Source ports vs recreations

More than a limit-implementing engine, I would love to see a limit-checking editor, or an analyzing tool that would warn you of any and all possibilities of such occurences. Of course it would need to follow the engine closely for that, e.g. simulating what sectors and walls would be in view by making an 8-way check starting from the middle of each sector.

What we have currently is GhostlyDeath's ChocoRenderLimits, which makes it easy to spot critical bugs without that forcing you to leave the level (such as by a VPO).

Graf Zahl said:
I wouldn't call keeping a limit that could be removed without any problems an 'option'. Any map that stays within a limit would play fine, no matter whether the limit is there or not.

That's just one aspect. Design is another, and there you need the limits if your aim is to work within them. People do that, in the same way people might make sketches manually with a pencil instead of taking photos or augmenting sketches with The GIMP. Just look at AV or PL2... or many other WADs released in the past few years.

This whole argument gives me a headache. Is it really that bizarre for me to honestly not care whether a wad is made for Doom.exe or ZDoom or anything in between? With the exception of some wads that go out of their way to be different (many of which succeed extremely well, such as Foreverhood, Action Doom 2, and others), I don't really feel like there's that much of a difference. Vanilla fundamentalism, just like sourceport fundamentalism, seems utterly ridiculous to me.

When I mod for a sourceport, or even when I'm not modding for a sourceport, I like to make the absolute most that I can out of the options that it makes available to me, both in terms of mapping and other features. I've found that many impressive effects from high-profile ZDoom wads can be recreated completely transparently using only Boom features, for example.

Are there a lot of things about vanilla mapping that drive me up the wall? I would be a terrible liar if I said there weren't, but it's a fun challenge and I enjoy doing things that haven't been done before (or even just things that haven't been done by me before). With vanilla, with ZDoom and Skulltag, with Eternity, with Boom and so on. I love creating fantastic natural-looking sloped structures and architecture in ZDoom, and I love creating eye-popping 3D layouts and designs in Eternity....and then I find that it's still just as fun, and nearly as flexible, to push Boom's features.

It irritates the living hell out of me when I see some people bitching and moaning that someone would dare make a Boom wad instead of a vanilla one, and it similarly aggravates me when I hear complaints if something made for vanilla or another port doesn't work properly in ZDoom -- or worse yet, to cite a recent case, when someone complains that Plutonia 2 "removed" jumping in ZDoom.

tl;dr: Vanilla-purist fundamentalism and any kind of advanced-sourceport fundamentalism is absolutely idiotic. It's still Doom, it'll always be Doom, it's really not all that different. Just play the damn game, have fun.

esselfortium said:
With the exception of some wads that go out of their way to be different (many of which succeed extremely well, such as Foreverhood, Action Doom 2, and others), I don't really feel like there's that much of a difference.

You don't seem to be into speedrunning or anything that makes the smaller things matter. In fact, you seem more of a mapper guy than a player guy, so you're concerned with the stuff your kind of person finds relevant.

I enjoy doing things that haven't been done before. [...] I love creating fantastic natural-looking sloped structures and architecture in ZDoom, and I love creating eye-popping 3D layouts and designs in Eternity....and then I find that it's still just as fun, and nearly as flexible, to push Boom's features.

This is what I mean above. Your objective is to innovate in the mapping department. Others are also mappers but have different objectives, thus eventually they may see designing or what roles engines play in that, differently.

Vanilla-purist fundamentalism and any kind of advanced-sourceport fundamentalism is absolutely idiotic.

That starts to sound like middle-ground "I like a bit of everything" fundamentalism, because you end up hating the other "extremes" yourself. Are you better because your interests are wider in some respects? Should others also necessarily care about all these different things you find appealing?

It's still Doom, it'll always be Doom, it's really not all that different. Just play the damn game, have fun.

