diff --git a/desktop_version/src/Game.cpp b/desktop_version/src/Game.cpp
index 84ba2851..24b9647a 100644
--- a/desktop_version/src/Game.cpp
+++ b/desktop_version/src/Game.cpp
@@ -6189,6 +6189,7 @@ void Game::createmenu( enum Menu::MenuName t, bool samemenu/*= false*/ )
         option("audio");
         option("game pad");
         option("accessibility");
+        option(loc::gettext("language"));
         option("return");
         menuyoff = 0;
         maxspacing = 15;
diff --git a/desktop_version/src/Localization.cpp b/desktop_version/src/Localization.cpp
new file mode 100644
index 00000000..8e9e2d60
--- /dev/null
+++ b/desktop_version/src/Localization.cpp
@@ -0,0 +1,389 @@
+#define LOCALIZATION_CPP
+#include "Localization.h"
+#include "LocalizationStorage.h"
+
+#include <utf8/unchecked.h>
+
+#include "Game.h"
+#include "UtilityClass.h"
+#include "VFormat.h"
+
+namespace loc
+{
+
+bool lang_set = false;
+std::string lang = "en";
+std::string lang_custom = "";
+LangMeta langmeta;
+
+// language screen list
+std::vector<LangMeta> languagelist;
+int languagelist_curlang;
+bool show_translator_menu = false;
+size_t limitscheck_current_overflow;
+
+int n_untranslated_roomnames = 0;
+int n_unexplained_roomnames = 0;
+int n_untranslated_roomnames_custom = 0;
+int n_unexplained_roomnames_custom = 0;
+
+int n_untranslated[COUNT_UNTRANSLATED_INDEX] = {0};
+
+const LangMeta* get_langmeta(void)
+{
+    if (game.currentmenuname == Menu::language && (unsigned)game.currentmenuoption < languagelist.size())
+    {
+        return &languagelist[game.currentmenuoption];
+    }
+
+    return &langmeta;
+}
+
+const char* gettext(const char* eng)
+{
+    if (lang == "en")
+    {
+        return eng;
+    }
+
+    return map_lookup_text(map_translation, eng, eng);
+}
+
+static const char* gettext_plural_english(const char* eng_plural, const char* eng_singular, int n)
+{
+    /* Do be consistent with negative number handling for other languages... */
+    if (n == 1 || n == -1)
+    {
+        return eng_singular;
+    }
+    return eng_plural;
+}
+
+const char* gettext_plural(const char* eng_plural, const char* eng_singular, int n)
+{
+    if (lang != "en")
+    {
+        unsigned char form = form_for_count(n);
+        char* key = add_disambiguator(form+1, eng_plural, NULL);
+        if (key != NULL)
+        {
+            const char* tra = map_lookup_text(map_translation_plural, key, NULL);
+
+            SDL_free(key);
+
+            if (tra != NULL)
+            {
+                return tra;
+            }
+        }
+    }
+    return gettext_plural_english(eng_plural, eng_singular, n);
+}
+
+void gettext_plural_fill(char* buf, size_t buf_len, const char* eng_plural, const char* eng_singular, const char* args_index, ...)
+{
+    /* Choose the right plural string based on a number, and then vformat that string.
