[factory] EoB engines' structure investigation - Page 3

JackAsser · 16 July 2006, 12:04

Quote:

Originally Posted by CFOU!

I a finished to resource all INF files of EOB1

and you are right all multi command finished by $EE are Branch (look my asm source joined) but not simple $ee command

not command $EE directly but:
$ff $ee
$ff $f9 $f8 $ee
$ff $f9 $ee
$fb $ee
$fe $ee
etc... and others...

Yes, and according to the JAD-sources I can tell it's stack based.
So for example, some command pushes current XP on stack, another pushes for example if a presure plate is pressed, then an $EE command followed by a sub-code pulls the last two arguments from stack and compare them depending on sub-code, for instance a <= operator. Etc.

Your asm-code, is it complete? I.e. do you know the complete meaning of the byte code?

CFou! · 16 July 2006, 12:23

Quote:

Originally Posted by JackAsser

Yes, and according to the JAD-sources I can tell it's stack based.
So for example, some command pushes current XP on stack, another pushes for example if a presure plate is pressed, then an $EE command followed by a sub-code pulls the last two arguments from stack and compare them depending on sub-code, for instance a <= operator. Etc.

Your asm-code, is it complete? I.e. do you know the complete meaning of the byte code?

$F2 is also an branch command

yes all relative codes has found !! i have already try to modify all text length of level 1 2 & 3

After to have compile & crunch files, i have test it in EOB1 without problem

My ressourcer works with the 12 levels so i have found all EOB1 commands with number of arguments but not the effect for the moment... (I have already replaced EOB WIKI SOurce)

At the end of the file there are a reloc or pointer table

Xyquaatu · 16 July 2006, 13:05

It seems that event.c is mainly for EOB2. Some examples:

interpreting main event code:

EOB1:

Code:

loc_1266_1F1C:			  
		les	bx, [bp+var_inner_elo_pt]
		mov	al, es:[bx]
		cbw	
		mov	si, ax
		inc	word ptr [bp+var_inner_elo_pt]
		or	si, si
		jge	loc_1266_1F4E
		cmp	si, -27		 ; <--- !! 26 cases !!
		jle	loc_1266_1F4E

EOB2:

Code:

lok_ev_main_read_code:
		les	bx, [bp+inf_tr_pt] 
		mov	al, es:[bx]
		cbw	
		mov	si, ax
		inc	word ptr [bp+inf_tr_pt]	
		or	si, si
		jge	lok_ev_main_next
		cmp	si, -31		; <--- !! 30 cases !!
		jle	lok_ev_main_next

There are more differences, for example 0xFE change wall:

EOB1:

Code:

		mov	al, es:[bx]
		cbw	
		cmp	ax, 0FFEAh	  
		jnz	loc_1266_A8B
		jmp	loc_1266_B4E	; 0xEA

loc_1266_A8B:				
		cmp	ax, 0FFF7h
		jz	loc_1266_A93	; 0xF7
		jmp	loc_1266_B64	; other

EOB2:

Code:

		mov	al, es:[bx]
		cbw	
		cmp	ax, 0FFE9h
		jnz	loc_152B_C46
		jmp	loc_152B_D14	; 0xE9

loc_152B_C46:				
		cmp	ax, 0FFEAh
		jnz	loc_152B_C4E
		jmp	loc_152B_D7D	; 0xEA

loc_152B_C4E:				
		cmp	ax, 0FFF7h
		jz	loc_152B_C56	; 0xF7
		jmp	loc_152B_D93	; other code

main event 0xF8 is also different, as you can see by the positioning of text strings

dlfrsilver · 16 July 2006, 19:31

Once finished, how do i recompile them ? ^^

thanks

Quote:

Originally Posted by CFOU!

I a finished to resource all INF files of EOB1

and you are right all multi command finished by $EE are Branch (look my asm source joined) but not simple $ee command

not command $EE directly but:
$ff $ee
$ff $f9 $f8 $ee
$ff $f9 $ee
$fb $ee
$fe $ee
etc... and others...

CFou! · 16 July 2006, 22:05

Quote:

Originally Posted by dlfrsilver

Once finished, how do i recompile them ? ^^

thanks

You must recompile it and pack it

Can you send to me your work in french ? (TExt.dat.asm/ et INF.asm)?

JackAsser · 16 July 2006, 22:12

@CFOU: Not sure if you missed my PM, but do you by any chance have the disassembled Amiga binaries for Eye of the Beholder I? I'd like them for comparison and reference since I'm about to reverse engineer the event byte codes and the byte code interpreter.

