http://www.terrybutler.co.uk/downloads/homm/IFC20.zip

Should be able to drop it straight in the main folder (where the main .exe lies).

Regards.

I have this IFC20.dll in the root Heroes 3 directory.

I take it that this doesn’t exist for you? I’ll upload it if that helps.

Did all the above but the program wont run. I get the following message – IFC20.dll is missing from your computer. Any idea’s?

Cheers

%pngo% %%G

to

%pngo% “%%G”

Otherwise, great tools, thanks.

I’m glad that you found it useful. Good point about doing a send-to on a single file, I can say that I had not considered it

I have posed this question before and have released some stuff but not much interest has been shown so far. I don’t have an awful amount of time for stuff these days so what I manage to get done is often just based on moments of whimsical excitement

I have every intention of improving this one day as well as releasing it as open source–after I clean up some embarrassing crappy code that is –which I’m hoping will also go a long way to improving it.

The extraction process is something I really hoped to improved as it’s somewhat shoddy at the moment!

Best regards
Terry

Will add them now

Best regards friend

I’m also not sure if all of the different games using the Asura Engine use a common “Asra format” or proprietary file formats.

I had a lot of people ask me for features and when I released the code as open-source, there wasn’t much of a buzz, but I at least hoped it’d provide the catalyst to get a few things going either way.

I took a look and I didn’t see anything that was obvious as to what they are. I’d imagine the contents of them are proprietary. I don’t really have the time to be investigating and writing extractors / readers for anything else at the moment though.

I will get around to it eventually, I promise. I was hoping to go through the code and fix up a bunch of stuff first though and remove some stuff.

I will post about it when I do!

That will be Greatful!!

http://www.megaupload.com/?d=THAEKSUY

I doubt it’ll give them the correct name however saying that, so it might still not be 100% what you need, but at least you have some files.

As far as I recall, the application just extracts all files.

I suppose I need to add some file type detection to the application but I’m not sure when I’ll be able to do this.

Once the resource file has been extracted by the app (find it in your User Temp directory if I recall) and run something like Dragon Unpacker’s HyperRipper tool over it.

It should be able to give you a list of known file types.

€ € 




 !$$$&)1$!.) ˆ8KˆNujQ”hu,x…Pl,>s,pY?v?`e&^c/i&?w$oX/~1cw_^o,Iw$rg:Œ:oƒggm&Hw$rg>Œ:olg&˜IX Y–c¥ešX‹`




!&+ &)&$ˆ0V}K}W9?YC1OG>s9Ep$iE??HR?SOSo9XrrHG?HOGGGSz,gƒ$wOUYGHS3>S~7g‰$wOSUUE?GUY¥Xo˜E˜eELHE?HO?

jÚÐÌÐÐÐ%) ,ˆµun}ÂÍEOGOGGUG/ˆ¢À»µÂå]OHH?SGS3ˆ¢š»©íSOGOGU.?5}ÈÓåÓÎÃ?G??.LHS5}ª……q…SG??::S>?Z[¡º‡OHHS5EGH)ˆˆ…€õÓ”ÒÐÒ×À
Îο©µ íӍ‚–
©”m s§ÔÓ´
dõÓµí' jÎíåÔ
Îbn…W

 2#
*' 
  
%)
! 

$&

 
$! &).?,!!
  % 
&&).%% !&1,,, 

%%

&


%3GH9%.5) 
99EH?95.
% &.% %



&LSH)5 

!%
8O[&$$>3
!%%5H8)?GfmS%
% .H\f`SH>.&&5G19G%.UU)
/9S\H)3

+\[&&>Y]YH??%&>Eh>98Wkf`O>L:8O]ie&fR,LkqfO:RRR\Yo?C[U:,EffE8:RK+OW·R> ZR¬a†O€’[/++(++++$/OOSUROOLGPOSPKLE
'
" '
¬WdŸLœZG3?GY?39$+""/?/?\SG9.OOEOG>LLLRGHGORL 

,,'#











SL>K1R>>Eˆ¢Èån”9:G>.E.?1ˆ²ˆ¿}`¢HLLSOLEOEˆˆ¡§CkWuÇ´»ÀÆÀ/

"RK;=<GOOY
395.9H.))%5.%9?)%%!!
!




