Note: Occasionally, my blog is a sober and objective account of the historical development of RPGs, and occasionally it's a very subjective and visceral account of my particular experience with a game. This post, in its original form (below) was one of the latter. As we discuss in the comments, there clearly were endgame graphics and text in the game's directory, so it's a mystery why they didn't trip when I defeated the Dark Lord. Perhaps I was supposed to defeat him some other way; perhaps there were issues with this (unofficial) English translation; perhaps there were other associated emulator issues. Whatever the case, it is clear that the issue is not that the developers didn't program an endgame and deliberately made the final battle impossible to hide that fact. When I asked that question below, it was never meant to be a serious accusation; just a reflection of the frustration I was feeling at the time. Nonetheless, I apologize for even raising it as a possibility without more information.
*****
Commenter dahauns wrote and offered to figure out the hex editing if I could send him the saved game file. I did, and he had it back to me within a couple of hours. All my characters had 255 in all attributes, and their hit points and spell points were close to 1,000.
*****
Commenter dahauns wrote and offered to figure out the hex editing if I could send him the saved game file. I did, and he had it back to me within a couple of hours. All my characters had 255 in all attributes, and their hit points and spell points were close to 1,000.
 |
| Super-Gideon. |
I re-engaged the combat and, predictably, found it not too difficult. My characters were able to withstand the trolls' barrages, and eventually I was able to kill them all with mass damage spells. It still took 8 rounds of attacks to defeat the Dark Lord.
 |
| End-of-battle statistics. |
Guess what happened then?
Nothing. Absolutely nothing. There was no congratulatory message. No, "you won!" screen.
Sighing, I realized that it was going to be that kind of game--the kind where you have to drag yourself out of the castle and back to the main city to see your victory screen. Consulting my maps, I marched down four levels and emerged triumphantly into the sunshine.
Nothing.
Could I have missed something? Or is it possible that there is no endgame? Is it further possible that, realizing they didn't have time or resources to develop an endgame before the disk had to ship, the developers deliberately made the final battle impossible?
I really hope I'm wrong. There is a file called "ENDGFX" in the directory, suggesting some kind of ending graphics. But I don't know what you need to do to trip them. Short of more information from the Internet, we need to regard the endgame as unachievable.
****
Edit from 23 March 2022: Years after I wrote this, commenter Busca directed me to a site whose author has exhaustively disassembled this game. He confirms that a bug causes the endgame message not to trigger but offers a workaround so that the player can view the final screen.
 |
| The "winning screen" to Dungeons of Avalon. |
He does not, however, offer a solution for the impossible final battle.
My original copy of Civilization had a corrupt PIC file in the intro. I copied the endgame PIC file over it and every time I started a new game, it showed me landing my spaceship on a new world. Maybe that would work for you here.
ReplyDeleteFor some reason I'm not very surprised this happened.
ReplyDeleteMaybe the fact that you are fighting Trolls in the last battle is a clue? ;-)
Anyway, I'd demand a refund from the computer game store!
My vote goes towards the gaming being made intentionally unwinnable. It took eight rounds with maxed-out characters to beat the Dark Lord, I imagine a normal party of characters never had a chance at all.
ReplyDeleteI think they ran out of time for playtesting. It dorsnt make sense to design FOUR levels of Dungeon And a battle to save the time for a winning screen.
DeleteProbably they wanted to do something clever, like an extra spell, item, area, trap, whatever that you need to win the battle (and no chance of winning without it) but they ran out of time to check if this alternative way is actually archieveble or possible to find.
Without a walkthrough or comments of people involved we wont find out Im afraid.
Peer's hypothesis seems sensible to me. Maybe there really IS an alternate way to defeat the Dark Lord, but with Quido and I both fully mapping the levels and not finding it, it seems doubtful.
DeleteThey probably forget to implement it (whatever it was) or it was bugged out. If they were using an Editor to design the dungeons its actually quite possible.
DeleteBtw in a later Amiga Joker there was a Hint about a ruby dagger that you had to find at a specific location by looking around. I dont think its related, but if items need apecific turning at a specific point, it might mean there might still be something around thats impossible to find without help (if the battles are so much more difficult in the castles, probably the puzzles are bordering impossible now? Think of Eye of beholder special quest level 1 impossible)
Its searching at level 0 location n34 o 17
DeleteThat Amiga Joker hint sounds vague enough to be misunderstood, but there's nothing in the game you find by "looking around"--there's not even any mechanism for that. The ruby dagger is an early quest item, found in a chest just like everything else.