Variety is not made up only of people that like everything. It's made of people that like all this, all that, everything, this, that, that and a bit of this, and so on. Some don't give a shit about this or that, others don't even remotely understand why some other guy might prefer that or this. All these are potentially "fundamentalist" but that depends more on their intolerance toward other perspectives than on how many things they appreciate. It's hating what other people prefer that makes one a bigot, not liking just one thing. In fact, tolerance often comes from understanding one is partial, and thus being able to appreciate one's own select preferences and acknowledge others' different choices because you can see they are different like you are different, and not from a general sameness.

what the hell are all of you arguing about

myk said:

Are you better because your interests are wider in some respects? Should others also necessarily care about all these different things you find appealing?

No, I don't consider myself better, but we're all here for Doom and I personally don't really see that much difference between the experience I get from playing well-made sourceport wads and playing well-made vanilla ones. That's how I feel about it, based on my own experiences. Of course when it comes to the things that could be wrong with a wad that's poorly made for one engine or another, that's a whole different story. :) I've played Doom a lot, as many of us have, and didn't really get to use many sourceports at all until a few years ago. After the initial veneer wore off, it's just Doom.

In terms of playing wads, I appreciate interesting and interconnected layouts, exciting and unpredictable gameplay, beautiful visuals, originality, etc. Those are the things that make me personally enjoy the Doom community and what they come up with year after year. I don't really consider any of those things to be exclusive to any one branch of the Doom source code, though many of them allow for those things to be taken in new directions in different ways.

I'm really not trying to be an elitist of any kind here, I just can't understand how something like this gets turned into such a fight.

As far as speedrunning goes, since you brought it up: IMO if you're competitive about it, it makes sense to do it in the engine the wad was made for, or one that's reasonably similar. That just makes sense, to get the same experience as was intended, and the same experience that others who speedrun the same wad will get. If you're just doing it for your own enjoyment and don't particularly care about whether your playing methods will work in quite the same way in vanilla or Boom or whatever, does it really matter that much if you're having fun playing the game?

esselfortium said:
After the initial veneer wore off, it's just Doom.

I went in the other direction. As I refined my understanding, I kept narrowing down many things I thought were essential, fun, and worth the time. I'm also expecting different people to take this matter in varying ways, and dependent on the person and his choices.

I'm really not trying to be an elitist of any kind here, I just can't understand how something like this gets turned into such a fight.

The differences will not go away (at least not because one wants them to), and knowing that is one factor in not getting overexcited about them (calling people idiots isn't exactly something that leads to less arguing, heh). The worst of the arguing comes when someone feels his likes are being stepped on. Sometimes this feeling is warranted, sometimes it's delusional, sometimes half way there. Acquiring a certain degree of tact allows different people or parties to develop their preferences while not stepping on other people's (at least not intentionally or too much).

If you're just doing it for your own enjoyment and don't particularly care about whether your playing methods will work in quite the same way in vanilla or Boom or whatever, does it really matter that much if you're having fun playing the game?

Most people value sharing the demos recorded, comparing with what others record, even when there's no explicit competitiveness involved (though it may come spontaneously). Stuff recorded in the same format is more easily comparable (this ends up affecting the level design, to some degree, as well as the more obvious engine behavior). Not to mention of course online play, where people have to connect their engines or clients and agree on settings to even start playing.

myk said:

You don't seem to be into speedrunning or anything that makes the smaller things matter.

Speedrunning and the demo scenes are the two biggest justifications for maintaining ports that replicate faithfully Doom's every little quirk and bug.

But even then, I don't really see the point of keeping bugs that result in insta-crash like visplane overflows; by oppositions to bug that merely affect behavior such as wallrunning.

Gez said:
But even then, I don't really see the point of keeping bugs that result in insta-crash like visplane overflows;

The main reason to do it, in my opinion, is to make maps that run without any other add-on whatsoever (including source ports). Another reason is stylistic; as an "artisan" you're making levels very much like in specs to previous (vanilla) creations. And a third, related to the second, that the levels play consistently in regard to previous creations; their features and sizes are similar, especially.