/
%%

.
.,



. %
$
.1 %,U\\O?3&!%OY?.%G9&%&!3??3$!%) &))..).)
!5OG,.S^crz
5?.13c‰¤­¸¸

).1^—¸ÅÉÉË
3XŒ¸ÉÉÉËÉE11?,UY)$O%Rf`>!!?>3O]E8[)O %&3 )?HhmH$U``.  %!Hmtii|†zzrz„‰—„rz~Ž›¤¦£¾¾ÅÉÏÏÜááçÝÝæèèØÏËÏØÜÞãóøùùúúúüöËÏÖØÜãëóùúüþþþþû*



;

 @#"#"""#-*@2'*;;B@;;;@@@@;-4;@@DMMJ@@@@B@@;;@@@DMWWND@BFDFB@@@@@FQWTMB@@B=4;@@@@@FJDB@@;
 




#-''''"
 ;;@@;;==0B@@DFNT[W=''4;D@DT[]]]WB’42″;B@@FNNQW][D-;@-2J4-4-;M\][F;@@*8




 #"  '' "-"'-;;4-'=("';B;;4@4;2'4QPT64BTD@4;2;'-DKCF@@F[J@4("&%^‘±ÅÉÉËÏÖ &G9))?rž±ÅÉÉÉËÖ3OGG??`ƒ›®¸ÅÉÏØÜ% %HsŒ•±¾ÅÉÕÜÞ,&&)_Ž“›±ÉÏÖÖÜÞ)O1&I~“ž±ÉØÞÞÞã3HH)?‘¦±ÉÖãÞÜãÏØÜÜãëóøùúúüýþþøÖÞéðððöüüþþÿÿÿÿüÞÞéëððóøüþþþÿÿÿüÞéðóðóøúýþÿÿÿÿÿýãëóóóöùùüÿÿÿÿÿÿüããðóööùúýÿÿÿÿÿÿüëóóîèèòùþÿÿÿÿÿÿýðóéË®­±Õòüþþÿÿÿý@6(2;@@@@@B@@;;*2D4'4@;;;4;@;-"24(-;2'##'-'"4@;4-''((((4@@4'';;444@6;@;-(-;@@@B;'#BTD4;6-44;4DQB4;DQB;6D;;@4;BMB44;=MdqhmZF@@*'<2DT[`f[hV=@@-*C;DNTPPW`T6444-C;@BMQW]]Q-4;;6N*;BNNT\`Q–4'AV*@FTWTfV-"'#JQ'*@F[TThV2'FQ;-BT[MQfV*""0M*B@@;<[Q@BCA422@8D4;@Dd[DN`dF--;VF;@@DZ]MNfqQ'';QD6@BDT`WThaC'@A6DD@BZmZMWM=;;BNDDTF@QmVBNNT=4MND6VM@F`Q@FWT-6NFB@DNBBTM@FJ6+JDBBB@;
NTMM=
JT[Q@# QdTJB* ZaFMF# ZFFD- F@BB*"
@@@JF4

!








  

 *

 

8"
!
 Af(

 =T"








E

I gather that you haven’t got the time to learn C#, which I can sympathize with, and you should sympathize with me too, as I haven’t got a lot of spare time either. Either that or you just want me to do the work for you

-S

Yeah, I was thinking that splitting off my rather nonsense framework to a whole separate project on Sourceforge would be quite overkill though.

At the same time, it’s hard to have duplicate files over all of my projects.

This is just one of those no-win situations.

YgNa:

I’ll try and get time to add them. I made this project open source so that others could get involved if they should desire.

Try and get others to help out if you can too

Someone here has a map of AvP 2010?

@Framework
I’d just split it off to it’s own page, SF or otherwise, and have it reference the .dll like the binary does. Then again, I’m honestly more of a scripter than a programmer, so I might be suggesting something that doesn’t make sense.

Time isn’t something I have in large amounts at the moment. I’d be happy to try something like that later, and if I sort the framework issue as stated above, I’d be very happy for somebody else to do it

Best regards,
Terry

Thanks for posting.

