Ö shows as Ã– in my database output, how do I fix it?

Fix the connection charset, not the stored rows. If the table holds proper UTF-8 and only the output is wrong, tell the client connection it's UTF-8; in MySQL or MariaDB that's `SET NAMES utf8mb4` right after connecting. Don't try a find-and-replace on Ã– first. The second character is an en dash, and any cleanup that normalizes dashes turns it into Ã-, which no decoder can turn back into Ö.

Is it OK to write Ö as OE?

In German, yes. OE is the accepted fallback when you can't type Ö, as in a username or an email address. In Swedish, Finnish and Turkish it's wrong, because Ö is a letter of its own there and OE just reads as two vowels. Dropping the dots altogether is worse in every language: in German, schon means already while schön means beautiful.

Is Ö the same letter as Ø?

No. They stand for a similar vowel, but they're separate characters. Swedish, Finnish and German write Ö, while Danish and Norwegian write Ø (U+00D8, byte 216). Plain string comparison treats them as unrelated, so a Norwegian name stored with Ö won't match a lookup typed with Ø. Keep whichever the language or the person uses, and only swap one for the other in a deliberate transliteration step.

Why does my old text box start with ÖÄÄ instead of ╓──?

The file is code page 437 being read as Windows-1252, where 214 is Ö. The rest of the frame follows the same swap: ╓──╖ comes out as ÖÄÄ·, the ║ sides turn into º, and the bottom edge ╙──╜ reads ÓÄÄ½. Long runs of Ä are the giveaway. Decode it as 437 on purpose, for example `open(path, encoding='cp437')` in Python, and save a UTF-8 copy.

ASCII 214

Uppercase O with umlautBox top-left corner, double vertical

ASCII 214 is Ö on Windows and ╓ in old DOS.

Ö (Capital O with Umlaut) is byte 214 in Windows-1252, Unicode U+00D6. German, Swedish, Finnish and Turkish all use it, and Hungarian has it next to a lookalike with two slanted strokes, Ő, as in ŐSZ (autumn). Windows-1252 has Ö but not Ő. Hungarian text saved in Windows-1250, where Ő is byte 0xD5, and then read as Windows-1252 shows Õ in its place, so ŐSZ becomes ÕSZ. Old DOS had Ö at 153. In UTF-8 it's C3 96, which misreads as Ã–. The HTML entity is Ö.

╓ (Double Down, Single Right) is byte 214 in code page 437, the original IBM PC character set, and maps to Unicode U+2553. It's the top-left corner of a box with double sides and a single top edge: ─ (196) runs right toward ╖ (183) and ║ (186) drops down. Encoding detection won't help you identify files built from pieces like this. Every box byte from 179 to 218 is also a valid ISO-8859-1 character, so the Linux file command reports a code page 437 frame as ISO-8859 text. You have to know, or guess from the gibberish, that it's 437. Code page 850 put Í on this byte. In UTF-8 it's E2 95 93, and the HTML entity is &boxDr;.

Latin LetterBox Drawing

Technical Details

Code Example

</>

#include <stdio.h>

int main(void) {
    /* On Windows (code page 1252) byte 214 is Ö.
       Modern terminals expect UTF-8, so print the code point, not the byte. */
    printf("\u00D6\n");   /* UTF-8: C3 96 */

    unsigned char b = 214;         /* the raw Windows-1252 byte */
    printf("%d 0x%02X\n", b, b);    /* 214 0xD6 */
    return 0;
}

Code Example

</>

#include <stdio.h>

int main(void) {
    /* Byte 214 is ╓ only on a console using code page 437
       (chcp 437 on Windows, or DOSBox). On a UTF-8 terminal the
       lone byte 0xD6 is invalid and prints as garbage, often �. */
    putchar(214);

    /* char is signed on x86, so a plain char holding 0xD6 is -42.
       Use unsigned char when you compare or index by byte value. */
    char c = (char)214;
    unsigned char u = 214;
    printf("\n%d %d\n", c, u);   /* -42 214 */