Limit removing levels will often end up larger unless someone dictates some limitations through design principles, such as "keep the levels within these detail levels or of this or that size", but who decides to do that? If they do they keep it simple by making it technical (10 sectors, 1 monster type, 1024 square area, &c). You can compare Scythe and Scythe2 here. Erik kept the vanilla levels smallish and simple, making vanilla mapping quick and smooth, while the levels naturally expanded a lot in the second installment.

The latter two reasons may work with other engines, as well. Making a DOOM II remake in ZDoom may make it comparable in many ways to KDiZD, for example, or another in Boom to P:AR. Both in design achievement and in gameplay (these engines can give you more variety in this than vanilla, but you'll still be judging how the features were used to achieve it).

Creaphis said:
what the hell are all of you arguing about

Creaphis said:
I will deliberately take a contrary position just for the sake of writing incredibly long arguments

This says something to me.

This has gone way OT. Look at the damn title, dammit.

myk said:

That's just one aspect. Design is another, and there you need the limits if your aim is to work within them. People do that, in the same way people might make sketches manually with a pencil instead of taking photos or augmenting sketches with The GIMP. Just look at AV or PL2... or many other WADs released in the past few years.

But what makes the 15 year old Doom.exe limits special, except for the sole reason that the engines you prefer to use are stuck with them?

esselfortium said:

This whole argument gives me a headache. Is it really that bizarre for me to honestly not care whether a wad is made for Doom.exe or ZDoom or anything in between? With the exception of some wads that go out of their way to be different (many of which succeed extremely well, such as Foreverhood, Action Doom 2, and others), I don't really feel like there's that much of a difference. Vanilla fundamentalism, just like sourceport fundamentalism, seems utterly ridiculous to me.

No, it's not bizarre at all. I play every type of WAD that comes along, be it 'classic' like Pl2, story driven like Daedalus or just something pushing the boundaries like KDiZD.

What I *do* have a problem with, and which is the reason for this debate, is the attitude of people like Myk, who constantly go out of their way to argue in favor of Doom's original limits in a way that does make no sense at all - unless you assume that he is personally bothered by a larger number of WADs he can't play on his favorite engines.

I honestly couldn't care less about others' habits but there is one point where I see this becoming a problem: When some mappers feel obligated to map 'vanilla compatible' - not because they want to but because they think they have to to 'maximize audience'. To go back about my previous statement about Chocolate Doom, my problem is not its existence. There clearly seems to be a niche market for such an engine (which I am clearly not part of) but as soon as its existence makes one mapper decide to strip down his project to be compatible I think the community loses.

tl;dr: Vanilla-purist fundamentalism and any kind of advanced-sourceport fundamentalism is absolutely idiotic. It's still Doom, it'll always be Doom, it's really not all that different. Just play the damn game, have fun.

You have to explain one thing to me: What's 'source port fundamentalism'? This phrase is a contradiction in terms if you ask me. A source port, especially those which expand the underlying engine are as anti-fundamentalist as imaginable.

--------------------------------------
Some general response to Myk's more recent posts:


Say what you want but everything you say convinces me more of one thing:

Every word you post is that of a hardcore fundamentalist. Even the path you say made you arrive at the point where you are fits perfectly.

Even in other fields like religion it's not those who grew up with it that are the one that develop this system of strict and rigid rules of 'what is right' but often it's those who converted late.

From what I deduce out of your posts is that at one point you 'saw the light' of 'classic' Dooming being the one and only True Way of Dooming and to justify your position you come up with arguments that make no sense whatsoever to people who don't share your position.

They sure don't make to me.

I believe that such boundaries are there to be tested - to be pushed outward - to see what can be done if they are removed. I'm not a fan of conservative attitudes like 'all is fine, let's keep it'. That'd be stagnation and stagnation inevitably leads to demise. Doom wouldn't be here anymore if it wasn't for some source ports trying to push the limits further outward. The original game - as much fun as it may be - is hopelessly dated and there's not much there that may attract new users. New users are gained by engines that either push the creative limits like ZDoom or the graphical limits like Doomsday.