I did state that my framework is absent and I need to go about splitting it off. The code is mostly for those curious to see how it works at this point in time.

I’m unsure how I’ll go about posting up the project. I think creating a new Sourceforge project might be a bit overkill, however, it makes no sense to put framework material in the project itself as I don’t want to be taking care of duplicates when I have future projects.

Not sure what the done thing is with this, so I’ve stalled on doing it for now.

Binary is here, by the way:

http://www.terrybutler.co.uk/tools/asura-engine-extractor/

Best regards,
Terry

FNFO, RSFL, RSCF (DDS), RSCF (WAV), RSCF (3D Models), TEXT, TXFL, MTRL, SHPD, SHAP, MTNM, EMOD, STPS, LITE, CRNA, PHON, MRVB, SBSN, NAV1, MLIN, SKYB, WTHR, FOG, AEPR, SMSG, CUTS, CTAC, CTTR, CTEV, CTAT, CTEV, HSKN, HSBB, HMPT, svt1, snav, sesk.
And you can add a Block Number search?
And option to delete blok or add

FNFO, RSFL, RSCF (DDS), RSCF (WAV), RSCF (3D Models), TEXT, TXFL, MTRL, SHPD, SHAP, MTNM, EMOD, STPS, LITE, CRNA, PHON, MRVB, SBSN, NAV1, MLIN, SKYB, WTHR, FOG, AEPR, SMSG, CUTS, CTAC, CTTR, CTEV, CTAT, CTEV, HSKN, HSBB, HMPT, svt1, snav, sesk

And you can add a Block Number search.
I suggest it because i do not understand programming in C#

If you (or anybody else) wants to be added as a developer on the project so you can commit source changes, just let me know.

http://www.terrybutler.co.uk/2011/03/12/asura-engine-now-open-source/

Check it out.

y apreciatte it please

I sourced most of mine for testing purposes only from the MD2 model pack for the modern Doom ports (JDoom and the like).

This is where I’d start if you need some for testing.

Terry

This game has been abandonware for a long time as far as I know, so here is a link to the full versions of Yendor 2 and 3:

http://www.terrybutler.co.uk/downloads/yendor/Yendor2+3-full.zip

Enjoy.

I am dying to play this game and I would really like to know where to find the full version. The demo does not let me go further than a certain point

My code only supports the 32-bit floats as far I recall, so watch out for that if you try and use doubles.

Thanks for commenting!

Tom:

This is one of the projects I’m going to release as open source, meaning that others will be able to contribute to the project, as well as allowing me to leverage the use of existing libraries and code more easily, meaning I’ll be able to provide a MUS player and extractor to MIDI, for example.

Johan:

I’m not sure why this is. I hope that you found a solution.

I’m probably going to clean it up and release it as open source along with a few other projects I have.

No idea when this will happen, however.

I’ll send you an e-mail to the address you’ve posted with, with contact details to some IM clients I have.

Hope to speak to you soon

I do believe there is a native Mac OS for the old PowerPC hardware. I’m not sure how you’d get that running on OS X.

You could also try that “CrossOver Mac” software to get the PC version running. That’s probably the route I’m going to go.

I’ve only just got into OS X myself, so I’m interested to know if you find anything out too

Received your email, as I said I also have most formats documented, and a bunch of C/C++ written tools. However, I didn’t have the time to put all this together, task which I started recently to build some kind of “engine”. About the language used, I made my old GUI tools using C# which I’ve found very productivity boosting but too limiting in terms of portability. As you said Java is an interesting option, it would be easy to port the Homm2 engine to android phones. For now, I’ll code everything in plain C, that way the work can be included in most other languages.

If you are planning to continue working on this topic, I’d like to keep in contact with you. Do you use any IM protocol like gtalk, facebook etc? Anyway, I’m willing to share code and info and maybe we can do some neat thing.

Thanks for replying! Keep up the good work, and take care

I know about fHeroes2. It’s a great project

This page is totally out of date and I’ve not put up half of what I know. I’m just looking for some free time to update it.