CFou! · 16 July 2006, 23:24

Quote:

Originally Posted by JackAsser

@CFOU: Not sure if you missed my PM, but do you by any chance have the disassembled Amiga binaries for Eye of the Beholder I? I'd like them for comparison and reference since I'm about to reverse engineer the event byte codes and the byte code interpreter.

I have send it this morning

i send agin it of you personal email...

CFou! · 16 July 2006, 23:26

Quote:

Originally Posted by CFOU!

I have send it this morning

i send agin it of you personal email...

I send again now

JackAsser · 16 July 2006, 23:36

Quote:

Originally Posted by CFOU!

I send again now

Ahh, got it! Superb!

So, your breaks in the .inf file ReSourcing regarding the event byte codes, are those assumptional, or based on the disasembly? In other words, did you assume number of arguments for each command or did you actually checked in the sources?

Could you please point out were in the code approx the .inf file parser is resided?

CFou! · 17 July 2006, 10:13

Quote:

Originally Posted by JackAsser

Ahh, got it! Superb!

So, your breaks in the .inf file ReSourcing regarding the event byte codes, are those assumptional, or based on the disasembly? In other words, did you assume number of arguments for each command or did you actually checked in the sources?

Could you please point out were in the code approx the .inf file parser is resided?

no for the moment i used more simple method, only function and number of argument are found (is created for translation project so not more inverstigation was needed)

but it's an another step to understand .inf

in the source looking for "btst #" (decrunch / crunch routine and inf decrypter it seems (but to confirme)

i will try to found the start of the code and will give to you the address...

JackAsser · 17 July 2006, 11:09

Quote:

Originally Posted by CFOU!

no for the moment i used more simple method, only function and number of argument are found (is created for translation project so not more inverstigation was needed)

but it's an another step to understand .inf

in the source looking for "btst #" (decrunch / crunch routine and inf decrypter it seems (but to confirme)

i will try to found the start of the code and will give to you the address...

Yep, ok. Based on the jad sources (which indeed seems inaccurate) i've started to understand how those scripts work.

For instance, the script that triggers at position 15,14 in the level1 maze starts at offset $725 (the table at the end points to that).

First it's an EE command which is a conditional, all conditionals are both started and terminated by EE.

So, the first few bytes EE,EF,03,00,FF,EE,34,07,EB,E2,20,03,F7,EF,03 simply means the following:

Code:

EE // Begin conditional
   EF,03 // Get maze flag 3 and push on stack.
   00 // Not sure here, but I think it pushed zero on stack.
   FF // Compare equality
EE // Conditional end
34,07 // If conditional is true, then jump to $0734

EB // Give experience
   E2 // To whole party (i think)
       20,03 // 0x320 xp

F7 // Manage flags
   EF // Set maze flag
      03 // Set maze flag #3

So, in C code:

Code:

push(getMazeFlag(3));
push(0);
if (pop()==pop())
   push(1);
else
   push(0);
if (pop()==0)
    goto _0734;
else
{
    giveExperience(WHOLE_PARTY, 0x320);
    setMazeFlag(3);
}

So, simply those few bytes are the polish stack based notation of a simple:

Code:

if (getMazeFlag(3)!=0)
{
   giveExperience(WHOLE_PARTY, 0x320);
   setMazeFlag(3);
}

So, to my understanding, clicking on those eyes in the first grate you see on level 1 will give you $320 (800) in experience (but only the first time). Then the script continiues and will change the wall set and display a message.

Zetr0 · 17 July 2006, 14:57

heys Jack,

Just a quick note, upon the eob.inf wiki page.

The declaration of 0xfb = wall mapping structure however a little further down the page is a snippet with command codes which state 0xfb = monster create.

I have put a depiction in the upload area for the wall mapping structure for use.

JackAsser · 17 July 2006, 17:22

Quote:

Originally Posted by Zetr0

heys Jack,

Just a quick note, upon the eob.inf wiki page.

The declaration of 0xfb = wall mapping structure however a little further down the page is a snippet with command codes which state 0xfb = monster create.

I have put a depiction in the upload area for the wall mapping structure for use.

Command 0xFB in the Decoration Data area == Define Wall Mapping
Command 0xFB in the Event data == Create Monster

Marcuz · 17 July 2006, 17:30

sorry if i get a lil' off topic, but some of you guys will turn this huge load of info into a nice editor?

JackAsser · 17 July 2006, 19:23

Quote:

Originally Posted by marco pedrana

sorry if i get a lil' off topic, but some of you guys will turn this huge load of info into a nice editor?

