Milestones in the History of Calculators

---

AER-I First version

The program (or its official name, line or area) can contain up to 9 parts of expressions, which must be separated by commas. Each part of the expression is a separate calculation, in case of evaluation their results are displayed one after another, together with the ANS text and the serial number of the part of the expression. The variable request statement above can be included in any part of the expression.

Basically, all the operations and functions that can be performed with the calculator can be included in the expressions, with which calculations can be performed in the same way as in manual calculation mode. The symbols used are the same as the symbols on the keys. You cannot enter the angle unit change (DRG), display format specification (F↔E or FSE, TAB) and delete instructions (CL, CA) in an expression. Because statistical calculations can be performed in a separate mode, statistical operations and variables cannot be used either.

The five areas in which expressions can be placed are not free to use, they can only be populated one after the other with expressions. If an area with a lower sequence number is deleted, the subsequent areas with a sequence number will take one smaller number. When typing, you can move to the next expression by entering the end-of-expression symbol (◺) at the end of that expression, which will cause the machine to display the next expression.

Expressions are evaluated in COMP mode by pressing the key sequence corresponding to the area number (2ndF 1:, 2:, 3:, 4: or 5:). The sign separating the area number from the expression is still a semicolon here.

When executing each area, after reaching the end of that area, the COMP key restarts the execution from the beginning of that expression so that manually controlled cycles can be created.

AER-I Second version

• 99 areas, up to 160 steps each, and label (name) may be entered for easier identification,
• you can also use variables with longer labels (with letters and numbers), the value is automatically requested when they first occur,
• conditional branches can be written to the expressions, which can contain three types of conditions and „true” and „false” branching parts.
• we can organize cycles that can even be nested,
• a maximum of 10 subroutines per area can be used to efficiently manage recurring parts,
• you can disable the display of irrelevant results of the expressions.

Mainly for compatibility reasons, the previous variable request statement (f()=) has been retained, to which we cannot directly link conditional branching. In this case, a multistatement delimiter (_) must be written after the closing )= instruction part.

In addition to multistatement delimiter (comma), a new delimiter is appeared (space or underline), which does not display the result of the expression part before it, and the further evaluation of the expression continues. Thus, we can prevent the display of sub-results that are not interesting from the user side, thus writing easier-to-use, simpler programs.

The handling of conditional branches and loops is different from the usual label jumping technique: after evaluating a given condition, the true part will be executed if the condition is true, and the false part if the condition is false. The three conditions that can be entered from the machine keyboard (>, ≥, ≠) are actually five, because the parts of the expression written to the left and right of the relational symbols can be interchanged. So in addition to the standard C<6, 8>D is also possible.

The use of the end-of-expression instruction (◺) has been extended to the true and false parts of branch statements, so that depending on the evaluation of a condition, we can stop further evaluation of the expression or the whole program.

It is also possible to organize loops, the instructions between the beginning of the loop (↳) and the end of the loop (↰) instructions are repeated until you exit the loop with a conditional branch instruction or by causing an error. It is not possible to use conditional branching in a branching instruction, but loops can be nested.

Repeating calculations can also be implemented as subroutines (SUB:), so shorter and more transparent programs can be written. Up to 10 subroutines can be created in an area, and these subroutines can be referenced (and called) by their number (2ndF and the number keys). Unfortunately, conditional branches and loops cannot be used within subroutines.

Each letter of the English alphabet can be used as a constant or a variable using the previously known f()= instruction. In addition, it is possible to use variables consisting of lowercase letters, special symbols and numbers, the value of which is requested by the machine when they first occur in the area. These variable names (or labels) can be up to 7 characters long and cannot be used with the STO statement.

The EL-5050 and EL-5150 models already include the SC61720 microprocessor, joint development of Hitachi and Sharp, which can be found in many pocket computers of the era.

The EL-5030 was based on the same architecture, but some changes were made to the AER program: the use of lowercase variables was removed (VAR mode), but instead the machine was able to handle arrays (matrices). The array of variables specified with the DIM instruction, and also be easily used to perform operations on matrices using loops.

Although all three models knew significantly more than their EL-5100 predecessors, most of these features could not be used in AER mode. Thus, conversions between different number systems, logical operations, and conversion of the result to the displayed value (MDF) could not be used in the AER expression.

This version can also be found in the EL-9000 (EL-5200) graphing calculator as AER-I mode, with the same limitation as the EL-5030 version, so that lowercase variable names could not be used here either.

AER-I Third version (simplified)

The simplified version can not handle subroutines (SUB), arrays (DIM) and lowercase variable names (VAR), but variables and functions that can be used in statistical mode could also be included in AER expressions, i.e. it was easy to write statistical programs.

The entered expressions can be evaluated immediately by pressing the = key twice, even without changing the mode. Taking advantage of the two-line displays used, the numeric values entered appeared in the bottom (seven-segment) line, thus avoiding syntax errors due to incorrect keystrokes.

In Japan, a more durable version of the EL-5010 called EL-566 has been developed for educational purposes. Their knowledge and internal structure are completely the same. The machine was manufactured for years, with a newer version for the 2007 school year, the EL-566E, which still used the AER programming model, with minimal changes: due to separation from the comma operator, the AER expression separator was replaced with a colon, the programming functions could be accessed from a menu system. For convenience, the closing square brackets for the yes and no parts of the conditional branches have been given a separate menu item, which is a great advantage for subsequent program modifications.

AER-II

The graphical (function) capabilities of the machine could also be used in AER programs, but the statistics and conversion functions between number systems could still not be programmed.

Unfortunately, although a multi-line display were used, the programs appear in full and even hard-to-read lines.

AER in Pocket Computers

The AER-II system was already started adopting some advantages from BASIC-language pocket computers which were developed in parallel with the EL-5100. Sharp’s next-generation graphing calculators (EL-9200, EL-9300) use a new programming model based on a menu system with instructions based on BASIC keywords, so easy-to-understand programs can be written with label-based jumping capabilities. The programs on the display are now easier to read because of using line breaks on multi-line displays.