DeleteThe only thing I can imagine that might solve this mystery is for someone to play the original, German, non-WHDLoad version and see if the same things happen.
Wow. This just gets funnier. I'm trying to imagine a scenario in which a developer might think it's a good idea to make a game intentionally unwinnable. If something's holding up making a proper ending, I'd rather go the other way and just script an ending with no fight, so that players at least can feel like they finished, rather than leave it in this condition. Seriously. A black screen with the words, "You use the scroll and defeat the dark lord. The end." would have been a more satisfying choice, wouldn't it? Not as good as a balanced fight and a nice screenshot, of course, but still better.
ReplyDeleteSo, I have to ask. In the past some lousy endings have been spoiled for you. This one seems to have caught you by surprise, and you've spent a lot of time trying to work your way around it. Now that you've had a chance to compare the two scenarios, do you think you would have been happier going into this one knowing it was unwinnable? Or is it still better to be surprised?
It's still better to be surprised. No question. Makes for a better experience writing, too.
DeletePlus, the "spoiler" isn't always right. Everyone told me that the ending to Eye of the Beholder was awful, but I just thought it was average. I spent the entire game anticipating a bad ending, which hurt my enjoyment of it.
"I'm trying to imagine a scenario in which a developer might think it's a good idea to make a game intentionally unwinnable."
DeleteFirst, don't think of them as developers. Think of them as a couple of European Amiga programmers who dashed off a quick game that was featured in a magazine. I can totally see them making the Dark Lord win as a joke. They'd find it hilarious.
"A balanced fight"? No way, this game did not contain 2015 design requirements.
This may be a game in only the broadest sense but did you ever read the "Choose your own Adventure" novels when you were young? There was one called "Inside UFO 54-40" which was nearly unique in that there was a specific special goal - reaching the planet of universal peace - Ulitma. There were pages in the book which described you reaching this goal. However you could never reach it because...'universal peace' can't be reached by following the rules.
DeleteSo you can add "deliberately frustrating children" to the list of reasons why an author might make a game intentionally unwinnable.
That one does stand out in memory. I referenced it in a post on Tunnels & Trolls, I think.
DeleteI remember it took me forever to "get" it. I found the ending in the middle of the book, and then I scoured the entire book looking for the steps that would take me to that ending, frustrated when I didn't find them. It was only after a lot of thinking that I realized what the passages meant about "you can't make it here by following directions."
Right, I thought you had mentioned it. Yes, I did the same thing. I recall figuring it out it was deliberate rather quickly but the impression the text left me with back then was more that ultimate peace was an impossible and purely abstract idea not that we need to push boundaries and break the rules. In any case I recall finding it lame and wanting to slap Packard.
DeleteI'm trying to imagine a scenario in which a developer might think it's a good idea to make a game intentionally unwinnable.
DeleteSadly, it happened on quite a few occasions. Guru Larry has a video on YouTube entitled "5 Purposely Broken, Unbeatable Games by D*ckish Developers" that compiles five examples. While a couple are edge cases, several were clearly done intentionally as a way of covering for incomplete or missing content, i.e. putting in an invincible boss, impossible time limit, or unreachable platform so you'll never find out there's no ending or last level.
BTW I dug Inside UFO 54-50 as a kid, though I don't remember how I figured out the key gimmick.
Haha holy crap. Intentional The Bad Guy Wins?
ReplyDeleteIf so, this is dreadful, like a petulant DM who doesn't want the characters to succeed in their quest.
Ahah, yes, it seems Chet is playing god and messing with the canon.
DeleteThe bad guy wins, and you can shelve the game on the "Won" category.
Well we found out if the sucessor starts with: After you got killed by the Dark Lord the first time, the posple decided to try another crew and another route...
DeleteBut I guess the kill screen of the first game would have hinted something...
I just had to check out if Peer was right and came to old review.
DeleteBefore investing lots of hours for playing the sequel, please check the portion about bugs.
http://www.syntax2000.co.uk/issues/51/avalon2.rev.txt
Good find. Maybe there's something like that here. There is one place in the castle where you're supposed to get a message, but it just says "Text 00." Maybe it was supposed to say, "Spin around three times before confronting the Dark Lord."
DeleteThat totally makes sense, realistically you never had a chance beating the DARK LORD. But I think the less realistic game would have been the better one.
ReplyDeleteRenaming ENDGFX to something that might be the intro is worth a try.
I tried it, but it oddly didn't seem to have any effect. I copied the ENDGFX file to several other files--"startup-sequence," "intro," "DGFX"--and didn't see any consequences when starting up the game.