Hehe! Not by me at least, but it will hopefully result in a playable version for the old Commodore 64.

Today I did a converted for the wallset data. I still need to squeeze the data some more otherwise I'll never manage to fit it all in those 64kb.

Actually I got surpriced by how good looking the convert actually looks in comparison to other C64 graphics. I'll post some shots tomorrow, including a replacment VCN/VMP for the amiga graphics which effectivly turns the amiga game into a c64-looking game.

wlcina · 17 July 2006, 21:36

wtf they r talking about ?

bx, [bp+var_inner_elo_pt]
mov al, es:[bx]

darn I have to improve my english

Marcuz · 17 July 2006, 21:56

Quote:

Originally Posted by JackAsser

Hehe! Not by me at least, but it will hopefully result in a playable version for the old Commodore 64.

Today I did a converted for the wallset data. I still need to squeeze the data some more otherwise I'll never manage to fit it all in those 64kb.

Actually I got surpriced by how good looking the convert actually looks in comparison to other C64 graphics. I'll post some shots tomorrow, including a replacment VCN/VMP for the amiga graphics which effectivly turns the amiga game into a c64-looking game.

cool screenshots galore

however i really hope someone do the editor thing, i look forward to do some graphic for an hack

CFou! · 17 July 2006, 22:27

Quote:

Originally Posted by JackAsser

Ahh, got it! Superb!

So, your breaks in the .inf file ReSourcing regarding the event byte codes, are those assumptional, or based on the disasembly? In other words, did you assume number of arguments for each command or did you actually checked in the sources?

Could you please point out were in the code approx the .inf file parser is resided?

how can i compile and lauch EoB viewer in java?

I am never coded in JAVa (JDK is needed?)

Zetr0 · 17 July 2006, 23:52

I am going to attempt to make an editor for the PC atfirst

, its been a long while since i last used wb/intuition.

now writing a windows version of uncps (should work for both amiga and pc), eventualy from this you could edit the cps based files. *starting with the inf*

Marcuz · 18 July 2006, 00:06

Quote:

Originally Posted by Zetr0

I am going to attempt to make an editor for the PC atfirst

, its been a long while since i last used wb/intuition.

now writing a windows version of uncps (should work for both amiga and pc), eventualy from this you could edit the cps based files. *starting with the inf*

that's cool! i prefer it windows based as i use windows anyway to do the work, as long it would be able to edit the amiga stuff. but that's my 2 cents.
i look forward to try it!

16 July 2006, 22:12	#46
JackAsser Registered User Join Date: Jul 2006 Location: Lund / Sweden Age: 45 Posts: 171	Amiga disassembled binaries. @CFOU: Not sure if you missed my PM, but do you by any chance have the disassembled Amiga binaries for Eye of the Beholder I? I'd like them for comparison and reference since I'm about to reverse engineer the event byte codes and the byte code interpreter.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Buglist] EoB 1 AGA - EoB 1 ECS Multilanguage	Marcuz	project.Amiga Game Factory	56	20 October 2019 13:17
Little games with script engines builtin	Leandro Jardim	Retrogaming General Discussion	3	07 May 2013 17:36
How do pseudo-3D racing game engines work?	absence	Coders. General	2	29 May 2010 00:05

17 July 2006, 14:57	#52
Zetr0 Ya' like it Retr0? Join Date: Jul 2005 Location: United Kingdom Age: 49 Posts: 9,768	heys Jack, Just a quick note, upon the eob.inf wiki page. The declaration of 0xfb = wall mapping structure however a little further down the page is a snippet with command codes which state 0xfb = monster create. I have put a depiction in the upload area for the wall mapping structure for use.

17 July 2006, 17:30	#54
Marcuz Registered User Join Date: Jun 2002 Location: . Age: 48 Posts: 5,562	sorry if i get a lil' off topic, but some of you guys will turn this huge load of info into a nice editor?

17 July 2006, 21:36	#56
wlcina Registered User Join Date: Aug 2004 Location: Czech republic Posts: 580	wtf they r talking about ? bx, [bp+var_inner_elo_pt] mov al, es:[bx] darn I have to improve my english

17 July 2006, 23:52	#59
Zetr0 Ya' like it Retr0? Join Date: Jul 2005 Location: United Kingdom Age: 49 Posts: 9,768	I am going to attempt to make an editor for the PC atfirst , its been a long while since i last used wb/intuition. now writing a windows version of uncps (should work for both amiga and pc), eventualy from this you could edit the cps based files. starting with the inf

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)