Graf Zahl said:

What I *do* have a problem with, and which is the reason for this debate, is the attitude of people like Myk, who constantly go out of their way to argue in favor of Doom's original limits in a way that does make no sense at all - unless you assume that he is personally bothered by a larger number of WADs he can't play on his favorite engines.

And then there's you who constantly goes out of your way to discredit all features of the original Doom engine because they are bad, bad, bad, and subsequently implement this philosophy into (G)ZDoom.

So you're both extremists and neither of you are necessarily right. You know what, though? The community pretty much requires both sides of extremism to progress, even if not many people want to be extremists themselves. People like myk ensure that the original Doom will always be playable -- new WADs can be targeted for it and be playable on every source port imaginable. People like Graf ensure that crazy amounts of new features go into source ports -- new WADs can do new tricks that vanilla could never dream of with all the DeHackEd in the world.

MikeRS said:

And then there's you who constantly goes out of your way to discredit all features of the original Doom engine because they are bad, bad, bad, and subsequently implement this philosophy into (G)ZDoom.

To be blunt, that's bullshit!

Tell me about one 'feature' (not bug!) that has been removed. And don't even mention stuff like wallrunning and similar glitches that were fixed because they caused other problems. Also, please name one vanilla WAD that doesn't work with GZDoom for anything else but rendering problems caused by hacks that are incompatible with GL rendering. As much as I'd like to make these work I'm well aware that with hardware rendering there's inevitably some problems that can't be resolved.

If you find such a WAD, please report it and don't post nonsense like this so Randy and I can have a look and either fix it or add a compatibility option.

Vanilla compatibility is important for me. Strict adherence to every minute detail of Vanilla's behavior on the other hand is not. If there's something that feels odd I'll fix it but if such changes have too bad an effect I'll think twice. I even reverted or defused some of Randy's changes that in my opinion were too harmful.

In the end, ZDoom's goal is not to preserve every single glitch but to present the overall experience in a more bug-free manner. For people that believe that all these glitches are necessary for a proper experience there's sufficient alternatives.

However, and here's the big difference: have I ever told anyone 'map for ZDoom because everything else is bad?' Because that's how many of Myk's posts come across to me.

Graf Zahl said:
But what makes the 15 year old Doom.exe limits special, except for the sole reason that the engines you prefer to use are stuck with them?

What makes anything in particular in the source special? What makes you think the limits are not part of the game like anything else? The reasons I usually prefer the original executables are exactly what you don't seem to go for; they keep the behavior of the game more or less intact in many respects. You don't have to care for that sort of thing, of course, but you can save yourself coming out as a loser who's hurt by other people's preferences.

Funny how your arguments have narrowed down to the limits only, because it wasn't so in the past. I'm glad we've come a long way in this respect, and commend you for broadening your perception to a degree. No one's perfect, but an improvement is always welcome. I've improved too, I think. I wasn't as keen on integrating purist play with other parts of community development in the past. In any case, now you can see something of the effects of the limits too, as shown above by comparing Scythe and Scythe2. As well as how making stuff comparable to each other sustains a recognizable way to play the game.

No, it's not bizarre at all. I play every type of WAD that comes along, be it 'classic' like Pl2, story driven like Daedalus or just something pushing the boundaries like KDiZD.

But what makes playing that variety of WADs special, except for the sole reason that the engines you prefer to use run them all?

And, no, I don't really have a problem with you playing whatever variety of WADs. I also play some of those, but nowhere near as often, I'm sure. This simply points out how technical choices, due to whatever reasons, help determine what one does. Not that I think everyone makes their choices according to an engine, though dedicated players and coders, to name two examples, may tend to concentrate on these for different reasons.

What I *do* have a problem with, and which is the reason for this debate, is the attitude of people like Myk, who constantly go out of their way to argue in favor of Doom's original limits in a way that does make no sense at all

How do I constantly go out of my way? Heh, you started the argument here, not me.

unless you assume that he is personally bothered by a larger number of WADs he can't play on his favorite engines.