DeleteThe only other possibility I see is it is a "trick" ending -- there is a way to win, but it requires something clever which triggers a special case win, and they never anticipate winning via combat in the code.
ReplyDeleteSounds reasonable, maybe you have to blow up the castle instead. That should be hinted on the scrolls, though, and several people missed it.
DeleteMaybe you have to kill him in his sleep?
DeleteThat's certainly a possibility, but if so, Quido and I both overlooked it despite exploring all the levels thoroughly. And the spirit of Kham definitely says to fight the Dark Lord--he won't even let you in the castle until you give him the Rune so he can make the special "anti-aura" scroll.
DeleteSo if there is an alternate way to do things, it still involves a battle.
What happens if you use the "anti-aura" scroll, and then run?
DeleteAlso, did you think to use the scroll when attacking with your maxed-out characters? If not, that might be the issue...
DeleteIf you use the anti-aura scroll and run, nothing really happens. If you enter the square with the final battle again, it's just as it was the first time. And yes, I did use it with my maxed characters. Until you use it, no spells can be cast.
DeleteIn the version from oldgames.sk there are two files in the S directory of the floppy image, TXE and TXD which seem to contain the winning messages in English an German.
ReplyDeleteIf you press 'G' in the "Adventurer Guild" you see a gallery of the monsters ('Left' and 'Right' to navigate, 'Space' to leave). There a two versions of the Dark Lord image. Could it be that you fought an Fake Lord?
But perhaps that graphic was intended for the sequel? Probably it is not worth the time to investigate.
The mystery deepens, but I don't know how to solve it.
DeleteThanks for the note about the text files. Here's the text:
****
As you bring the good message to your king, he looks at you. You cry, "The Dark Lord is dead." But the king gets angry and shouts, "If you are not telling me the truth, I will kill you. Many men retained the same, but they are all dead. They were all great liars! Show me evidence and I will believe you."
The leader of the party steps forward. He takes out from the bag the skull of the Dark Lord. "Look, my king. Is this enough?" He answers to the words of the king.
A long silence covers the old halls. Everybody is staring at the party. The king answers remorsefully. "Accept my pardon, heroes. Spread the good message all over my kingdom." He tells to his messengers. "This is a great day in our history. The reward belongs to your party." The land is defeated from an evil lord. See you in Part 2 of Avalon?
Cue credits.
Sounds a bit inspired by the Amiga ending to Eye of the Beholder. But it's otherwise weird. There's no mention of a king anywhere else in the game, and I have no idea how you get to him. Also, I cleaned up the language a lot. The English is MUCH worse in this file, with numerous spelling mistakes, mixed tenses, and so forth.
So I don't feel like playing this game, but I spent some time digging around in its files...
Delete- There's one mention of a king: when you enter the castle in Ghale, the game says something like "You stand before the castle of the former king Walter" (translated from the German version). This contradicts the manual though, which tells that the Dark Lord had the castle constructed.
- I dumped the dungeon data, and Quido's maps are essentially complete and correct, you didn't miss any part of the game.
- The "ENDGFX" file is a proper end-game screen. It's titled "Heros of Avalon", has portraits of your party and room for text on the right. See https://i.imgur.com/dHqYf5B.png (not sure if the colors are correct).
- I didn't find any endgame hints.
- Random idea: perhaps it's just the English version that's broken? (The translation certainly didn't receive any quality control.) Has anybody tried with the German version?
Oh, man, I love this kind of stuff. I can't offer any kind of help, but I'm excited to see where these researches lead. (e.g. Will the German version work? Was there actually an alternate path to victory, or is the endgame just one big screw-up? Can the bug be patched? etc.)
DeleteI too find this fascinating. I'd love to know more about how one goes about reverse engineering such a graphics format. I tried to, but I didn't get very far: http://i.imgur.com/DQHImrS.png
DeleteWell, I can walk you through decoding ENDGFX. It's mostly a lot of tinkering though.
DeleteFirst, take a look at the file in a hex editor. The graphics files all start with the characters "XX50". I found some text file that described this as "fake PP20", and PP20 apparently stands for PowerPacker 2.0, some Amiga archiver. On Aminet I found a "ppunpack" utility that could deal with these files, after it was modified to recognize the XX50 header. You seem to have cleared this hurdle.
