Should I use ÷ or / in formulas?

Use / or a stacked fraction anywhere beyond basic arithmetic. The international standard for mathematical notation, ISO 80000-2, advises against ÷ for division, and programming languages and spreadsheets only accept /. ÷ is fine in teaching materials and on calculator keys, where readers expect it.

Why does ÷ 10% on a Danish receipt mean a discount?

In Denmark and Norway ÷ can mean minus. On receipts and in commercial text, ÷ 10% can mean minus 10%, a discount rather than a division. Keep that in mind before parsing Nordic figures.

Why is 1e-20 ≈ 0 false in Julia?

Julia's ≈ is isapprox, which by default only uses a relative tolerance, and a tolerance relative to zero is zero, so nothing but 0 itself counts as close to 0. Pass an absolute tolerance when comparing near zero: `isapprox(x, 0; atol=1e-12)`. Python's math.isclose has the same default and takes abs_tol for this case.

Why does Alt+247 give ≈ instead of ÷?

The short code comes from the DOS table, and on a US system 247 there is the almost-equal sign, which is why ≈ appeared. Add the zero, Alt+0247, and Windows uses its own table, where 247 is ÷. That leading zero is the whole difference between the two lookups.

ASCII 247

Division signAlmost equal to

ASCII 247 is ÷ on Windows and ≈ in old DOS.

÷ (Division Sign) is byte 247 in Windows-1252, Unicode U+00F7. It's the obelus from school arithmetic and calculator keys. Its position causes a quiet regex bug. Windows-1252 and Unicode put ÷ in the middle of the lowercase accented letters, just as × (215) sits among the capitals, so a name pattern like [A-Za-zÀ-ÿ]+, written to allow accented names, also accepts × and ÷. Exclude them explicitly, or use a Unicode letter class such as \p{L}, which leaves both out. Old DOS had ÷ at 246. In UTF-8 it's C3 B7, which misreads as Ã·. The HTML entity is ÷.

≈ (Almost Equal To) is byte 247 in code page 437, the original IBM PC character set, and maps to Unicode U+2248. It's the everyday approximately sign, as in π ≈ 3.14. Maths has close relatives that aren't interchangeable with it: ≃ (U+2243) means asymptotically equal, and ≅ (U+2245) marks congruent shapes or isomorphic structures. In plain text people write ~5 min for the same idea, and search won't connect the tilde with ≈. Code page 850 put the cedilla ¸ on this byte. In UTF-8, ≈ is E2 89 88, and the HTML entity is ≈.

Math SymbolMath Symbol

Technical Details

Code Example

</>

#include <stdio.h>

int main(void) {
    /* On Windows (code page 1252) byte 247 is ÷.
       Modern terminals expect UTF-8, so print the code point, not the byte. */
    printf("\u00F7\n");   /* UTF-8: C3 B7 */

    unsigned char b = 247;         /* the raw Windows-1252 byte */
    printf("%d 0x%02X\n", b, b);    /* 247 0xF7 */
    return 0;
}

Code Example

</>

#include <stdio.h>

int main(void) {
    /* Byte 247 is ≈ only on a console using code page 437
       (chcp 437 on Windows, or DOSBox). On a UTF-8 terminal the
       lone byte 0xF7 is invalid and prints as garbage, often �. */
    putchar(247);

    /* char is signed on x86, so a plain char holding 0xF7 is -9.
       Use unsigned char when you compare or index by byte value. */
    char c = (char)247;
    unsigned char u = 247;
    printf("\n%d %d\n", c, u);   /* -9 247 */

    /* Portable: print the Unicode character as UTF-8 instead. */
    printf("\u2248\n");          /* E2 89 88 */
    return 0;
}

How to Type It

Windows: Hold Alt and type 0247 on the numeric keypad, with Num Lock on. The leading zero picks the Windows code page. Alt+246, without the zero, types it too. No numeric keypad? In Word, type 00F7 and press Alt+X.
Mac: Press Option+/.
Linux: In GNOME, or any desktop running IBus, press Ctrl+Shift+U, type f7, then Space. With a Compose key set up, press Compose, colon, minus.
Phone: iPhone: tap the character at the top of this page to copy it. Android (Gboard): tap ?123, then =\< and tap ÷.

How to Type It

Windows: Hold Alt and type 247 on the numeric keypad, with no leading zero. US PCs insert ≈, but PCs set to code page 850, common in Western Europe, insert ¸. No numeric keypad? In Word, type 2248 and press Alt+X.
Mac: Press Option+X.
Linux: In GNOME, or any desktop running IBus, press Ctrl+Shift+U, type 2248, then Space. With a Compose key set up, press Compose, tilde, tilde.
Phone: Tap the character at the top of this page to copy it.

Frequently Asked Questions

Use / or a stacked fraction anywhere beyond basic arithmetic. The international standard for mathematical notation, ISO 80000-2, advises against ÷ for division, and programming languages and spreadsheets only accept /. ÷ is fine in teaching materials and on calculator keys, where readers expect it.
In Denmark and Norway ÷ can mean minus. On receipts and in commercial text, ÷ 10% can mean minus 10%, a discount rather than a division. Keep that in mind before parsing Nordic figures.
The short code comes from the DOS table, and on a US system 247 there is the almost-equal sign, which is why ≈ appeared. Add the zero, Alt+0247, and Windows uses its own table, where 247 is ÷. That leading zero is the whole difference between the two lookups.

Frequently Asked Questions

Julia's ≈ is isapprox, which by default only uses a relative tolerance, and a tolerance relative to zero is zero, so nothing but 0 itself counts as close to 0. Pass an absolute tolerance when comparing near zero: `isapprox(x, 0; atol=1e-12)`. Python's math.isclose has the same default and takes abs_tol for this case.
The short code comes from the DOS table, and on a US system 247 there is the almost-equal sign, which is why ≈ appeared. Add the zero, Alt+0247, and Windows uses its own table, where 247 is ÷. That leading zero is the whole difference between the two lookups.

ASCII 247

Division signAlmost equal to

ASCII 247 is ÷ on Windows and ≈ in old DOS.

Math Symbol

#include <stdio.h>

int main(void) {

/* On Windows (code page 1252) byte 247 is ÷.

Modern terminals expect UTF-8, so print the code point, not the byte. */

printf("\u00F7\n"); /* UTF-8: C3 B7 */

unsigned char b = 247; /* the raw Windows-1252 byte */

printf("%d 0x%02X\n", b, b); /* 247 0xF7 */

return 0;

}

#include <stdio.h>

int main(void) {

/* Byte 247 is ≈ only on a console using code page 437

(chcp 437 on Windows, or DOSBox). On a UTF-8 terminal the

lone byte 0xF7 is invalid and prints as garbage, often �. */

putchar(247);

/* char is signed on x86, so a plain char holding 0xF7 is -9.

Use unsigned char when you compare or index by byte value. */

char c = (char)247;

unsigned char u = 247;

printf("\n%d %d\n", c, u); /* -9 247 */

/* Portable: print the Unicode character as UTF-8 instead. */

printf("\u2248\n"); /* E2 89 88 */

return 0;

}