Bothered how? Because I don't play them? This is the delusion I was pointing out above. I don't care about something, so then you end up thinking it's under attack. I have nothing against the making of or playing of WADs that I might not play much or at all. I just don't generally play them myself. I actually help people doing things I'm not into, even. Not as much as I support what I prefer, of course, but it's a contribution.

If I play football but not basketball, I hate the latter?

I honestly couldn't care less about others' habits but there is one point where I see this becoming a problem: When some mappers feel obligated to map 'vanilla compatible' - not because they want to but because they think they have to to 'maximize audience'. To go back about my previous statement about Chocolate Doom, my problem is not its existence. There clearly seems to be a niche market for such an engine (which I am clearly not part of) but as soon as its existence makes one mapper decide to strip down his project to be compatible I think the community loses.

Your own argument defeats itself. You do indeed have a problem with people who decide to make vanilla WADs for that reason (because they are NOT obligated), and do indeed have a problem with what the existence of such an engine does.

Do you think mappers owe you something when they make a WAD, and that they should choose a format or extension that suits your desires? Because you just came out that way, clearly.

How is choosing some other format, like Boom, VaVoom or EDGE different? And don't tell me that it's because they offer more mapping possibilities, because that can get in the way of what some people want.

And why stop at the limits? How are they not "stripping their WAD" when using only limit-removing, Boom or ZDoom and not all GZDoom features or the like? You have a problem each time someone releases a Boom WAD?

A source port, especially those which expand the underlying engine are as anti-fundamentalist as imaginable.

Way to try and give carte blanche to any mong that takes using an advanced port as an excuse to molest or accuse others who don't. A zealot is a person who with prejudice attacks others who are different, not someone who prefers simpler things for reasons based on usage or style.

From what I deduce out of your posts is that at one point you 'saw the light' of 'classic' Dooming being the one and only True Way of Dooming and to justify your position you come up with arguments that make no sense whatsoever to people who don't share your position.

You have to deduce things because you have to adapt what you're reading to your preferences and objectives. If, instead of taking my word for what I prefer and leaving it at that, you attempt to analyze me through your viewpoint, you'll be prejudiced. You argument falls flat on its face considering how tolerant I tend to be as far as what people may do is concerned. I'm not the one trying to stop people from doing certain things.

I believe that such boundaries are there to be tested

Why? Is it even a matter of belief? To me, it isn't. Changing things for the sake of change isn't any less arbitrary than sticking to a set of technical specifications without a motivation. Experience, not preconception, guides my choices. I've tried things, and selected what I thought was most enjoyable, and didn't worry whether my choices were going to sound reasonable to others (except inasmuch as I may give an explanation to those concerned or interested), and don't really dwell much on whether other people's choices make that much sense to me. If I were to go with what you do I'd feel I'd be destroying the game, especially for myself. Obviously this is not the case, because I can do whatever the hell I want and not have a problem with those who do things I wouldn't do.

New users are gained by engines that either push the creative limits like ZDoom or the graphical limits like Doomsday.

I think new users are gained when people start sharing their playing with each other, intensifying it, especially by recording demos or through online play. As far as I'm concerned, DOOM is a game and playing is in the foreground, not coding or mapping. Keeping certain standards to strengthen this is quite beneficial. Changes can be good too, because they broaden the spectrum of people that end up attracted to the games. The central hub or engine, however, is the essential and simple action packed gameplay delivered even in the standard game.

Graf Zahl said:
However, and here's the big difference: have I ever told anyone 'map for ZDoom because everything else is bad?' Because that's how many of Myk's posts come across to me.

Where? In that recent Freedoom thread? Vanilla was one of the choices there, and I argued for it because I like it. How am I forcing anyone by defending one of the eligible choices?

If I argue anything in regard to vanilla, vanilla is already part of the discussion. I don't go to a thread about a JDoom project and argue "don't do it like this, do it for vanilla". Keep in mind though that you DID literally say you had things against people choosing to make their projects vanilla compatible.