At this point we have (presumably) raw image data. Generally when dealing with an unknown format, I first try to figure out how wide an image is (i.e., how many bytes per line). To do this, I use the raw data as 8-bit grayscale pixels to form test images of various widths, e.g.:
https://i.imgur.com/uTBoYeZ.png -- 36 bytes per line, just noise
https://i.imgur.com/Mm34Itu.png -- 40 bytes per line, we have a winner
(If nothing looks plausible, you most likely have a compressed image.)
If the file as a header, it will also be obvious: there will be some garbage on the beginning of the first line(s) of the image, and there will be a vertical tear in the image. There isn't one here though.
My initial guess of 8 bit per pixel was obviously wrong. The image looks much better with one bit per pixel: https://i.imgur.com/gjaioeR.png
But we still have five separate images. This indicates that we're dealing with planar graphics ( https://en.wikipedia.org/wiki/Planar_%28computer_graphics%29 ). ENDGFX uses five bits per pixel (max 32 different colors), and each iteration that we see represents a separate bit of the source image.
Combining the bits gives the following image: https://i.imgur.com/sXCN6nl.png
(The details are very dark here, this is the same thing with the brightness boosted: https://i.imgur.com/zcAQYfY.png)
Finally we have to figure out how the color works. There are still 64 bytes at the end of the file that we haven't used. If you have a small amount of "extra" data at the end of an image, it's almost always a palette. Here it looks like we have 32 12-bit RGB values (4 bits for each channel). After applying it to the previous image I get this: https://i.imgur.com/dHqYf5B.png
Getting the colors right without a reference screenshot can be a bit tricky though. The red background matches the rest of the game, but green brick is weird. Maybe I made a mistake here.
Thanks a lot for taking the time to explain. I'm still having trouble with the colour palette, though. When I try to interpret it as 32 12-bit RGB values, I end up with this: http://i.imgur.com/ujsMMpm.png
DeleteYou've combined the planes in the wrong order. (I made the same mistake initially.) The first plane gives the least significant bit.
Delete@Peter
DeleteAn amazing post and great sleuthing. You deserve some kudos.
*applause* Very nice.
DeleteThis site has an owner of the original listed: http://www.kultboy.com/magazin/4214
ReplyDeletePerhaps its worth to ask him if there are further contact details listed in the magazine (or a walkthrough in one of the later magazines, he seem to own a lot of this magazines)
It doesn't seem to be a space issue, since the castle is so long (as commented) - but making a game unwinnable isn't unheard of. In the case of Gladiator, "The Spectrum and Amstrad versions notably were programmed by one David Perry of later Shiny Entertainment fame, who hilariously made the final boss invincible because there was no memory space for the ending left." - from Hardcore Gaming.
ReplyDeleteOf course this is a port. It's hard to imagine someone making an original unwinnable, especially after investing so much work...
This has gone from tragedy to comedy to somewhere in between. Slow claps are in order.
ReplyDeleteThe entire mess reminds me of an endgame encounter with Namtar in Dragon Wars. You're *supposed* to do something special, but if you just march up to him then you fight his entire army ("It's you against an army, folks!"), a battle that appears to be unwinnable without hexedits.
ReplyDeleteIt's possible the developers meant to program something like that in this game, but failed due to time constraints or bugs.
That's a good analogy. Hopefully, we can get one of the developers here to clear things up.
DeleteOr maybe the BBEG was supposed to drop his head which you then take to the king ?
ReplyDeleteBut judging from the difficulty of the end fight it seems that something is simply messed up in data file.
Something like too many 0s in BBEG's HP total and not the first game to have such bugs.
When I had fought with Dark Lord, I didn´t try choose to "Run Away" from fight. Did You try "Run Away"? Mayby after try to fight with Dark Lord it could change something in Dungeons.
ReplyDeleteThere were a couple of combats in which I did "Run Away" just so I could reload, after it became clear that the party was going to lose. I confess I never ran away, then ran around the dungeon to see if anything had changed. That seems to be expecting an awful lot, especially if you don't get a message that anything notable has occurred. But I agree that it's another possibility.
DeleteJudging from everyone else's very adept detective work, I find the following scenario plausible:
ReplyDeleteThe game is likely winnable in the original German version. Perhaps the Dark Lord's instant kill is set to be more resistable, or the initiative system functions differently. Some small tweak would make the English version winnable as well, but it was never done because the code went to localization before the endgame was fully playtested. A similar bug prevents the win screen from launching, but it was never noticed because the localized version wasn't thoroughly playtested.
Things like this have happened before. It's not a perfect analogy, but consider the lack of harmonic gem drops in the PC port of Bard's Tale 3. No harmonic gems makes the early game almost impossible for a fresh party, but playtesters apparently never noticed.