I’ve got most of the information on the file formats. I’ve just been woefully bad at updating my Heroes 2 page. I keep meaning to do it, and a thousand other things!

Since this application, I actually did start work on a unified reader for Heroes 1, 2 and 3, although it was never released, it did cover a fair few of the file formats from each game.

Hopefully I get chance to shape everything up over Christmas.

I’ve wanted to write my own Heroes 2 or 3 engine myself regardless of what already exists

Terry

please help me?

About .82m sound files, you just have to append a proper WAV (RIFF) header at the beggining of each file and then they’ll play fine even on windows media player. Check you are using the proper values for each setting (frequency, bitrate etc)

I like to see I’m no the only one which likes researching about this fantastic (great I mean) old game. Now I’m trying to make an H2 remake of mine (I already know about fheroes2 engine, but I want to do it myself)

Many thanks for writing! It’s made my day! ;D

I’m a big fan of the music that you did that ended up in LineWars II, amongst other pieces. Excellent stuff!

Best regards,
Terry

Cheers

Jim / U4ia

i like your projects, it’s possible to share some source code ?

It definitely speeds things up. I’ve been playing with WebGL and furthering the MD2 loader. I’ve currently got it a WebGL context rendering the MD2 with vertex shader interpolation based on some old code I had kicking around.

The load times for such a small model in JavaScript is quite horrendous mind you. I’m sure it takes longer to load the model than it does to download it.. Well, almost

I do enjoy how available JavaScript is. I’m really excited to see WebGL take off.

I’ll hopefully have this WebGL demo I’ve been talking about uploaded soon, I just want to add some more goodies to it

Hope to hear from you soon!

Terry

This is very clever to read the file once and use the buffer. It makes things faster and easier to do since you don’t have to use fread and can work directly with pointers.

About the number of elements, ie accessing another part of the structure is something I have been faced but I haven’t found a proper solution.

What I’ve done in m2 is that ‘struct’ is a function that reads 2 ints: the number of elements and the offset. It jumps to the offset, and reads `n` times the structure given as argument.

I really need to get back on this project as it really makes innovating use of javascript

for example, making Sorceress start with something other than Navigation?

I vaguely recall digging through the main executable for the game and this is what I figured out on my cursory glance at the time.

00..04 - Creature ID (with null terminator e.g. "titn"\0)
5
6
7
8
09..10 - Gold Value
15 -
16 -
17..18 - Hit Points
19 - ?
20 - Speed (6 = Very Fast, 7, = Ultra Fast, 8 = Blazing, 9 = Instant)
21 - attack or defence
22 - attack or defence
23 - Speed Min
24 - Speed Max
25 - Number of shots

It’s pretty bare information, but you should be able to find reference to “titn”, which for me lies at offset 985243 (0xF089B) for Heroes II Gold v2.1.

I might pick this up again soon if I can find the time. I also can’t verify that it’s 100%, but I would have modified data and then checked the results in-game, so it should be okay.

Taking this section of the ‘m2′ structure:

1

vertices: ['struct', 'vertex'],

If you’ve been given a number of expected vertices in some header, how would you match that up so the binaryParser can read it properly?

I would have expected to have seen something like this:

1

vertices: ['struct', 'vertex', numVertices],

With numVertices having been populate earlier on to maintain this recursive, elegant solution you’ve said about?

I know that I’m probably just missing something!

You are right, execution speed is quite honestly the crippler for these elegant methods and it makes sense to keep these lower level languages for when we need high performance at the expense of simplicity perhaps.

Seeing as you gave me some knowledge on JavaScript, I hope I can repay the favour by giving a quite example of what I believe to be, In my humble opinion, some elegant C code

The code below is an excerpt from a Quake 1 MDL loader. It took a lump of raw memory and didn’t use freads, but instead, casts raw memory to desired structures, which you can’t get much faster than. It literally takes a lump of raw memory and uses the structures as buckets to dump that raw memory straight over, which virtually involves no processing logic.

This is the preferred version of loading stuff in C for me as I feel it has a greater degree of extensibility, reduces disk reads and just generally speaking, is a hell of a lot faster.

