the presence of replaceable type heads, i.e., it is possible to program texts of the form:

$W. Smith$ asserts that $F(x)=Aexp(-\rho/\gamma\delta)$

or

“INPOLIGRAFMASH-69” “inpoligrafmash-69” “INPOLIGRAFMASH-69”

The device under consideration can operate in the following modes:

1) setting mode;
2) printing mode;
3) semiautomatic correction mode;
4) reperforation mode.

When operating in the setting mode, the device implements relations (1) and (2). At the same time, signaling is carried out for the process of filling the line, on the basis of which the operator decides on the optimal completion of the line.

As a result of setting, we obtain a typewritten original layout, fully coinciding with the subsequent typographic proof, and a punched tape—the control program for line-casting or phototypesetting machines.

Fig. 2.

Studies carried out by the authors at the printing house of the PIK VINITI showed that the percentage of detection of an erroneously printed character as a function of the number of characters printed after it obeys the dependence presented in Fig. 2. Analysis of this dependence shows that more than 95% of detected errors fall within the first eight characters. Therefore, to implement relation (2), a device is provided that makes it possible to exclude any of the last eight printed characters within a line, since providing the possibility of correcting an erroneously printed character within the entire line would significantly increase the size of the device. Errors are corrected without reversing the perforator, which accelerates the programming process and reduces wear on the mechanical parts of the perforator.

Quantitative estimates of the acceleration of the programming operation and the increase in wear resistance of the perforating mechanisms will be determined upon completion of the studies.

To check the punched tape obtained in the setting mode, the device provides a printing mode, the result of which is a typewritten original layout of the text encoded on the punched tape. The reading of codes from the punched tape is performed by an electromechanical reading device.

The semiautomatic correction mode makes it possible to introduce any changes into the punched tape—the control program for line-casting and photo-

by typesetting machines and, in the process, obtain a corrected typewritten camera-ready copy and a new punched tape.

To obtain several identical punched tapes, operation of the device in reperforation mode is provided. Reperforation mode and printing mode can, when necessary, be combined.

Practical results from testing the line model for \(G = 1, 2, 3, 4\), using data from cast Linotype lines, gave satisfactory agreement. The number of actual lines not corresponding to the model does not exceed \(0.5\%\).