Lemmings 1 music problem in DOSBox

[ccexplore]

One of these days I'll have to check the floppy version against the abandonia version to see what if anything exactly changed.  Actually, I guess at the moment, I don't even know yet whether my floppy version is actually free of this problem or not.

[DoubleU]

I don't honestly think it would have happened by accident, though; it would have been simpler to do it the other way.  And for other versions, perhaps they changed their minds, or had different ideas?

Truth be told, I think the CD version, or the "cracked" version (CD or not), is a bit wussy. In addition to "fixing" the music "bug", they increased the drop height, making some levels much easier than necessary.

Anyway, it's not harsh or malicious; the idea is to be encouraging, even after mistakes have been made - i.e. to discourage players from giving up.

Besides that, have you considered that it might be normal for developers to do things like that? Otherwise, Mario games would have infinite lives, you could skip levels, there would be no need for passwords in Lemmings or other games, etc.

The SNES version has something similar: If you fail a level, you have 10 seconds to opt to try again or you get returned to the main menu.

[DoubleU]

bump

[ccexplore]

Wow, you must be really bored to bump on such an old topic from over 2 years ago.

By now I don't even remember where I copied the files from the floppy to.  I'm sure there's a backup somewhere at my place but I'll have to dig it up.  I think I can actually find the floppy disks themselves faster except I don't think I still have any computers that even have floppy disk drives, and I don't remember if I still have the USB floppy drive either.

Anyway, the old laptop with the debugging tools that I used to examine Lemmings EXEs has long been dead and while I backed up most of the files on it, I never bothered to setup any of my new computers so far with the same tools I'd need.   So sadly, I'm not exactly ready to go to look at this again even if I have the time or motivation.

I would ask a simple question though:  IIRC, the Amiga and Atari ST versions are developed basically around the same time as DOS versions, and all 3 by DMA.  If you think the music reset behavior is on purpose then it would make sense for DMA to implement the same behavior on those other two versions as well.  Anyone knows if those versions reset the music?  If memory serves I want to say the Amiga version does not reset the music like that, though to be fair I haven't played that for so long I don't really remember for sure.

[DoubleU]

Actually, no, that doesn't make as much sense as you might think. The Super NES version gives you 10 seconds to choose to retry a failed level; after which it goes back to the menu. It also has a new introduction with different music, and cutscenes in between ratings. The Genesis version does neither of these things, and is more like the Amiga version that way. By your reasoning, shouldn't they be a little more similar? They were both made by Sunsoft.

The DOS music order is also different from the Amiga, Atari ST, and Mac order. Why didn't they keep it the same? (I suppose I'm grasping for straws with this one.)

Back to the point, the entire idea is to encourage players not to give up until their ship's sunk (although, admittedly, the idea doesn't work so well with challenges or certain levels that rely on chance). Anyway, aren't you glad it's such a small penalty rather than something big like resetting the game every time you fail a level, so that you don't need to put in the code every time you wanna play again? It's certainly a lot more lenient than the way Mario or Sonic games work: enough deaths and you must start over.

[Dullstar]

"Encouraging the player not to give up" is an interesting spin on "punishing failure." It's kinda like doing a task at gunpoint (except with much lower stakes) - I'm not sure I'd exactly call the gun "encouraging" even if it does provide a major disincentive for giving up (failure by default).

I'd argue lives serve as a holdover from arcades. My reasoning for this is that they make the most sense in the arcade context - where the player pays money for a set number of tries. As games have progressed, lives have become more rare, and games that still use them tend to be more generous with them, and less punishing for running out of them. In fact, within the Mario series, by Mario 64 game overs were a mere annoyance, but didn't actually cost you any progress (unless you'd been intentionally not saving). As such, I suspect game developers put lives in their games not because they had a design reason to do so, but because "we've always done it this way."

Passwords do not constitute a good example of a failure-punishment mechanism. They were typically used to allow a playthrough to be stopped and resumed later before saving was common. In fact, it is believed that the original Metroid uses a password system to save on production costs (the Japanese FDS version saves, and the North American cartridge uses a similar board to the original Zelda, which has battery-backed saves. They probably didn't want to spend money for the internal batteries, so they replaced the save system with a password system. In many cases, such as the original Metroid and Lemmings, you don't need to use the passwords on failure.

I'd also argue that the SNES's 10 second countdown is a strange design choice rather than a mechanic actually intended to do something. I'd hazard a guess that, "A lot of other games are doing it" was probably involved in the decision-making process.