Here’s where I load the entire file to memory from disk:

1
2
3
4
5
6
7
8
9
10

// Get the length of the file
fseek(fp, 0, SEEK_END);
fileLength = ftell(fp);
fseek(fp, 0, SEEK_SET);
 
// Allocate memory for buffer with size of fileLength
char * buffer = new char [fileLength];
 
// Read the entire file to memory
fread(buffer, sizeof(char), fileLength, fp);

I have my data structures, the first structure (mdl_vertex_t) is how Quake stores them and the second structure (mdl_vertexReal_t) are the uncompressed values which are valid for rendering. Carmack compressed the vertices to save space.

1
2
3
4
5
6
7
8
9
10
11

struct mdl_vertex_t
{
	unsigned char v[3];
	unsigned char normalIndex;
};
 
struct mdl_vertexReal_t
{
	Vertex3f_t v;
	float normalIndex;
};

Note: Please take for granted in the below example that buffer which references our entire file in memory, its pointer has been moved along during the loading process and now lies just before the vertices data.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

// Cast "buffer" to verts, so that verts as a pointer lays straight over our raw buffer Notice that we 
// haven't actually defined a size for "verts", but memory is unmanages and all pointers "bleed" into 
// neighbouring memory! Scary business, but it is nice and fast!
mdl_vertex_t * verts = (mdl_vertex_t *)buffer;
 
// We can now seek buffer along equal to the size of a mdl_vertex_t structure multipled by the number of vertices
buffer += sizeof(mdl_vertex_t) * mdl->numVerts;
 
// Allocate some proper memory to store our "real" (i.e. uncompressed) vertices
this->vertices = new mdl_vertexReal_t[mdl->numVerts];
 
// Loop through and uncompress vertices
for (int i = 0; i < mdl->numVerts; i++)
{
	vertices[i].v[0] = (mdl->scale[0] * verts[i].v[0]) + mdl->translate[0];
	vertices[i].v[1] = (mdl->scale[1] * verts[i].v[1]) + mdl->translate[1];
	vertices[i].v[2] = (mdl->scale[2] * verts[i].v[2]) + mdl->translate[2];
}

I think this is quite simple and elegant as well as lightning fast!

I hope that you found that interesting anyway

http://tfc.duke.free.fr/coding/md2-specs-en.html

If you look at the code to read the model, he does some fread, but the rest of the code is boring stuff like doing loops, making checks, reading ints and jumping … All this is already in the structure. You are basically writing things twice.

It’s okay when your structure is simple. However when working with models that got like 300 rules of structure, the parsing code is about 1000, 1500 lines of pure repetition. Which makes it really bug-friendly and hard to maintain (because you don’t want to touch it!)

I’ve been studying the possibilities in C. Since there is no introspection you can’t really do beautiful things, so you can use macros to simplify things a bit, or write a parser generator. But both are quite ugly :p

The downside of my approach however is execution speed. Since it’s using a lot of generic stuff, it’s really slow. Parsing a 2meg model takes roughly 2 seconds on chrome and 10 on firefox.

(The method isn’t the only problem, Javascript is dynamic and therefore slower than dedicated C code, the BinaryReader should be written in C and using processor convertion instructions.)

Anyway the goal was just to make a proof of concept and try to use Javascript powerfulness to solve this very specific problem.

I was also really amazed by the power given compared to the size of the code. The parsing function takes only 30 lines, and about the same amount for the standard lib (basic types: int32, … and arrays)

Grrr I really need to debug the demo and blog about it

I do disagree with the C comment with regards to it being overly cumbersome due to the lack of options. I preferred the old style of dumping a lump of raw memory over the top of an expected structure. When I started using C# quite heavily, I had to get used to the memory marshalling to try and emulate what I used to do in C/C++

1
2
3
4
5
6
7
8
9
10
11
12

struct md2Header
{
	int ident;
	int version;
	int skinWidth;
	int skinHeight;
	// etc
}
 
md2Header		header = { 0 };			// Model header
 
fread(&header, sizeof(md2Header), 1, fp);

After defining the structure and then doing the above, you’ve got access to the header’s information quite simply.

