diff --git a/font_manipulator.py b/font_manipulator.py
new file mode 100644
index 0000000..c8800b9
--- /dev/null
+++ b/font_manipulator.py
@@ -0,0 +1,221 @@
+# adapted from https://svn.code.sf.net/p/unifraktur/code
+
+# -*- coding: utf-8 -*-
+
+import fileinput
+import os
+import shutil
+import ast
+
+def string_auswerten(string):
+	try:
+		string = ast.literal_eval(string)
+	except ValueError:
+		pass
+	return string
+
+def glyph_nach_datei(glyph):
+	if glyph.startswith("uni"):
+		return glyph + ".glyph"
+	elif glyph == " ":
+		return "space.glyph"
+	else:
+		chars = list(glyph)
+		i = 0
+		while i < len(chars):
+			if chars[i].isupper():
+				chars.insert(i, "_")
+				i+=1
+			i+=1
+		return "".join(chars) + ".glyph"
+
+def stems_zerlegen(stemstring):
+	ausgabe = []
+	for stem in stemstring.split(">"):
+		if stem.strip():
+			ebene = stem.split("<")[0].split()
+			positionen_roh = stem.split("<")[1].split()
+			positionen = map(list,zip(*[iter(positionen_roh)]*2))
+			ausgabe += [[ebene, positionen]]
+	return ausgabe
+
+def stems_vereinen(stemslist):
+	ausgabe = []
+	for stem in [stem for stems in stemslist for stem in stems]:
+		neu = True
+		for bestehender_stem in ausgabe:
+			if stem[0] == bestehender_stem[0]:
+				bestehender_stem[1] += stem[1]
+				neu = False
+				break
+		if neu:
+			ausgabe += [stem]
+	return ausgabe
+
+def stems_nach_string(stems):
+	ausgabe = ""
+	for stem in stems:
+		ausgabe += " ".join(stem[0])
+		ausgabe += "<"
+		ausgabe += " ".join([komp for position in stem[1] for komp in position])
+		ausgabe += "> "
+	return ausgabe
+
+class Schrift:
+	def __init__(self, name, quellordner, zielordner):
+		self.name = name
+		self.ordner = zielordner
+		if self.ordner in os.listdir("."):
+			raise Exception("Fehler: Zielordner existiert bereits.")
+			return None 
+		else:
+			os.mkdir(self.ordner)
+		
+		self.props = []
+		
+		for datei in os.listdir(quellordner):
+			if os.path.isfile(quellordner + "/" + datei):
+				shutil.copy(quellordner + "/" + datei, self.ordner)
+		
+		self.props_oeffnen()
+		for i,zeile in enumerate(self.props):
+			for nameString in ["FontName:", "FullName:", "FamilyName:"]:
+				if zeile.startswith(nameString):
+					self.props[i] = nameString + " " + self.name + "\n"
+	
+	def __del__(self):
+		self.props_schliessen()
+	
+	def props_oeffnen(self):
+		if not self.props:
+			with open(self.ordner + "/font.props", "rb") as propsdatei:
+				self.props = propsdatei.readlines()
+	
+	def props_schliessen(self):
+		if self.props:
+			with open(self.ordner + "/font.props", "wb") as propsdatei:
+				propsdatei.writelines(self.props)
+
+	def ersetzen_durch_Kompositglyph(self, ziel, teil1, teil2):
+		teile = (teil1,teil2)
+		self._ersetzen_durch_Kompositglyph(ziel, teile)
+
+	def _ersetzen_durch_Kompositglyph(self, ziel, teile):
+		breiten = []
+		encodings = []
+		hstems = []
+		vstems = []
+		dstems = []
+		verschiebungen = []
+		verschiebungen.append(0)
+		
+		self.props_oeffnen()
+		in_kerningtabelle = True
+		anzahl_vor = None
+		kerning_start = None
+		indizes = [None, None]
+		j = None
+		kerning = 0
+		for i,zeile in enumerate(self.props):
+			if zeile.startswith("KernClass2"):
+				in_kerningtabelle = True
+				anzahl_vor = int(zeile.split()[1])
+				j = anzahl_vor-1
+				anzahl_nach = int(zeile.split()[2])