+     * The first vararg determines the specific plural form. */
+
+    va_list args;
+    va_start(args, args_index);
+    int count = va_arg(args, int);
+    va_end(args);
+
+    const char* tra = gettext_plural(eng_plural, eng_singular, count);
+
+    va_start(args, args_index);
+    vformat_buf_valist(buf, buf_len, tra, args_index, args);
+    va_end(args);
+}
+
+std::string getnumber(int n)
+{
+    if (n < 0 || n > 100)
+    {
+        return help.String(n);
+    }
+
+    if (number[n].empty())
+    {
+        return help.String(n);
+    }
+    return number[n];
+}
+
+static bool is_script_custom(const char* script_id)
+{
+    return SDL_strncmp(script_id, "custom_", 7) == 0;
+}
+
+const TextboxFormat* gettext_cutscene(const std::string& script_id, const std::string& eng, char textcase)
+{
+    hashmap* map;
+    const char* map_script_key;
+    if (is_script_custom(script_id.c_str()))
+    {
+        map = map_translation_cutscene_custom;
+        map_script_key = &script_id.c_str()[7];
+    }
+    else
+    {
+        if (lang == "en")
+        {
+            return NULL;
+        }
+
+        map = map_translation_cutscene;
+        map_script_key = script_id.c_str();
+    }
+
+    uintptr_t ptr_cutscene_map;
+    bool found = hashmap_get(map, (void*) map_script_key, SDL_strlen(map_script_key), &ptr_cutscene_map);
+    hashmap* cutscene_map = (hashmap*) ptr_cutscene_map;
+
+    if (!found || cutscene_map == NULL)
+    {
+        return NULL;
+    }
+
+    size_t alloc_len;
+    char* key = add_disambiguator(textcase, eng.c_str(), &alloc_len);
+    if (key == NULL)
+    {
+        return NULL;
+    }
+
+    uintptr_t ptr_format;
+    found = hashmap_get(cutscene_map, (void*) key, alloc_len-1, &ptr_format);
+    const TextboxFormat* format = (TextboxFormat*) ptr_format;
+
+    SDL_free(key);
+
+    if (!found)
+    {
+        return NULL;
+    }
+    return format;
+}
+
+const char* get_roomname_explanation(bool custom_level, int roomx, int roomy)
+{
+    /* Never returns NULL. */
+
+    if (!fix_room_coords(custom_level, &roomx, &roomy))
+    {
+        return "";
+    }
+
+    const char* explanation;
+    if (custom_level)
+    {
+        explanation = explanation_roomnames_custom[roomy][roomx];
+    }
+    else
+    {
+        explanation = explanation_roomnames[roomy][roomx];
+    }
+    if (explanation == NULL)
+    {
+        return "";
+    }
+    return explanation;
+}
+
+const char* get_roomname_translation(bool custom_level, int roomx, int roomy)
+{
+    /* Only looks for the translation, doesn't return English fallback.
+     * Never returns NULL.
+     * Also used for room name translation mode. */
+
+    if (!fix_room_coords(custom_level, &roomx, &roomy))
+    {
+        return "";
+    }
+
+    const char* tra;
+    if (custom_level)
+    {
+        tra = translation_roomnames_custom[roomy][roomx];
+    }
+    else
+    {
+        tra = translation_roomnames[roomy][roomx];
+    }
+
+    if (tra == NULL)
+    {
+        return "";
+    }
+    return tra;
+}
+
+const char* gettext_roomname(bool custom_level, int roomx, int roomy, const char* eng, bool special)
+{
+    if (!custom_level && lang == "en")
+    {
+        return eng;
+    }
+
+    if (special)
+    {
+        return gettext_roomname_special(eng);
+    }
+
+    const char* tra = get_roomname_translation(custom_level, roomx, roomy);
+    if (tra[0] == '\0')
+    {
+        return eng;
+    }
+    return tra;
+}
+
+const char* gettext_roomname_special(const char* eng)
+{
+    if (lang == "en")
+    {
+        return eng;
+    }
+
+    return map_lookup_text(map_translation_roomnames_special, eng, eng);
+}
+
+bool is_cutscene_translated(const std::string& script_id)
+{
+    hashmap* map;
+    const char* map_script_key;
+    if (is_script_custom(script_id.c_str()))
+    {
+        map = map_translation_cutscene_custom;
+        map_script_key = &script_id.c_str()[7];
+    }
+    else
+    {
+        if (lang == "en")
+        {
+            return false;
+        }
+
+        map = map_translation_cutscene;
+        map_script_key = script_id.c_str();
+    }
+
+    uintptr_t ptr_unused;
+    return hashmap_get(map, (void*) map_script_key, SDL_strlen(map_script_key), &ptr_unused);
+}
+
+uint32_t toupper_ch(uint32_t ch)
+{
+    // Convert a single Unicode codepoint to its uppercase variant
+    // Supports important Latin (1 and A), Cyrillic and Greek
+
+    // Turkish i?
