This article is mostly good, but saying machines are deterministic breaks down at the limits - - some emulators start hitting cases where physical chips enter inconsistent states, which non-deterministically then fall into another state. Emulating such things requires treating machines as analogue.
I probably should have clarified that my "point" here isn't really about genuinely replicating the behavior 1:1, but rather preserving the experience at a level where our culture & media can still be accessed in a form that maintains its core experience and intentions. Sure, an emulator might have certain bugs or idiosyncracies when compared to the real hardware, but you're still able to utilize a game or piece of software for its intended function. It still delivers on the promise of what it was created for. Looking into the future, this means that children of the year 4862 might still be able to play Super Mario Bros. It might get some pixels wrong, it might crash or freeze under some weird edge case scenarios, but the experience is preserved.
I guess, I'm a bit divided on this. While software is abstract (and deterministic), the experience of running the software is not. Meaning, as you become part of a feedback loop, the sensorical side of things matters. And this is really hard to recreate.
E.g., not only do emulators introduce lag, also modern systems tend to be laggier than vintage ones. Graphics taylored to the specific attributes of a display technology won't convey the original feeling on modern hardware. Or, interacting with a game using modern controllers, by D-pad or thumb stick, which are operated by fine motor control, feels much different from using joysticks that involve the entire body, by this establishing a rhythm. (This is also true the other way round, e.g., Pong, meant to be controlled by fine motor controls using a paddle, is half the fun using the brute means of a digital joystick, either jammed up or down.)
This may sum up to the effect of the user entirely missing the point of a game. (There are some VCS games, I always deemed to be noteworthy, which are generally rated low in modern reviews. I had a hard time understanding why reviewers would miss the rather obvious qualities of those games – until I tried them using a D-pad controller.)
P.S.: This doesn't mean that emulators aren't worth it (to the contrary, it's great to have them), but it may be worth keeping in mind that you are not necessarily having the original experience.
> E.g., not only do emulators introduce lag, also modern systems tend to be laggier than vintage ones.
Some emulators have run-ahead now, which removes internal processing frames from games. Combined with an adaptive sync monitor, 1000hz USB polling, and other tricks, emulators can actually respond to inputs faster than the original gaming consoles did on CRTs. It's pretty much the closest thing we have to magic.
While I agree with you on this for the most part, aren't we going to start hitting a ceiling with what's possible to do run-ahead with?
For example, I know in that article you mention bsnes, but wouldn't it be exponentially more difficult to have that feature in something like a GameCube or N64 emulator (let alone something like a PS3 with its funky Cell architecture)?
The system setup of old devices was quite different from what we're used to now. CRT TV's, no OS in the way. Just inputs and near instant display of what was going on.
The more modern your game system gets, the harder it is to perform runahead. But it's also required less and less. Newer consoles got more 'laggy' by default, with OSes, USB/Bluetooth input processing, triple buffering and so on. So runahead requirements have become less and less to recreate experiences.
As to you examples, the PS3 was meant to be played on LCD TVs much like our current setups. I don't know how RPCS3 compares to a real console, but I expect the difference isn't huge. Dolphin (GameCube) and Mupen (N64) can emulate their consoles so efficiently that you should be able to achieve 1 frame of runahead on somewhat powerful systems for many games.
Yes it definitely would be. Every frame of run-ahead requires emulating another instance of the system. So a run-ahead of four frames means needing 5x the processing power. Generally speaking you'd only want 1-2 frames of run-ahead, however. And in practice you can usually skip video rendering during the run-ahead frames to reduce overhead a bit.
> Some emulators have run-ahead now, which removes internal processing frames from games. Combined with an adaptive sync monitor, 1000hz USB polling, and other tricks, emulators can actually respond to inputs faster than the original gaming consoles did on CRTs. It's pretty much the closest thing we have to magic.
Super Smash Bros Melee has a little bit of polling lag on console (the tournament standard) due to poor coding. By removing this lag (or rather, making it configurable), you can "hide" network latency behind this lag people are used to, and online play can feel like local play.
I address this near the beginning - and you are correct that there are some examples that can have a significant impact on the resulting experience of the software. Emulation is not a 1:1 recreation of the original experience including all of its intricacies. But I'm thinking extremely long-term here, far beyond "what our kids will think about the games we played". I'm thinking about what historians of the distant future could learn from our creations, due to how well they can be preserved compared to other forms of media. Ideally people would have an intuitive understanding that input methods and displays make a tangible difference in the experience, and that an emulator is never going to be a full replication of that - so that hopefully they can study and judge a piece of software with those differences in mind.