+				kerning_start = i
+			elif in_kerningtabelle:
+				if zeile.startswith(" "):
+					if ("{}" in zeile):
+						if all(indizes):
+							a = indizes[0] - kerning_start
+							b = indizes[1] - kerning_start - anzahl_vor + 1
+							kerning = int(zeile.split("{}")[a*anzahl_nach+b])
+					else:
+						inhalt = zeile.split()[1:]
+						if teile[j>0] in inhalt:
+							inhalt += [ziel]
+						if teile[j<=0] in inhalt:
+							indizes[j<=0] = i
+						neuer_inhalt = " ".join(inhalt)
+						self.props[i] = " " + str(len(neuer_inhalt)) + " " + neuer_inhalt + "\n"
+						j -= 1
+						continue
+				
+				in_kerningtabelle = False
+				kerning_start = None
+				anzahl_vor = None
+				indizes = [None, None]
+				j = None
+
+		for i,teil in enumerate(teile):
+			for zeile in open(self.ordner + "/" + glyph_nach_datei(teil), "rb"):
+				if zeile.startswith("Width:"):
+					breiten += [int(zeile.split()[1])]
+				elif zeile.startswith("Encoding:"):
+					encodings += [zeile.split()[2:]]
+				elif zeile.startswith("HStem:"):
+					hstems += [stems_zerlegen(zeile[7:-1])]
+				elif zeile.startswith("VStem:"):
+					vstems += [stems_zerlegen(zeile[7:-1])]
+				elif zeile.startswith("DStem2:"):
+					dstems += [stems_zerlegen(zeile[8:-1])]
+		
+		zieldatei = glyph_nach_datei(ziel)
+		
+		first = True
+		for j,teil in enumerate(teile):
+			if(first):
+				first = False
+				continue
+
+			verschiebung = kerning
+			r = range(0, j)
+			for x in r:
+				verschiebung = verschiebung + breiten[x]
+			verschiebungen.append(verschiebung)
+
+			if j < len(hstems):
+				for hstem in hstems[j]:
+					for position in hstem[1]:
+						for i in (0,1):
+							position[i] = str(string_auswerten(position[i]) + verschiebung)
+
+			if j < len(vstems):
+				for vstem in vstems[j]:
+					vstem[0][0] = str(string_auswerten(vstem[0][0]) + verschiebung)
+
+			if j < len(dstems):
+				for dstem in dstems[j]:
+					for i in (0,2):
+						dstem[0][i] = str(string_auswerten(dstem[0][i]) + verschiebung)
+		
+		for zeile in fileinput.input (self.ordner + "/" + zieldatei, inplace=1):
+			if (
+					   zeile.startswith("StartChar:")
+					or zeile.startswith("Encoding:")
+					or zeile.startswith("VWidth:")
+					or zeile.startswith("GlyphClass:")
+					or zeile.startswith("Flags:")
+					or zeile.startswith("EndChar")
+				):
+				print zeile,
+			elif zeile.startswith("Fore"):
+				print zeile,
+				for j,teil in enumerate(teile):
+					print "Refer: %s %s N 1 0 0 1 %s 0 2" % (encodings[j][1], encodings[j][0], verschiebungen[j])
+			elif zeile.startswith("Width:"):
+				print "Width: %i" % (sum(map(int, breiten))+kerning)
+			elif zeile.startswith("LayerCount:"):
+				print zeile,
+				
+				stemstring = stems_nach_string(stems_vereinen(hstems))
+				if stemstring:
+					print "HStem: " + stems_nach_string(stems_vereinen(hstems))
+					
+				stemstring = stems_nach_string(stems_vereinen(vstems))
+				if stemstring:
+					print "VStem: " + stems_nach_string(stems_vereinen(vstems))
+					
+				stemstring = stems_nach_string(stems_vereinen(dstems))
+				if stemstring:
+					print "DStem2: " + stems_nach_string(stems_vereinen(dstems))
diff --git a/font_mod_character_pairs b/font_mod_character_pairs
new file mode 100644
index 0000000..0f47998
--- /dev/null
+++ b/font_mod_character_pairs