+    if (get_langmeta()->toupper_i_dot && ch == 'i') return 0x130;
+
+    // a-z?
+    if ('a' <= ch && ch <= 'z') return ch - 0x20;
+
+    // Latin-1 Supplement? But not the division sign
+    if (0xE0 <= ch && ch <= 0xFE && ch != 0xF7) return ch - 0x20;
+
+    // ß? Yes, we do have this! And otherwise we could only replace it with SS later on.
+    if (ch == 0xDF) return 0x1E9E;
+
+    // ÿ?
+    if (ch == 0xFF) return 0x178;
+
+    // Let's get some exceptions for Latin Extended-A out of the way, starting with ı
+    if (ch == 0x131) return 'I';
+
+    // This range between two obscure exceptions...
+    if (0x139 <= ch && ch <= 0x148 && ch % 2 == 0) return ch - 1;
+
+    // The rest of Latin Extended-A?
+    if (0x100 <= ch && ch <= 0x177 && ch % 2 == 1) return ch - 1;
+
+    // Okay, Ÿ also pushed some aside...
+    if (0x179 <= ch && ch <= 0x17E && ch % 2 == 0) return ch - 1;
+
+    // Can't hurt to support Romanian properly...
+    if (ch == 0x219 || ch == 0x21B) return ch - 1;
+
+    // Cyrillic а-я?
+    if (0x430 <= ch && ch <= 0x44F) return ch - 0x20;
+
+    // There's probably a good reason Cyrillic upper and lower accents are wrapped around the alphabet...
+    if (0x450 <= ch && ch <= 0x45F) return ch - 0x50;
+
+    // Apparently a Ukrainian letter is all the way over there, why not.
+    if (ch == 0x491) return ch - 1;
+
+    // Time for Greek, thankfully we're not making a lowercasing function with that double sigma!
+    if (ch == 0x3C2) return 0x3A3;
+
+    // The entire Greek alphabet then, along with two accented letters
+    if (0x3B1 <= ch && ch <= 0x3CB) return ch - 0x20;
+
+    // Unfortunately Greek accented letters are all over the place.
+    if (ch == 0x3AC) return 0x386;
+    if (0x3AD <= ch && ch <= 0x3AF) return ch - 0x25;
+    if (ch == 0x3CC) return 0x38C;
+    if (ch == 0x3CD || ch == 0x3CE) return ch - 0x3F;
+
+    // Nothing matched! Just leave it as is
+    return ch;
+}
+
+std::string toupper(const std::string& lower)
+{
+    // Convert a UTF-8 string to uppercase
+    if (!get_langmeta()->toupper)
+    {
+        return lower;
+    }
+
+    std::string upper;
+    std::back_insert_iterator<std::string> inserter = std::back_inserter(upper);
+    std::string::const_iterator iter = lower.begin();
+    bool ignorenext = false;
+    while (iter != lower.end())
+    {
+        uint32_t ch = utf8::unchecked::next(iter);
+
+        if (get_langmeta()->toupper_lower_escape_char && ch == '~')
+        {
+            ignorenext = true;
+            continue;
+        }
+
+        if (!ignorenext)
+        {
+            ch = toupper_ch(ch);
+        }
+        utf8::unchecked::append(ch, inserter);
+
+        ignorenext = false;
+    }
+
+    return upper;
+}
+
+std::string remove_toupper_escape_chars(const std::string& _s)
+{
+    // No-op, except if langmeta.toupper_lower_escape_char, to remove the ~ escape character
+
+    if (!get_langmeta()->toupper_lower_escape_char)
+    {
+        return _s;
+    }
+
+    std::string s = std::string(_s);
+    for (signed int i = s.size()-1; i >= 0; i--)
+    {
+        if (s[i] == '~')
+        {
+            s.erase(i, 1);
+        }
+    }
+    return s;
+}
+
+} /* namespace loc */
diff --git a/desktop_version/src/Localization.h b/desktop_version/src/Localization.h
new file mode 100644
index 00000000..4ef98a2a
--- /dev/null
+++ b/desktop_version/src/Localization.h