Technically this is true of anything that can be represented digitally, but software's interactive nature gives it a much more profound ability to represent our culture and elicit a strong and personal reaction in those who experience it.
I totally consent with this. However, doing as much as we can to reproduce the experience is of some importance. E.g., there has been a basic Spacewar! emulator in Java since 1997, but it only recreated the game on an abstract level, missing the intricacies of the display. Many, including myself, somewhat shrugged their shoulders on this. It has only been, since there have been efforts to recreate the display experience on a more or less analog level that the game is appreciated for what it is, a quite perfect beginning, not to the least for its esthetics. (Disclosure, the web emulation of the PDP-1 running Spacewar! is by me.)
Based on this experience, I'd say, there's need to constantly update emulators to the current state of technical media representation. We are never on the safe side… This will only become trickier as any original hardware and/or media recordings of the software running are lost. Much will be lost in translation, since software has context.
> there has been a basic Spacewar! emulator in Java since 1997, but it only recreated the game on an abstract level, missing the intricacies of the display
Did this version actually emulate PDP-1 machine code, or was it a purely mechanical clone? Because if it's the latter, then that isn't an emulator.
Either way though, Spacewar is definitely one of those examples where the hardware directly impacts the experience. This is also why I am saddened by the fact that the only truly good CRT emulation I've personally witnessed is from Super Win The Game!, and since I believe that game is closed source, it unfortunately cannot be repurposed for other games or emulators. Unless you happen to know of a particularly good reshade implementation or something.
For a game like Super Metroid, the "CRT experience" essentially boils down to "a greater sense of atmosphere and an enhanced sense of depth in some of the sprites". Does this arguably improve the experience? Yes. But the core experience of Super Metroid is not lost if you can't experience it this way - I would argue Spacewar does lose a core part of its experience without that iconic display, and I absolutely believe we should make an effort to preserve that experience.
Regarding the Spacewar! emulation: The original one was by Barry & Brian Silverman and Vadim Gerasimov (in fact, every PDP-1 emulation, including simH, is based on this code). The Java emulation was a bit clunky on average machines (this being 1997), but later they ported it to JS and the HTML5 canvas [1]. It does instruction level emulation of the original code, but doesn't emulate the specific display, thus having to clamp two frames into one and halving the frame rate.
My version [2] built on this (with kind permission), adding further instructions and some enhanced fidelity in order to support a broader range of software. Also, there's now a more accurate screen emulation – well, as good as you could do in 2012 in a web browser on average hardware. Which is also of some importance for running classic "display hacks", which would otherwise just display a cloud of dots without much of a sense or meaning.
Generally, I like the media aspect and the material aspect of media, so I tend to put a bit of importance to this. Another aspect is the kind of artistic stroke in these media surfaces. E.g., Spacewar! owes much to the point plotting x/y display, combining both drawing and painted effects, e.g., for the explosions (these were known as painted displays for a reason). Maybe the most interesting video games in this respects are the early Atari coinops, which expose their internal construction in their compositions of counter-based video windows like Gothic architecture. (Something, which is missed as soon as there were bitmapoed graphics.)
We should try to preserve this kind of context. Not to the least, binary encoding is nothing without context – there's no clue to what this might be without external knowledge. (Is it a binary image, a music stream, a program, a data set? What is the compression algorithm used? What are the dependencies, on what kind of machine did it run? What was its purpose? Why was it worthwhile? How did it compare to other solutions? What were contemporary preferences?) There's actually an entire encyclopedia to go with every file. Preserving this will be a huge aspect of conservation, as digital artifacts pile up.