@@ -0,0 +1,2 @@
+Str|ate|Ret|urn|Adm|ira|l C|Cho|ose
+Sel|ect
\ No newline at end of file
diff --git a/font_mod_run.py b/font_mod_run.py
new file mode 100644
index 0000000..8672a58
--- /dev/null
+++ b/font_mod_run.py
@@ -0,0 +1,18 @@
+#! /usr/bin/env python
+# -*- coding: utf-8 -*-
+
+from font_manipulator import *
+import itertools
+
+start_unicode = "0xE000"
+unicode_point_int = int(start_unicode, 0)
+pairs = [line.rstrip('\n').split("|") for line in open('font_mod_character_pairs')]
+pairs = filter(None, list(itertools.chain.from_iterable(pairs)))
+
+name = "A-OTF-UDShinGoPro-Regular"
+M20 = Schrift(name, "../"+name+".sfdir", "../"+name+"_mod.sfdir")
+
+for pair in pairs:
+    name = "uni%X" % unicode_point_int
+    M20._ersetzen_durch_Kompositglyph(name, list(pair))
+    unicode_point_int = unicode_point_int + 1
diff --git a/kc_tr_notes.txt b/kc_tr_notes.txt
new file mode 100644
index 0000000..756c860
--- /dev/null
+++ b/kc_tr_notes.txt
@@ -0,0 +1,156 @@
+1. get app and patch
+2. decrypt both <url>
+3. copy app, then patch into dir "original", overwriting as necessary
+4. copy dir "original" to dir "translation", overwrite with contents of translation patch
+
+next up, unityex unpacking, sadly gameobjects tree view cannot be unpacked from command line, and ui unpack doesn't create nested trees
+
+first task though should be creating a script to create translation from original using the repo
+
+the unity asset manipulations tools <url> are not great, so they'll mostly be used manually and then binary patches created, i think (xdelta)
+
+dds files extracted via conversion from asset files don'd loat in faststone, but irfanview can handle them. support contacted
+
+
+assembly-csharp was unpacked with jetbrains c# decompiler but not yet repackable. binary patch?
+
+wtf is a monobehavior, is the file format described? found mention that it's a base class of unity objects and the exact type is clarified by the ids. needs more research.
+
+quest descriptions have space for 4 lines, seem to need newlines as they break below ui elements otherwise,
+fit about this much: "Have 2 ships in your main fleet. Mmmmmmm mmm"
+
+ship greeting texts have space for 4 lines and auto-break,
+fit about this much: "mmm mmmmmm mmmmmm mmmmm mmmmmm", 630 px
+
+font for both appears to be A-OTF-ShinGoPro-Regular.ttf, 18, line leading setting of -5
+
+
+選択 6 "Select"
+提督コマンド 18 "Admiral Command   "
+戦略へ 9 "Strategy "
+決定 6 "Choose"
+戻る 6 "Return"
+
+wrote a script to inject english versions of jp strings via raw string-matching
+
+wrote a script to inject modified binary files back into .asset files in bulk
+
+
+
+=head1 Assembly-CSharp.dll
+
+Contains mostly strings in UCS-2 LE. But however a FEW strings (enums) are also UTF8.
+No idea if the latter show up ingame.
+
+> "Most tutorials like this, that I've seen,
+> always use .NET Reflector with the Reflexil plugin.
+> However, I'm going to be using ILSpy,
+> which is a free alternative, with the Reflexil plugin."
+
+https://github.com/icsharpcode/ILSpy/releases
+https://github.com/sailro/Reflexil/releases
+https://github.com/0xd4d/dnSpy/releases
+jetbrains c# decompiler
+
+None of these that i tried were able to recompile the dll into something that doesn't
+crash the vita, so i went the way of binary modding the dll.