Graf Zahl said:

To be blunt, that's bullshit!

Graf, don't take anything he says seriously. MikeRS is known to be quite the troll.

I think porting or recreating Doom cannot be lossless. Todays compilers work differently than the NextGen one used, operating systems work differently, buggy code ported to other programming languge works differently.

I think consideration should be made before deciding to port about how much the loss would be. Like removing Doom's limit would ruin all the demos, speedruns and some (I think small number) of wads. I admire the heroic efforts some port authors make to even try to emulate the bugs in the original game. Other port authors decided the loss of demo compatibility is acceptable, some even removed demo support altogether. The gamers and map designers have the choice which port to choose. Its a free world, right?

rpeter said:

Other port authors decided the loss of demo compatibility is acceptable, some even removed demo support altogether. The gamers and map designers have the choice which port to choose. Its a free world, right?

Demo support can be one of the biggest obstacles when adding new features. You just can't change certain things - even if you'd need to to implement what you'd like.

So, as long as we are having a few engines that specialize in demo support all is fine. Do we really need more or all ports to support this? I think the only people for which this is the problem are those who have one - and only one - port on their system and expect it to run everything they come across and refuse to install something else.

D'oh, removing certain limits such as purely visuals limits shouldn't hurt compatibility, including demo compatiblity.

E.g. imagine that a map is playable, demoable, etc. on vanilla doom with the usual static maximum limit of visible SSEGS of 256. Now, if you make a port that supports 257 visible SSEGS without altering anything else, would that be able to break gameplay in any way? The answer is no, assuming that visuals and gameplay are entirely decoupled in the programmer's intentions and that there's no undocumented/unintended dependency, compiler bug, randomity or kludge in the code that would make the original limit of 256 a 'magic number'.

Under these conditions, you would have just made a superset of the original doom engine, with all of its properties, minus one weakness. A better Doom, if you wish. The general case of N+1 SSEGS is left as an exercise to the reader.

In the case of a crash-inducing bug, like VPO, a port with a tolerance for N+1 visplanes would not crash when the offending part of a level was reached, but continue playing until it hit its own limits. In the case of a demo recorded with a N-visplane port, which would have crashed at this point, it would just stop playback, but would be identical up to the offending point, again assuming the lack of "IT JUST WORKS THAT WAY, DON'T TOUCH IT" quirks in the code.

I don't see why Doom shouldn't be "Turing complete" and repeatable, in this sense. Unless we admit that there is some sort of "magic code" in the SC that none can really fully comprehend no matter how much effort is thrown at it. At that point, comprehending DOOM becomes more of an act of faith.

Graf Zahl said:

But what makes the 15 year old Doom.exe limits special

I think that's where the "must map within the limits because they are the limits" people lose me. From a mapping perspective, even a playing perspective, if the limits were there because of good aesthetic or gameplay reasons then I could see why people would cling to them, but they are not. The limits are mainly there, I suspect, because id didn't need any more whateveritwas, or perhaps because of 1993 hardware requirements, or maybe id didn't have time to expand certain areas or perhaps, even, they hadn't figured out how to remove limits that may have even limited their creativity.

In other words, the limits are there because the engine, at the time and within its development restrictions, did everything the game developers needed. It was not set up for, and the limits do not suit, a modding community where pushing the boundaries and, for example, things like line type choice are important. So, in today's context, the limits are actually pretty arbitrary. They are there for someone else, and for someone else's reasons. They aren't in there because the game doesn't work without them or going beyond them somehow spoils the game. They are there because that's all that was needed at the time.

So it isn't like making a pencil drawing when a camera to take a photograph is available. It's more like choosing to use a blunt pencil to make your sketch, even if a sharp one is available, for no better reason than because the guy originally in charge of making pencils never had the need for a sharp one.