I still agree that when you’ve got there with the binaryParse is also very elegant though It also introduces me to some new techniques that I haven’t yet covered in JavaScript

I made the BinaryReader class to do a loader like yours but for World of Warcraft models. It has never been finished but I’ve done some work that may interest you.

When parsing binary files, you have to do boring stuff like

1
2
3
4
5
6
7

define the structure:
        var skinwidth;
        var skinheight;
 
parse the structure:
    header.skinwidth        = reader.readInt32();
    header.skinheight       = reader.readInt32();

and this for every field in the model. this is repeating stuff that could be automated. In C this is not possible because the language lacks of a good toolset. But in Javascript we can do it!

I’ve written a small lib called BinaryParser (that uses the Reader) http://fooo.fr/~vjeux/blog/armory/scripts/binaryParser.js which allows you to write clean code. Here’s an example of the code to read the m2 (WoW) models ( http://fooo.fr/~vjeux/blog/armory/scripts/m2.js )

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

 
		nofs: {
			count: 'uint32',
			offset: 'uint32'
		},
 
		animationBlock: {
			interpolationType: 'uint16',
			globalSequenceID: 'int16',
			timestamps: 'nofs',
			keyFrame: 'nofs'
		},
 
		color: {
			color: 'animationBlock',
			alpha: 'animationBlock'
		},

And to parse the model you just have to call BinaryReader.parse(description, file).

And all the values are either defined structures that are recursively parsed (alpha of color is calling animationBlock that is calling nofs that is calling unit32) or custom functions.

For example in the m2 there are structure headers at the beginning of the file that contains the offset and the number of elements. Therefore we can make a function to handle this

1
2
3
4
5
6
7
8

		struct: function (config, type) {
			var nofs = config.parse('nofs');
			var pos = config.binaryReader.getPosition();
			config.binaryReader.seek(nofs.offset);
			var result = config.parse(['array', type, nofs.count]);
			config.binaryReader.seek(pos);
			return result;
		},

And use it in the code

1
2
3
4

		m2: {
			globalSequences: ['struct', 'uint16'],
			animations: ['struct', 'animation'],
			animationLookup: ['struct', 'uint16'],

In the end the code is muuuucchhh more cleaner than the version I wrote in Delphi at the time for the war3 model parser.

However this has never been finished and i’ve been confronted to many conception problems such as how to make it generic enough to support a wide range of cases. How to you allow the functions to call back the parser … This version is not perfect at all, I haven’t handled the error parsing for example, but I am pretty sure you can grab some ideas for your future projects

http://fooo.fr/~vjeux/blog/armory/modelviewer.html

[I am pretty sure all the code will be unreadable due to the lack of monofont text, please can you edit it back, thanks)

In the original _decodeFloat function, it’s slightly overcomplicated and hard to debug due to it not really splitting the calculation of the Exponent and Mantissa and performing a lot of logic at the end to split it and join it again.

I’m curious to see what the error is _decodeFloat, but I’m also happy to replace it too

Thanks for visiting!

I just saw that you used my binary reader refactor, i’m really glad you like it. Do you have example of wrong values being parsed?

I took the math lib on a random website and it was unreadable. I tried to make it cleaner as I could but have no idea how it works behind. Maybe I can find the issue

javascript: (function (){var newSS, styles = ‘* { border:none !important; background: black ! important; color: #aaa !important; } b,strong,h1,h2,h3 {color:white !important} input, button, textarea { border:1px solid blue !important; } a:link, a:link * { color: yellow !important; text-decoration: none !important; }a:visited, a:visited * { color: #a00 !important; }a:hover, a:hover * { color: red !important; } * pre { color:white !important; } ‘;var elemHead = document.getElementsByTagName(“head”)[0];var elemCSS = document.getElementById(“darkenCSS”);if (elemCSS){elemHead.removeChild(elemCSS);}else{newSS = document.createElement(‘link’);newSS.rel = ‘stylesheet’;newSS.href = ‘data:text/css,’ + escape(styles);newSS.id = “darkenCSS”;elemHead.appendChild(newSS);}})();