    /* Portable: print the Unicode character as UTF-8 instead. */
    printf("\u2553\n");          /* E2 95 93 */
    return 0;
}

How to Type It

Windows: Hold Alt and type 0214 on the numeric keypad, with Num Lock on. The leading zero picks the Windows code page. Alt+153, without the zero, types it too. No numeric keypad? In Word, type 00D6 and press Alt+X.
Mac: Press Option+U, then Shift+O.
Linux: In GNOME, or any desktop running IBus, press Ctrl+Shift+U, type d6, then Space. With a Compose key set up, press Compose, quote, O.
Phone: Tap Shift, then long-press O and pick Ö.

How to Type It

Windows: Hold Alt and type 214 on the numeric keypad, with no leading zero. US PCs insert ╓, but PCs set to code page 850, common in Western Europe, insert Í. No numeric keypad? In Word, type 2553 and press Alt+X.
Mac: Press Control+Command+Space to open the Character Viewer, then search for "box drawings down double and right single".
Linux: In GNOME, or any desktop running IBus, press Ctrl+Shift+U, type 2553, then Space.
Phone: Phone keyboards have no key for it. Tap the character at the top of this page to copy it.

Frequently Asked Questions

Fix the connection charset, not the stored rows. If the table holds proper UTF-8 and only the output is wrong, tell the client connection it's UTF-8; in MySQL or MariaDB that's `SET NAMES utf8mb4` right after connecting. Don't try a find-and-replace on Ã– first. The second character is an en dash, and any cleanup that normalizes dashes turns it into Ã-, which no decoder can turn back into Ö.
In German, yes. OE is the accepted fallback when you can't type Ö, as in a username or an email address. In Swedish, Finnish and Turkish it's wrong, because Ö is a letter of its own there and OE just reads as two vowels. Dropping the dots altogether is worse in every language: in German, schon means already while schön means beautiful.
No. They stand for a similar vowel, but they're separate characters. Swedish, Finnish and German write Ö, while Danish and Norwegian write Ø (U+00D8, byte 216). Plain string comparison treats them as unrelated, so a Norwegian name stored with Ö won't match a lookup typed with Ø. Keep whichever the language or the person uses, and only swap one for the other in a deliberate transliteration step.
The file is code page 437 being read as Windows-1252, where 214 is Ö. The rest of the frame follows the same swap: ╓──╖ comes out as ÖÄÄ·, the ║ sides turn into º, and the bottom edge ╙──╜ reads ÓÄÄ½. Long runs of Ä are the giveaway. Decode it as 437 on purpose, for example `open(path, encoding='cp437')` in Python, and save a UTF-8 copy.

Frequently Asked Questions

The file is code page 437 being read as Windows-1252, where 214 is Ö. The rest of the frame follows the same swap: ╓──╖ comes out as ÖÄÄ·, the ║ sides turn into º, and the bottom edge ╙──╜ reads ÓÄÄ½. Long runs of Ä are the giveaway. Decode it as 437 on purpose, for example `open(path, encoding='cp437')` in Python, and save a UTF-8 copy.

ASCII 214

Uppercase O with umlautBox top-left corner, double vertical

ASCII 214 is Ö on Windows and ╓ in old DOS.

Latin Letter

Box Drawing

#include <stdio.h>

int main(void) {

/* On Windows (code page 1252) byte 214 is Ö.

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

printf("\u00D6\n"); /* UTF-8: C3 96 */

unsigned char b = 214; /* the raw Windows-1252 byte */

printf("%d 0x%02X\n", b, b); /* 214 0xD6 */

return 0;

}

#include <stdio.h>

int main(void) {

/* Byte 214 is ╓ only on a console using code page 437

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

lone byte 0xD6 is invalid and prints as garbage, often �. */

putchar(214);

/* char is signed on x86, so a plain char holding 0xD6 is -42.

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

char c = (char)214;

unsigned char u = 214;

printf("\n%d %d\n", c, u); /* -42 214 */

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

printf("\u2553\n"); /* E2 95 93 */

return 0;

}