+
+Sadly due to the strings being in UTF16 and binary modding not allowing
+changes to the byte length, it's necessary to mod the fonts in order to
+show more characters per byte than usually possible.
+
+A-OTF-ShinGoPro-Regular.ttf in assets 2, used for quest name texts, ship greetings, ...
+A-OTF-UDShinGoPro-Regular.ttf in assets 3, used for the popup button guide at the bottom, ...
+
+online font tester: http://torinak.com/font/lsfont.html
+
+For modding fonts i use FontForge. When using it there might be
+an error window in the taskbar but not actually visible,
+click on that and hit esc once.
+
+Detailed steps from start to finish:
+
+- put unityex in ../unity_tools
+- put the decrypted game in ../kc_original
+- run `unpack_original_files.pl`
+- open ../kc_original\Media\Managed\Assembly-CSharp.dll in JetBrains dotPeek
+  - right-click Assembly-CSharp, export to project (might need two clicks (???))
+  - ..\kc_original_unpack\Media\Managed\
+  - progress bar in bottom right
+
+<dll dictionary updater here>
+
+But first it makes sense to run translate_utf16_binary.pl to regenerate
+Assembly-CSharp.dll. It might complain, depending on whether the
+translation dictionary has changed, that pairs are missing. Add them
+judiciously to the font_mod_character_pairs file until it stops
+complaining about that.
+
+The following steps set up a font to prepare it for injecting character pairs.
+
+- do this with:
+  - ..\kc_original_unpack\Media\Unity_Assets_Files\sharedassets2\A-OTF-ShinGoPro-Regular.ttf
+  - ..\kc_original_unpack\Media\Unity_Assets_Files\sharedassets3\A-OTF-UDShinGoPro-Regular.ttf
+    - open the ttf file in font forge
+      - CID > flatten
+      - save as directory in d:\vita\
+      - close
+      - open font dir in font forge
+      - save (wait a little)
+      - close
+      - open font file in font forge
+      - TODO: explain how to add a char for "space"
+      - copy text from unifraktur-maguntia/set_new_glyphs_as_unicode.txt
+      - ctrl+end, file > execute script, paste, select FF, OK (this'll take a while)
+      - save (wait a little)
+      - close
+
+
+
+Now we have a directory with the specification of the font that contains
+5000+ prepared glyphs in the first "Private Use Area". In the following
+steps we make a copy of that and inject multi-character pairs into the
+prepared glyphs.
+
+After that, these steps will take a copy of the earlier prepared directory
+and add the glyphs with the character pairs in the pairs file.
+
+- # cd unifraktur-maguntia
+- # python mod_kc_font.py
+- open this directory in fontforge: unifraktur-maguntia/A-OTF-UDShinGoPro-Regular.sfdir
+- file > generate fonts > opentype (cff)
+- close ff
+
+And these steps inject them into the assets file,
+ready for sending to the vita.
