Race Fonts (*.fnt files)

[Benman2785]

here is the correct tga with correct headers

working with UE 0.9.0dev3

EDIT: newer version available

[Benman2785]

Feds have different ÄÖÜ - but all are working with preedited 1366x768 wdf (ship type on galaxy map)

OLD version

12px_fonts.jpg (637.11 KiB) Viewed 1848 times

[Benman2785]

i fixed some issues with the 12px fonts

NEW VERSION - added % (for Intercept)



1. no more black bars (simply needed to edit 3px to be transparent)
this needs to be the upper right corner of tga

transparent.gif (95 Bytes) Viewed 1821 times

2. added some often used chars like . , - %

3. fixed the A and ÄÖÜ (i prefer this style)

here the result:

12px_fonts_fixed.jpg (699.05 KiB) Viewed 1821 times

[Spocks-cuddly-tribble]

Looks nice and clean!

[Benman2785]

here i edited the 12px and 9px fonts - now containing "ß" needed in german

**outdated**

in 9px i am not sure if "%" is placed correctly

hu9.gif (1.17 KiB) Viewed 1499 times

hu12.gif (909 Bytes) Viewed 1498 times

[thunderchero]

in 9px i am not sure if "%" is placed correctly

if it displays correctly it is placed correctly, the header info determines where it needs to be placed iirc.

[Benman2785]

now i finished modding 9px and 12px // *updated*

BotF_9px_12px_fonts_+_tga.zip

(23.63 KiB) Downloaded 54 times

here is the template i use for all (300% enlarged)
9px

hu9.gif (5.56 KiB) Viewed 1438 times

12px

hu12.gif (4.42 KiB) Viewed 1402 times

[Benman2785]

i forgot to mod generic9.fnt - now it can be used for german as it has "ß" (seen on Pakled next to Cardassian space)

generic9.fnt.zip

(2.48 KiB) Downloaded 48 times

generic9.jpg (1012.27 KiB) Viewed 1396 times

[thunderchero]

While testing font files I noticed;

{ (opening CURLY BRACKET) will cause CTD if in star name. other locations (lexicon, ect untested)
crash log

Code: Select all

File: textdef.c, Line: 156, ptr == NULL, this means there was no data in the embedded string

} (closing CURLY BRACKET) will not display but no CTD

note; this is unrelated to Benman2785 files

[Spocks-cuddly-tribble]

{ (opening CURLY BRACKET) will cause CTD

sub_51EEB0 checks hex code for { 0x7B (only, other switches use jnb), might be a bug. But do we use/need fonts for this?

Code: Select all

0051EEBF           cmp   byte ptr [eax], 7Bh  // hex code for { opening CURLY BRACKET
0051EEC2           jz    short loc_51EECE

[thunderchero]

{ (opening CURLY BRACKET) will cause CTD

sub_51EEB0 checks hex code for { 0x7B (only, other switches use jnb), might be a bug. But do we use/need fonts for this?

Code: Select all

0051EEBF           cmp   byte ptr [eax], 7Bh  // hex code for { opening CURLY BRACKET
0051EEC2           jz    short loc_51EECE

UE could prevent symbol from being added and prevent CTD

[Spocks-cuddly-tribble]

UE could prevent symbol from being added and prevent CTD

I'd first try (might even display CURLY BRACKETs?):

Code: Select all

0051EEB0           push  ecx   // -> retn   // disable sub_51EEB0

Or nop the jz above (no crash, but no display either?).

[thunderchero]

After more testing code dose not like any type of brackets, no other crashes but displayed no transparences or text stopped displaying after bracket.

repost of new header info
0x000 height of font +1
0x001 1 (all others) or 0 (9 fnt)
0x002 - 0x003 260 start of tga file
0x004 - 0x005 10 (unknow/unchanged)
0x006 - 0x007 120 total blocks
0x008 - 0x009 width of tga
0x00a - 0x00b height of tga
0x00c - 0x00F 47354 tga sile size in bytes
0x010 - 0x011 1536 (unknow/unchanged)
0x012 - 0x013 86 (unknow/unchanged)
0x014 - start of row 1 (10 entries) 1 - 0
0x028 - start of row 2 (26 entries) A - Z
0x05c - start of row 3 (26 entries) a - z
0x090 - start of row 4 (32 entries) symbols
0x0d0 - start of row 5 (26 entries) German and specials symbols

note; first entry of each row is font area width only, all others entries in that row are font area width +1

[thunderchero]

here is more odd info on fnt files

here is text of all numbers, letters and symbols as in fnt files

Code: Select all

1234567890
ABCDEFGHIJKLMNOPQRSTUVWXYZ
abcdefghijklmnopqrstuvwxyz
`-=\[];',./~!@#$%^&*()_+{}|:"<>?
äâàêèéïîöôüñÇçÄÖÜß®©™…–¿¡•

here is an example of new 12 fnt Benman2785 and I have been working on

fnt.jpg (87.4 KiB) Viewed 1321 times

% does not display/ignored anything before or after displays with no space when used in lexicon
… UE changes hex 85 -> 3F (?), if hex edit file anything before this symbol displays, anything after this symbol does not display
– UE changes hex 96 -> 3F (?), if hex edit file anything before this symbol displays, anything after this symbol does not display
• UE changes hex 95 -> 3F (?), if hex edit file anything before this symbol displays, anything after this symbol does not display
™ UE changes hex 99 -> 3F (?), if hex edit file anything before this symbol displays, anything after this symbol does not display
® symbol does not display, anything before this symbol displays, anything after this symbol does not display
© symbol does not display, anything before this symbol displays, anything after this symbol does not display

Important; when adding font to tga file the top left of tga and each font area is the transparency color.

topleft.jpg (103.42 KiB) Viewed 1314 times

[Flocke]

late response, but...

% does not display/ignored anything before or after displays with no space when used in lexicon

'%' is an escape character used for text placeholders like %s (string), %c (character), %d (decimal), %f (flloat). To actually print a '%' symbol, it must be doubled "%%".

… UE changes hex 85 -> 3F (?), if hex edit file anything before this symbol displays, anything after this symbol does not display
– UE changes hex 96 -> 3F (?), if hex edit file anything before this symbol displays, anything after this symbol does not display
• UE changes hex 95 -> 3F (?), if hex edit file anything before this symbol displays, anything after this symbol does not display
™ UE changes hex 99 -> 3F (?), if hex edit file anything before this symbol displays, anything after this symbol does not display

By documentation of java.nio.charset.Charset "The native character encoding of the Java programming language is UTF-16."
BOTF however does not support unicode, but uses some variant of ISO 8859-1. DCER therefore opted to apply the lava.lang.string.getByte conversion with java.nio.charset.StandardCharsets.ISO_8859_1 encoding, the only standard charsets ISO_8859_1 variant listed.

The characters above however are not covered by ISO 8859-1, which is why they are auto-replaced with 3F = "?" by lava.lang.string.getBytes.
Most likely BOTF actually uses the Windows-1252 character encoding:
https://de.wikipedia.org/wiki/Windows-1252
But maybe it is just the .fnt files using it, which would explain why BOTF fails to display these characters.

Since Windows-1252 is an extension to ISO 8859-1 and there is no use in replacing these special characters with '?', I fixed UE to now use Windows-1252 for text encoding.

® symbol does not display, anything before this symbol displays, anything after this symbol does not display
© symbol does not display, anything before this symbol displays, anything after this symbol does not display

These are part of both ISO 8859-1 and Windows-1252. UE converts them to:
® = 0xAE
© = 0xA9
They should be working, unless BOTF uses some weird character encoding variant.