[ccexplore]

Actually, no, that doesn't make as much sense as you might think. The Super NES version gives you 10 seconds to choose to retry a failed level; after which it goes back to the menu. It also has a new introduction with different music, and cutscenes in between ratings. The Genesis version does neither of these things, and is more like the Amiga version that way. By your reasoning, shouldn't they be a little more similar? They were both made by Sunsoft.

There are of course differences between even the Amiga, Atari ST and PC DOS versions, like the music ordering you mentioned.  My point is more that it would be far more convincing of an intentional design if they actually adopted the same music reset behavior across all 3 versions.  For example, like you mentioned, you can see that all 3 versions kept a password system and did not have anything that would eventually force the player to have to replay from "the beginning", so the evidence is much clearer and stronger that we have an intentional design there with the passwords system.  Heck, it would be more convincing if we even just see the music reset behavior on more than merely the DOS version as it seems to be the case.

The fact that DMA didn't have consistency with the music reset behavior across the 3 versions that were co-developed at basically the same time, suggests that even if it were intentional and not a programming error, it was at least not universally agreed upon even within DMA.

Sometimes some version differences can be attributed to accommodating differences between the respective platforms, like the reduction from 100 to 80 lemmings and removal of many water objects in DOS version, likely for performance reasons.  The difference in music reset behavior does not seem to be something that can be explained that way either.

I don't remember for the SNES and Genesis cases whether Sunsoft is just the game publisher or is also game developer.  Either way, I think those two versions weren't developed quite at the same time like the case with DMA, and probably had less overlap of personnel compared to the DMA situation.

[DoubleU]

"Encouraging the player not to give up" is an interesting spin on "punishing failure." It's kinda like doing a task at gunpoint (except with much lower stakes) - I'm not sure I'd exactly call the gun "encouraging" even if it does provide a major disincentive for giving up (failure by default).

I'd call that an exaggeration. This idea isn't like a gun, just something cosmetic. That is, the music doesn't affect the gameplay, the way that (say) being forced to start from a checkpoint in Super Mario just past a powerup does. You have the same advantages and disadvantages as before, you just can't go on to more music until you beat the level. Therefore, the more often you beat a level (esp. the first time), the more different kinds of music you hear, while hearing "Lemming 1" less often.

A question of my own, that I might ask, is that it was programmed that way by accident, how would that have come about? It seems more complicated to me than just tying music to levels, or the other DOS versions of starting with #1 and changing under the sole condition of victory. True, there's another version that might be so described (i.e. doesn't reset to Lemming 1 each time, like on the CD version), but it also increases the splat height from 63 to 66, making some levels a lot less fun, being a lot less challenging. (Like, um, Lemmingology. Well, heck, "Snow Dragon" seems to disagree with me, but still.) I'm all for not being to harsh with players, but I know I like a good challenge.

[ccexplore]

A question of my own, that I might ask, is that it was programmed that way by accident, how would that have come about?

One class of programming errors involves changing more data in memory than intended, resulting in corrupting data that is unrelated to the thing the programming is trying to achieve, leading to unintended side effects.  So you can imagine one case where during the programming's processing of "stop playing music" (as would be the case when you leave the level, victorious or otherwise), a programming error like this always ends up resetting the music to track #0 or similar.  It doesn't matter on victory because the game has to load the next level and there is already programming there to update the music accordingly, but on losing, the level is already loaded and does not need to be reloaded, so no code was run that would undo the resetting done earlier.  If the game is actually intentionally programmed to reset the music on the condition of losing, you would expect the code to do so to reside alongside all the related programming around detecting and handling losing a level, rather than in the sound-handling programming as something that would happen whenever it is stopping music for any reasons.

Anyway, I stress that the above hypothetical is complete speculation and not based on any actual examination of the game's programming.

It is especially interesting to me that the thread suggests either a version difference, or possibly a difference between running the game in the environment of DOSBox versus an actual running of DOS, as causing some seeing the music resetting behavior in DOS Lemmings but not others.  Examining the game's programming should provide useful clues in further understanding those cases.

[DoubleU]

It is especially interesting to me that the thread suggests either a version difference, or possibly a difference between running the game in the environment of DOSBox versus an actual running of DOS, as causing some seeing the music resetting behavior in DOS Lemmings but not others.  Examining the game's programming should provide useful clues in further understanding those cases.

If it helps, I can attest to the fact that, at least in the demo, that's exactly what happened with DOS Lemmings, which I'd often play in Windows 3.x before discovering 95 and DOSBox.