Just an idea: If (hopefully) the versions are save game compatible, it could be worth a try to load that last game into the German version and see if it makes any difference for the final battle and/or endgame sequence.
DeleteThe English version I have is a WHDLoad version, but all the German versions I can find use .ADFs. I would somehow have to transfer the save game file onto the German .ADF, right? I'm not sure how to do that since I can't "open" the ADF from within Windows, and I'm not sure how to mount and edit the ADF in the emulator. I'll be glad for hints on that.
DeleteWithin the workbench, I mean. I obviously know how to mount a floppy file in WinUAE, but not so I can access and edit the files.
DeleteThat would probably require the use of the Amiga's command shell (CLI) or some dedicated tool, since the Workbench itself doesn't show the entire contents of a disk. Not sure if you want to dig that deep, but if you can send me the save game I'll gladly give it a try.
DeleteHaha, had the same idea :)
DeleteThe savegame seems to work in the german version, but I didn't have time yet to test it further.
Copying it to the ADF depends on what you have on your installation. The Classic Workbench images, and IIRC also the Cloanto ones, all have Directory Opus (the Norton Commander of the Amiga world) installed, which makes this rather simple, otherwise you have to set Window->Show->All Files (see: wiki.amigaos.net/wiki/AmigaOS_Manual:_Workbench_Using)
Okay, dahauns sent me a German ADF with my save file copied to it. Nothing is any different (aside from being in German). The final battle involves the same enemies and nothing happens afterwards. Of course, this doesn't necessarily mean anything if the saved game itself got corrupted somehow.
DeleteGreat, so that's one possibility eliminated (more or less). Is a German-speaking reader willing to play through the game, ideally on real hardware, and see if they encounter the same show-stopper -- or if they pick up on some poorly-translated secret (clear in German, opaque in English) required to beat the Dark Lord?
DeleteIf that doesn't work, the next step would be to see if a routine to access the ending is at least still present in the program code, and is getting branched around or otherwise not activated.
I was hoping that the Addict's addition of a disclaimer to this post meant some big revelation was imminent, but things seem to have died down. Unfortunately, I can't help solve this riddle in any meaningful way, so it'd be gauche of me to "bump" too much.
DeleteBut still, I do want to say that I hope someone eventually figures out (or otherwise discloses) what's really happening with Dungeons of Avalon's inaccessible ending & impossible final boss fight, and will be very appreciative of any efforts in that direction.
Oh, and: actually, Shin Megami Tensei, Final Fantasy, you should really play that on a console, traffic lights are gorgeous, Amigas are easy and intuitive computers where you "just have to _____", and Kenny G is great. There, did I hit every bullet point of obnoxiousness? :D
Thanks once again for all your effort in investigating the shadowy, sketchy, forgotten corners of the CRPG world.
Did the dispel Aura scroll work some place else. Can you read it outside of fight? Can you read it to Kham (in DM the quest giver is kind of the bad guy)? It would be so disapointing if it's only a bug. May be your save can work on a german version, and someone can get to the bottom of that.
ReplyDeleteJust had the same idea. ;-)
DeleteNo, the aura scroll only works in combat. In any event, Kham's visage disappears once you solve his riddles and complete his quests. There's now way o find him or get him to come back that I can see.
DeleteJust to add my 2 cents... I played the sequel many years ago, which was given away free on a magazine coverdisk. It was bugged so you couldn't progress beyond a very early point, as the game refused to recognize you possessed a key item in your inventory. It looks like the programming for the first game is better (you can get to the end) but kind of worse as you can likely never finish it!
ReplyDeleteThis one continues to bother me a bit, and I wish we could get a definitive answer on what was supposed to happen in that final combat or how we were supposed to win the game.
DeleteI hope the bug you mention doesn't happen in every version, or the second one is going to be a quick play, too.
Did you ever make contact with any of the developers? I thought that when you softened the language of your post, it meant you'd gotten in touch with one of them and were worried about offending them.
DeleteNo, I was never able to track them down. I put that disclaimer up there because upon re-reading my post the next day, I realized it was misleading and needlessly accusatory. I posted it in a state of post-game agitation instead of letting the entry cool for a couple days like I usually do.
DeleteThis one still bothers me (as I'd imagine it bothers you, no doubt infinitely more). I wish I knew something I could do to help figure it out, but short of a testimonial from the developers, I think the only other option would be a full disassembly, which I don't have the chops to pull off.
Delete