@@ -0,0 +1,93 @@
+#ifndef LOCALIZATION_H
+#define LOCALIZATION_H
+
+#include <stdint.h>
+#include <string>
+#include <vector>
+
+/* The translator menu will appear in any of the following circumstances:
+ * - The "lang" folder is NOT next to data.zip, but it is found in "desktop_version" within which the game is running
+ * - The command line argument "-translator" is passed
+ * - ALWAYS_SHOW_TRANSLATOR_MENU is defined
+ */
+// #define ALWAYS_SHOW_TRANSLATOR_MENU
+
+namespace loc
+{
+
+struct LangMeta
+{
+    bool active; // = true, language is shown in the list
+    std::string code;
+    std::string nativename;
+    std::string credit;
+    std::string action_hint;
+    bool autowordwrap; // = true; enable automatic wordwrapping
+    bool toupper; // = true; enable automatic full-caps for menu options
+    bool toupper_i_dot; // = false; enable Turkish i mapping when uppercasing
+    bool toupper_lower_escape_char; // = false; enable ~ to mark lowercase letters for uppercasing
+    std::string menu_select;
+    std::string menu_select_tight;
+    unsigned char font_w;
+    unsigned char font_h;
+};
+
+struct TextboxFormat
+{
+    const char* text;
+    unsigned short wraplimit; // = 36*8-pad_left-pad_right; no effect if tt or !langmeta.autowordwrap
+    unsigned short wraplimit_raw; // original value of wraplimit, only used for language file sync
+    bool tt; // teletype, don't auto-wordwrap
+    bool centertext; // whether the text should be centered inside the box
+    unsigned char pad_left; // pad with X characters
+    unsigned char pad_right;
+    unsigned short padtowidth; // pad to X pixels (0 to disable)
+};
+
+extern bool lang_set;
+extern std::string lang;
+extern std::string lang_custom;
+extern LangMeta langmeta;
+extern std::vector<LangMeta> languagelist;
+extern int languagelist_curlang;
+extern bool show_translator_menu;
+extern size_t limitscheck_current_overflow;
+
+extern int n_untranslated_roomnames;
+extern int n_unexplained_roomnames;
+extern int n_untranslated_roomnames_custom;
+extern int n_unexplained_roomnames_custom;
+
+enum n_untranslated_index
+{
+    UNTRANSLATED_STRINGS = 0,
+    UNTRANSLATED_NUMBERS,
+    UNTRANSLATED_STRINGS_PLURAL,
+    UNTRANSLATED_CUTSCENES,
+    UNTRANSLATED_ROOMNAMES_SPECIAL,
+    COUNT_UNTRANSLATED_INDEX
+};
+extern int n_untranslated[COUNT_UNTRANSLATED_INDEX];
+
+
+const LangMeta* get_langmeta(void);
+
+const char* gettext(const char* eng);
+const char* gettext_plural(const char* eng_plural, const char* eng_singular, int count);
+void gettext_plural_fill(char* buf, size_t buf_len, const char* eng_plural, const char* eng_singular, const char* args_index, ...);
+std::string getnumber(int n);
+const TextboxFormat* gettext_cutscene(const std::string& script_id, const std::string& eng, char textcase);
+const char* get_roomname_explanation(bool custom_level, int roomx, int roomy);
+const char* get_roomname_translation(bool custom_level, int roomx, int roomy);
+const char* gettext_roomname(bool custom_level, int roomx, int roomy, const char* eng, bool special);
+const char* gettext_roomname_special(const char* eng);
+
+bool is_cutscene_translated(const std::string& script_id);
+
+uint32_t toupper_ch(uint32_t ch);
+std::string toupper(const std::string& lower);
+std::string remove_toupper_escape_chars(const std::string& _s);
+
+} /* namespace loc */
+
+#endif /* LOCALIZATION_H */