Imploder compression algo

[StingRay]

They need code for compression, not for uncompressing

[Exl]

Yeah, there's plenty of Imploder decompression routines about, so I was able to use one of them. Thanks though hit.

[jotd]

I have made more progress. The output data is still corrupt, but not very much and I have a lead. So I think this will work very soon.

@Exl: do you still need this, or did you find some imploder packer written in C? I did not understand your post.

[Exl]

Oops, I meant decompression routines in my previous post. I still need it yeah. I've been working on the editor, and it's almost finished. Save for some polishing and some Imploder crunching

[jotd]

Same here. Imploding in C almost works. I can pack a small file 100%, but for bigger files it still does not work 100%. the file is decrunched with errors. Give me a few days and I'll deliver a full version of fimp with C source.

[gemanix]

any of these help?
http://www.amiga-stuff.com/crunchers-download.html

[Exl]

You can download Amiga crunchers from it, yeah. We're looking for\Jotd is engineering a cruncher for Imploder in C, so that it can run on pretty much any platform.

[gemanix]

Originally Posted by Exl

You can download Amiga crunchers from it, yeah. We're looking for\Jotd is engineering a cruncher for Imploder in C, so that it can run on pretty much any platform.

Oh, I see... I might see what I can do

[jotd]

Almost there...

[Exl]

I can't wait

[gemanix]

Originally Posted by gemanix

Oh, I see... I might see what I can do

i looked into it. while i believe i could do it, the number of lines makes this is a very time consuming exercise (5 hours+). so i'm leaving it alone, sorry.

[jotd]

rewriting compression code without knowing the algorithm at all is very difficult. The testing and bughunting is very hard because we don't know functionnally how it works.

68000 -> C translation is better because it can be adapted to others "blackbox" algos as well.

@Exl: ok, I think it works. Get fimp C version from here, test and extract the implode routine for your proggy:

http://pagesperso-orange.fr/jotd/

bye

[gemanix]

seen here?
http://www.dstoecker.eu/xpklibs.html#IMPL

[jotd]

Originally Posted by gemanix

seen here?
http://www.dstoecker.eu/xpklibs.html#IMPL

I dunno, but if it's here, it's probably 68000 source code. We need C code.

BTW Exl I tried to run the editor but it cannot find "renderlib". It does not work for me.

[Exl]

Originally Posted by jotd

I dunno, but if it's here, it's probably 68000 source code. We need C code.

That is one of the first things I came across when looking for cruncing code, but it's assembly as well as decrunching only.

BTW Exl I tried to run the editor but it cannot find "renderlib". It does not work for me.

It'll require the Visual C 2008 runtimes, unfortunately. See http://www.microsoft.com/downloads/d...displaylang=en

I tested your fimp port with Superfrog and it works perfectly I'll put it in the editor today and finish some last bits. A lot has changed since the 0.5 version that I posted. Thanks a million, it's impressive work you've done.

[jotd]

still did not work because MSCOMCT2.OCX was missing.

After adding this file in the current directory it worked. I suggest that you include this file in the editor distribution and that you document the fact that it requires the VC 2008 runtime.

I'm glad the compression routine works. I can't wait to play the custom levels. How many did you create?

[gemanix]

Originally Posted by jotd

still did not work because MSCOMCT2.OCX was missing.

After adding this file in the current directory it worked. I suggest that you include this file in the editor distribution and that you document the fact that it requires the VC 2008 runtime.

I'm glad the compression routine works. I can't wait to play the custom levels. How many did you create?

well done jotd. i downloaded your code. the conversion looked complicated. impressive!

[Exl]

Originally Posted by jotd

still did not work because MSCOMCT2.OCX was missing.

After adding this file in the current directory it worked. I suggest that you include this file in the editor distribution and that you document the fact that it requires the VC 2008 runtime.

My current version uses GCC, so that'll no longer be necessary. When it's done I'll make a small installer for the other shared files.

I'm glad the compression routine works. I can't wait to play the custom levels. How many did you create?

I haven't really modified levels much yet because testing them is tiresome (copy to Amiga HD directory, compress in WinUAE from a shell, hex edit, copy back, etc.). But it should be a piece of cake now.

[jotd]

@germanix: it's not because it is complicated that it is impressive The cool thing about it is that any 68000 algo can be effortlessly converted now. Don't read "any game can be converted". Only sections of code manipulating the memory. I don't intend to rewrite an emulator!

@Exl: for the ATN compression, you could have mounted your data directory in WinUAE directly instead of copying/copying back, and make a modified version of fimp to output ATN header instead of IMP using an hex editor . But you're right, it's better with the native compression module (which should work on a big endian architecture BTW but I haven't tested that yet), and even better with gcc instead of msvc.

[Exl]

Originally Posted by jotd

@Exl: for the ATN compression, you could have mounted your data directory in WinUAE directly instead of copying/copying back, and make a modified version of fimp to output ATN header instead of IMP using an hex editor . But you're right, it's better with the native compression module (which should work on a big endian architecture BTW but I haven't tested that yet), and even better with gcc instead of msvc.

Yeah mounting is what I meant, but it's still a pain

Anyway, here it is: Installer with shared components, some accompanying text (readme + FAQ), and source code!

Installer: http://members.home.nl/meuwissenth/f...etup_0.9.9.exe
Source: http://members.home.nl/meuwissenth/f...urce_0.9.9.zip