+
+- # cp unifraktur-maguntia/UDShinGoPro-Regular.otf d:\vita\kc_translation_mod_candidate\Media\Unity_Assets_Files\sharedassets3\A-OTF-UDShinGoPro-Regular.ttf
+- # cd d:\vita\kc_translation_mod_candidate\Media\
+- # ..\..\unity_tools\UnityEX.exe import sharedassets3.assets
+- copy to vita
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
diff --git a/set_new_glyphs_as_unicode.txt b/set_new_glyphs_as_unicode.txt
new file mode 100644
index 0000000..53341a8
--- /dev/null
+++ b/set_new_glyphs_as_unicode.txt
@@ -0,0 +1,22 @@
+start = 0xE000
+chars_to_add = 5822
+# would prefer to add 6400 and use up the entire PUA,
+# but UDShinGoPro-Regular.otf already has a char at F6BE
+
+# copy this so we can add a shape to the new chars so ff will save them
+Select("#")
+Copy()
+
+# add the new slots
+oldsize = CharCnt()
+SetCharCnt(oldsize + chars_to_add)
+
+# select all the newly added slots
+Select(oldsize, oldsize + chars_to_add - 1)
+
+# add the "e" shape, set the correct unicode value and increase the counter
+foreach
+    Paste()
+    SetUnicodeValue(start)
+    start++
+endloop
diff --git a/translate_utf16_binary.pl b/translate_utf16_binary.pl
new file mode 100644
index 0000000..4f6e645
--- /dev/null
+++ b/translate_utf16_binary.pl
@@ -0,0 +1,150 @@
+use 5.020;
+use strictures 2;
+use IO::All -binary;
+use List::Util 'sum';
+use Encode qw' decode encode ';
+use utf8;
+
+run();
+
+sub get_hits {
+    my ( $content, $jp_enc ) = @_;
+    my @hits;
+    my $pos = 0;
+    while ( ( my $hit = index $content, $jp_enc, $pos ) != -1 ) {
+        push @hits, $hit;
+        $pos = $hit + 1;
+    }
+    return @hits;
+}
+
+sub map_str_to_multi_chars {
+    my ( $tr, %prepared ) = @_;
+    my @mapped;
+    my @chars = split //, $tr;
+    my @pair;
+    while (@chars) {
+        push @pair, shift @chars;
+        next if @pair < 3;
+        my $combination = join "", @pair;
+        my $mapped = $prepared{$combination};
+        if ( !$mapped and $combination ne lc $combination ) {
+            say "combination missing: '$combination', lower-casing";
+            $combination = lc $combination;
+            $mapped      = $prepared{$combination};
+        }
+        if ( !$mapped ) {
+            say "combination missing: '$combination'";
+            push @mapped, encode "UTF-16LE", shift @pair;
+            next;
+        }
+        push @mapped, $mapped;
+        @pair = ();
+    }
+    push @mapped, shift @pair if @pair;
+    return encode "UTF-16LE", join "", @mapped;
+}
+
+sub pad_multi_char_w_spaces {
+    my ( $l_src, $l_tra, $obj ) = @_;
+    my $diff = ( $l_src - $l_tra ) / 2;
+    $obj->{tr_mapped} .= encode( "UTF-16LE", " " ) x $diff;
+    return length $obj->{tr_mapped};
+}
+
+sub map_tr_to_multi_chars {
+    my ( $jp, $obj, %prepared ) = @_;
+    $obj->{tr_mapped} = map_str_to_multi_chars $obj->{tr}, %prepared;
+    my $l_src = length encode "UTF-16LE", $jp;
+    my $l_tra = length $obj->{tr_mapped};
+    $l_tra = pad_multi_char_w_spaces $l_src, $l_tra, $obj if $l_tra < $l_src;
+    die "translation '$jp' => '$obj->{tr}' doesn't match lengths: $l_src => $l_tra, probable char count: " . ( $l_src / 2 ) . "\n"
+      if $l_src != $l_tra;
+    return;
+}
+
+sub trim_nl {
+    my ($s) = @_;
+    $s =~ s/[\r\n]//g;
+    return $s;
+}
+
+sub run {
+    $|++;
+    binmode STDOUT, ":encoding(UTF-8)";
+    binmode STDERR, ":encoding(UTF-8)";
+
+    my %tr = (
+        "選択" => {
+            tr   => "Select",
+            desc => "explains numpad (the skips should be rechecked probably)",
+            ok   => 4562125,
+            skip => [
+                qw( 4487986 4495884 4539834 4539862 4539876 4539906 4548242 4548242 4548404 4550010 4550048 4550486 4552806
+                  4562095 4562119 4495298 4563095 4563131 4563513 4564079 4566677 4570003 4595855 4595927 4596751 4596779
+                  4596809 4596951 4598553 4598719 4598751 4633240 4698799 4698860 4698896 4699107 4699472 )
+            ]
+        },
+        "提督コマンド" => {
+            tr   => "Admiral Cmd",                                                           # quick menu?
