It seems that you're using an outdated browser. Some things may not work as they should (or don't work at all).
We suggest you upgrade newer and better browser like: Chrome, Firefox, Internet Explorer or Opera

×
Here you go, guys

https://youtu.be/N2DVKj8lBeg
avatar
stryx: I have installed the game freshly from the Masterpiece Collection CD, patched it to version 1.1 and the game crashed during the final cutscene. See the attached screenshot.
avatar
SweetieViaPony: Hi, what's that scene there? I don't know Stone Prophet yet. I've only tested that the movie of Ankh vs. the Heirophant plays. Sorry, haven't don't the game before. Return to city?

---

Found my way out of temple. Got two more movies to play, one while leaving, then one as I ran to the town. I didn't see the guy saluting.
Once you left the buried temple of Ra turn to the east and head into the (now defunct) wall of Ra, till you hit an invisible wall. Then head to the north, where a skeleton lies. Pick up special scroll of return, put it into a character's hands and use it. This triggers the end game sequence, which will then produce the error.
avatar
stryx: When I install the game from the Masterpiece Collection CD (which I just did today) it is version 1.0, that's why no rollback patch is needed. I couldn't find one on my disk either.

I downloaded the 1.1 patch from the internet and applied it manually. And kept a back-up of the old 1.0 version of the game just in case.
avatar
rmeakins: Must have been several different distributions! I got mine in Australia so may have been different from the US-published edition?
That could very well be the case.
Post edited February 09, 2020 by stryx
Okay, with you again. Thank you for leading me. I was wandering all over that temple until I realized I just had to turn around and walk straight... When I previously thought I beat the game, it was because a timer ended. The credits did roll, but I didn't realize that was the bad ending.

I've found the scroll now and replicated the bug with the kid. Seems to be that his audio files aren't loading. The error is somewhere in Code.2. I'll start digging into it later tonight.
Best I have so far is a 1.1 version that is only missing one movie, the one with the little boy talking to you. By bypassing that cinema, I've been able to make the game go right from him running up to the party entering the portal. Something in Code.2 is making an invalid call, I believe to URCHSPK.DFA, which is the animation for him talking. I'm unable to make any sense out of much of the hex for Code.2, I believe it's compiled.
avatar
SweetieViaPony: Best I have so far is a 1.1 version that is only missing one movie, the one with the little boy talking to you. By bypassing that cinema, I've been able to make the game go right from him running up to the party entering the portal. Something in Code.2 is making an invalid call, I believe to URCHSPK.DFA, which is the animation for him talking. I'm unable to make any sense out of much of the hex for Code.2, I believe it's compiled.
Well, I guess it's time to bring out the debugger and disassembler, and brush up on your knowledge of x86 assembly language.
So close. I guess if it's compiled, there is no easy way to analyse the pre-patched version of code.2 and the post-patched version. It would be pretty cool if it was possible though. Kudos for the attempt :)
avatar
rmeakins: So close. I guess if it's compiled, there is no easy way to analyse the pre-patched version of code.2 and the post-patched version. It would be pretty cool if it was possible though. Kudos for the attempt :)
Well, you could look for differences in the files, just like any other file. You could also disassemble the files and look for differences in the disassembly.
I did spend hours last night looking for differences, values that would create 71/6, calls to urchspk... no luck. Most of the files values are under a head called code that makes up 90% of the file and are almost completely different. If I could figure out how to decompile the things, yes, I think progress could be made.
avatar
SweetieViaPony: I did spend hours last night looking for differences, values that would create 71/6, calls to urchspk... no luck. Most of the files values are under a head called code that makes up 90% of the file and are almost completely different. If I could figure out how to decompile the things, yes, I think progress could be made.
One other possible avenue... would it be possible to alter URCHSPK.DFA so that the invalid call from code.2 then became valid?
avatar
SweetieViaPony: I did spend hours last night looking for differences, values that would create 71/6, calls to urchspk... no luck. Most of the files values are under a head called code that makes up 90% of the file and are almost completely different. If I could figure out how to decompile the things, yes, I think progress could be made.
One idea is to get a debug build of DOSBox and set a breakpoint whenever the game calls DOS to open a file, check the filename being opened, and see when the file containing the video is opened. Once you reach that point, you can try stepping through the machine code to try to figure out what's going on.

As I said, you might need to brush up on your knowledge of x86 assembly language, as you'll be working with machine code. DOSBox's debugger will disassemble it for you, but converting it into a higher level language can't be easily done. Also, you might need to look at how DOS interrupts work, specifically how you call DOS from user programs.

(Also, what CPU mode is the game running in? It might help to figure this out.)
Debugs and Assembler are way above my knowledge level, I'm afraid. I've just learned what few tricks I've managed to teach myself over the years. I look at the files, switch them around, delete them, muck with their bytes, and see what happens.

Just thinking aloud here: There's one thing about the file structure I don't get. If you look at the CINE folder, you see every cinema has two files. DFA for video, ACT for audio. Except for four. WINCAP1-3 and URCHSPK. They only have DFA files. But we can see that URCHSPK and the WINCAPs do indeed have audio. It plays in version 1.0, and WINCAP also plays if we bypass URCHSPK in 1.1. So where is the audio being stored and why was the programmers convention for the audio changed for four of the five final videos of the game?
avatar
SweetieViaPony: Debugs and Assembler are way above my knowledge level, I'm afraid. I've just learned what few tricks I've managed to teach myself over the years. I look at the files, switch them around, delete them, muck with their bytes, and see what happens.

Just thinking aloud here: There's one thing about the file structure I don't get. If you look at the CINE folder, you see every cinema has two files. DFA for video, ACT for audio. Except for four. WINCAP1-3 and URCHSPK. They only have DFA files. But we can see that URCHSPK and the WINCAPs do indeed have audio. It plays in version 1.0, and WINCAP also plays if we bypass URCHSPK in 1.1. So where is the audio being stored and why was the programmers convention for the audio changed for four of the five final videos of the game?
A long time ago I had some old generic dos tools (MRIP I think?) for extracting other files from packaged game files, if I find some time, I might go exploring to find the location of the audio.
I tried the generic extraction tool and it found a bunch of midi files in RES5, but not much else.

I did find this however - looks like someone has done a lot of the advance work for the file format.
https://wiki.multimedia.cx/index.php/DFA
Thank you, sir. I hope i fixed my arrow damage numbers with quick hex edit =)
Killed my first wolf with arrows only - so its working after all