javascript: (function (){var newSS, styles = ‘* { border:none !important; background: black ! important; color: #aaa !important; } input, button, textarea { border:1px solid blue !important; } a:link, a:link * { color: yellow !important; text-decoration: none !important; }a:visited, a:visited * { color: #a00 !important; }a:hover, a:hover * { color: red !important; }’;var elemHead = document.getElementsByTagName(“head”)[0];var elemCSS = document.getElementById(“darkenCSS”);if (elemCSS){elemHead.removeChild(elemCSS);}else{newSS = document.createElement(‘link’);newSS.rel = ‘stylesheet’;newSS.href = ‘data:text/css,’ + escape(styles);newSS.id = “darkenCSS”;elemHead.appendChild(newSS);}})();

If you haven’t installed the entire .NET framework, that might be the reason. It’s probably missing parts of it or isn’t configured right.

If you’ve installed the whole thing and it’s having issues, I’m quite honestly not sure what the problem is.

Let me know either way.

Best regards,
Terry

Try searching for the names of the creatures and look for numbers that match their statistics. I’ll try this myself if I get time and report on my HoMMII page

i want to balance the game out as certain aspects are fairly strong when used in certain ways..

I need to investigate the files and see what they are sometime.

Is there any possible way to import AvP3 models to 3DS MAX?

Thanks in advance.

Cheers.

# Asura engine (RSCF chunks) (script 0.2)
# Sniper Elite
# Rogue Trooper
# Guard Shield
# Possibly other Rebellion’s games
# Script for QuickBMS http://aluigi.org/papers.htm # quickbms

get PACK_SIZE asize
getdstring SIGN 8
SIGN f == “AsuraCmp”
ZSIZE get long
get long SIZE
math ZSIZE = PACK_SIZE
math ZSIZE -= 8
comtype huffboh
clog TEMPORARY_FILE 8 ZSIZE SIZE
open “.” TEMPORARY_FILE
get PACK_SIZE asize
getdstring SIGN 8
elif SIGN == “AsuraZlb”
get ZERO long
ZSIZE get long
get long SIZE
savepos OFFSET
math ZSIZE = PACK_SIZE
math ZSIZE -= OFFSET
comtype zlib
clog TEMPORARY_FILE OFFSET ZSIZE SIZE
open “.” TEMPORARY_FILE
get PACK_SIZE asize
getdstring SIGN 8
endif
if SIGN! = “Asura”
cleanexit
endif

for
savepos CHUNK_OFFSET
math TMP = CHUNK_OFFSET
math TMP + = 16
if TMP> = PACK_SIZE
cleanexit
endif
getdstring chunk 4
CHUNKSZ get long
DUMMY get long
DUMMY get long
if chunk == RSCF
TYPE get long
DUMMY get long
get long SIZE
get string NAME
math OFFSET = CHUNK_OFFSET
OFFSET + math = CHUNKSZ
math -= OFFSET SIZE
log OFFSET NAME SIZE
# else
# print “Chunk% CHUNK_OFFSET chunk%%%%% CHUNKSZ”
# savepos OFFSET
# math SIZE = CHUNK_OFFSET
# math SIZE + = CHUNKSZ
# math SIZE -= OFFSET
# log “OFFSET SIZE # comment here if you do not want to extract these useless chunks
endif
CHUNK_OFFSET + math = CHUNKSZ
goto CHUNK_OFFSET
next

Sorry for the late response!

Great project, my friend! I’m going to go out on a limb here and say you’ve gone far beyond my amateur experimentations! That looks fantastic

I want to start posting up some source, but none of it is robust enough right now and I’d hate to give false information to anybody.

I’ll probably start with posting source code clips in posts that I can make sure are robust enough to be informative.

My ray caster in XNA merely did a standard ray cast (I actually copied verbatim my entire function from a C# GDI project version) and then rendered simple lines to the screen for the columns with the appropriate colour (or appropriate texture slice).

Nothing too impressive at the moment I’m afraid. The ray casting algorithm itself is very much mundane also.

Best regards,
Terry

by the way, I also have been working on a tiny raycast engine you can check it out here: http://wolf3dx.codeplex.com

The full source code is available.