There are some ways in which DOOM's behavior can become undefined, and the engine can actually run in a destabilized state on occasion without crashing, wherein some strange emergent properties occasionally come to bear - stuff that cannot be explained by the individual lines of code, but rather through their interaction in unexpected ways. I have witnessed some of these with my own eyes. Memory overflows and underflows are one example of something that can result in this, but it requires great serendipity. The usual result is just a crash.

But port authors do understand every bit in the source code, so they know those limits just don't always work the way you described. Just think of array overflows: the C language does not have a mechanism to detect them. In the original Doom under DOS most overruns went unnoticed, the engine simply used some garbage from memory. Under current operating systems the crash is however imminent. Up the limit: no more crash, however demos recorded under DOS go out of sync now.

Another kind of limit is in the collision code: sometimes you shoot through enemies (and vica versa). Fix it: you're also hit more often by fire and the spiderdemon gets weaker against rockets. Which case do you prefer?

How about this bug: mancubi fireballs sometimes hit you through the wall. Fix it: there goes demo compatibility again.

Even bounding box checking is broken in the original Doom. Fix it: God knows how it affects things.

True, a lot of those limits only affect demo compatibility. However, for some demo compatibility is paramount.

... and there's enough engines that provide it. So I don't see a problem.

rpeter said:

A number of bugs

For the record, I like all of those to be in their "fixed" state. But that's primarily because I spend a lot of my time using the Doom engine to make new monsters and situations and I want them to behave like they are meant to and not as the result of an internal quirkiness or an overlooked bug. I concede, however, that some of those do have an impact on the way the game plays and may not be viewed as desirable.

However, none of the things you mention are the mapping limits that I was talking about: the ones that people try to edit within, despite the fact they only exist because id never needed anything outside those limits.

But these are gameplay-altering fixes which are quite certain to mess-up demo playback. I was referring to leveraging statically imposed limits, which should normally have no effect if the code is robust enough. If developer's experience shows otherwise however, I respect that, it just means the codebase contains unintended behavior which depends on memory garbage, dangling pointers, undefined behavior etc. so that a computationally-complete DOOM recreation would also have to partially emulate DOS behavior and memory contents to some extent (since that's the particular version of Doom that is considered the Mother of All Doom, and versus which compatibility is claimed).

In that case, there should be occasions of demos going out of sync even in vanilla doom, using the "correct" version, if anything at all ends up being dependant on memory garbage, which was different even between MS-DOS machines. E.g. tutti frutti will look different even on the same machine on different occasions...or not?

Those limits are usually lifted in popular ports. You can even hack the original Doom.exe to up some limits. Map authors usually indicate if they wad needs a limitless engine to play.

Another question is whether such ports offer full backward compatibility with or without the demo stuff.

The general consensus here appears to be that there is more than meets the eye in the original source code, and that subtle changes may result in a sort of butterfly effect to the whole game. Ironically, source ports may have much more stable and predictable behavior than the original DOS Doom after years of polishing, yet DOS doom is still used as a comparison standard even if its behavior is, by admission of expert coders, at least partially undefined.

DOS Doom still remains the etalon, see the argument above. But you don't have to map for it, don't have to play with it.

I could imagine Chocolate-Doom with a menu item: ORIGINAL/LIMITLESS.

rpeter said:

DOS Doom still remains the etalon, see the argument above. But you don't have to map for it, don't have to play with it.

I could imagine Chocolate-Doom with a menu item: ORIGINAL/LIMITLESS.

It would require to fix the insta-crash bugs, and replace the crash by something else. Maybe an instantaneous return to the titlescreen with a message the limit has been overflown, continue yes/no? Yes then returns to the game, but in "limitless mode", no make you return to the main menu.

Maes said:

The general consensus here appears to be that there is more than meets the eye in the original source code, and that subtle changes may result in a sort of butterfly effect to the whole game.

That's only the type of overflows that's not being checked. Many of the problems cause intentional aborts (which by itself was a very stupid thing - aborting a program for the most banal of errors is just plain retarded) so if such limits are removed nothing undefined can be removed because Doom.exe would not work anymore if such limits are exceeded.