+            desc => "usually opens quick menu or changes the buttons to an alternate set",
+            ok   => [ 4518854, 4682604 ]
+        },
+        "戦略へ" => { tr => "Strategy", desc => "leads to fleet strategy management screen", ok => 4518868 },
+        "決定"    => { tr => "Choose",   desc => "select the currently chosen option",        ok => 4518802 },
+        "戻る"    => { tr => "Return",   desc => "back out of the current screen",            ok => 4518826, skip => [ 4566833, 4682617 ] },
+        "らしんばんをまわしてね！" => { desc => "some kind of compass fairy text, using this to find it" },
+        "敵艦隊"                            => { desc => "" },
+        "見ゅ"                               => { desc => "" },
+        "見ゆ"                               => { desc => "" },
+        "被弾回避率補正"                => { desc => "" },
+    );
+
+    my $unicode  = 0xE000;
+    my @pairs    = grep $_, map split( /\|/, $_ ), map trim_nl($_), io("font_mod_character_pairs")->getlines;
+    my %prepared = map +( $pairs[$_] => chr( $unicode + $_ ) ), 0 .. $#pairs;
+
+    for my $jp ( grep !$tr{$_}{tr}, sort keys %tr ) {
+        say "no translation for $jp, skipping";
+        delete $tr{$jp};
+    }
+    map_tr_to_multi_chars( $_, $tr{$_}, %prepared ) for sort keys %tr;
+
+    my $rel_path = "Media/Managed/Assembly-CSharp.dll";
+    my $src      = "../kc_original/$rel_path";
+    my $tgt      = "../kc_translation_mod_candidate/$rel_path";
+
+    io($tgt)->unlink if -f $tgt;
+
+    my $content = io($src)->all;
+
+    my ( %found, $error );
+    for my $jp ( reverse sort { length $a <=> length $b } sort keys %tr ) {
+        my @hits = get_hits $content, encode "UTF-16LE", $jp;
+        if ( !@hits ) {
+            say "no hits at all for $jp";
+            next;
+        }
+        my %obj = $tr{$jp}->%*;
+        $obj{$_} = !defined $obj{$_} ? [] : !ref $obj{$_} ? [ $obj{$_} ] : $obj{$_} for qw( ok skip );
+        for my $hit (@hits) {
+            next if grep $hit == $_, $obj{skip}->@*;
+            if ( !grep $hit == $_, $obj{ok}->@* ) {
+                my $mod = $hit - ( $hit % 16 );
+                my ( $offset, $extract ) = (0);
+                while ( $offset < 3 ) {
+                    $extract = decode "UTF-16LE", substr $content, $hit - 16 + $offset, 32 + 16;
+                    last if $extract =~ /$jp/;
+                    $offset++;
+                }
+                say sprintf "hit $hit %08x %08x for $jp not marked skipped or ok, please verify $jp in >$extract<", $mod, $hit;
+                next;
+            }
+            substr( $content, $hit, length $obj{tr_mapped} ) = $obj{tr_mapped};
+        }
+    }
+
+    io($tgt)->print($content);
+    say "done";
+    return;
+}
diff --git a/unpack_original_files.pl b/unpack_original_files.pl
new file mode 100644
index 0000000..358db29
--- /dev/null
+++ b/unpack_original_files.pl
@@ -0,0 +1,22 @@
+use 5.010;
+use strictures 2;
+use IO::All -binary;
+use Capture::Tiny 'capture';
+
+run();
+
+sub run {
+    my $target = "../kc_original_unpack/Media";
+    io($target)->mkpath;
+    chdir "../kc_original_unpack/Media";
+    my @files = qw( sharedassets2.assets sharedassets3.assets );
+    io->file("../../kc_original/Media/$_")->copy(".") for @files;
+    my $unity_ex = io("../../unity_tools/UnityEX.exe")->absolute->pathname;
+    for my $file (@files) {
+        my ( $out, $err, $res ) = capture { system qq["$unity_ex" export $file] };
+        warn "\n$out" if $out;
+        warn "\n$err" if $err;
+        io($file)->unlink;
+    }
+    return;
+}