> "For a game like Super Metroid, the "CRT experience" essentially boils down to "a greater sense of atmosphere and an enhanced sense of depth in some of the sprites". Does this arguably improve the experience? Yes. But the core experience of Super Metroid is not lost if you can't experience it this way"
If there's some to "the media is the message", well, this was originally based on a study on the effects of projected media and the value attributed to it regardless of the content. So, yes, the specific mode of projection and the mode of bodily sync might be well the better part of the experience. However, the Great North American Video Game Crash (appropriately spelled and pronounced GNAVGC) may be still an argument to the extent that the contents still matters, at least a bit. :-)
>> "For a game like Super Metroid, the "CRT experience" essentially boils down to "a greater sense of atmosphere and an enhanced sense of depth in some of the sprites". Does this arguably improve the experience? Yes. But the core experience of Super Metroid is not lost if you can't experience it this way"
In all honesty I really have trouble buying the idea that a CRT is necessary to enjoy super metroid(or these old games in general). That's my favourite game, I grew up playing it on a snes, when we put the snes away after we got an n64, I kept playing it on emulators, the last time I played it I used an 8bitdo sn30 over Bluetooth on retroarch on my android phone, the game is the same. The atmosphere is the same, playing it felt the same, there was nothing 'missing' by not playing it on my snes on the old crappy TV in the hallway while my family walks back and forth in front of me.
Eight year old me would have been amazed and exctatic if he could have carried around and played the entire catalogue of 8,16 and 64 bit console games on a small computer that fits in my pocket using a wireless controller.
I've always thought of consoles as simply a means to play games. I love that there's hardware accurate emulators or even inaccurate emulators, that I have the ability to play these games on nearly any popular device that exists with graphical enhancements or using one of the many hacks or fan patches available. Being freed from the original hardware these games were developed for has allowed them to live on and be enjoyed in ways nobody back then even considered and for me that's far more awesome and exciting than trying desperately to recapture the exact look and feel of all these games.
CRT emulation of some kind is sometimes necessary to accurately render effects that were intended by the original artists. For example, this waterfall transparency: https://www.youtube.com/watch?v=s4QOHe4bOZc
Then there's the fact that a lot of early console artwork was designed to be output on a composite video cable instead of as separate RGB channels, so the crosstalk may need to be emulated to get the right color blending: http://www.chrismcovell.com/gotRGB/rgb_compare.html
(The reduced effective spatial resolution is often more of a detriment, so plenty of games do look much better in pure crisp RGB from an emulator.)
As indicated, this last section was meant to be read with a grain of salt. It's fair to debate the sensorical context of such a game and to assign varying degrees of importance to this, depending on individual perspective. But I'd also argue that a point like this may be made. E.g., early LCD games are generally not valued highly, while their bigger VFD cousins, which are probably not more complex, tend to be welcomed as charming devices, accordingly enjoying much better chances of preservation. Super Mario Brothers on a cheap NES-on-chip with a tiny built-in display (and several hundreds of games at the tip of your finger to switch to anytime, as well) makes it certainly harder to understand, why this could have been the fascination of an entire generation of kids.
Now we could argue that, for example, Alexandre Dumas' novels were for the most originally published as a series of episodes, which doesn't affect us, who know them only as books. However, it does matter to understand what this meant for some of the text strategy in those novels. And, since we still understand what this kind of publication is, we have indeed a chance to understand what this means. On the other hand, if a medium is entirely gone, it may be real hard to reconstruct this kind of context and it's meaning. (E.g., speaking of the C64, if you never have seen RF output on a customary consumer color cRT TV set, it may be hard to understand why Commodore switched from the rather elegant font of the PET to those blocky characters. Was it just for fancy? Right, it must have been so, since the 1980s loved bold fonts, weren't they?)
Surely, these are minor details, but they may built up to an extent, were it becomes hard to understand the worth of abstract code. In some cases, like Spacewar!, the threshold may be lower, in other cases it may be more robust, but I wouldn't dare to bet future history on this. I think, we still may require constant awareness and constant involvement in adapting resources like emulators. ("Who cares, it will probably still run in 20 layers of emulation" may be not enough.) But then, it may be just me with my interest into those kinds of media types.
However, for a beginning, we should come up with a format of ROM or media images, which includes the original instructions/documentation in a standard format. This may actually help quite a bit. (Compare the notion of ET on the VCS as the "worst game of all times", which only came to being in the 2000s, quite a while after the fact, and was based mostly on the fact that the game was now played without the instructions that would have been included with every cartridge. As I recall it, the most disappointing game on the system was the rushed Atari Pac-Man.)
It's the difference between events and artifacts, and there absolutely are differences between the two as time goes on that make even the very best preserved materials different in their essence. An event is experienced once and remembered as its artifacts; an artifact retains some substance allowing similar events to reoccur.
But on some level all artifacts are just echoes of the event and therefore are prone to decay and lose their meaning. Making it a digital artifact just allows some of those axes of decay to slow down, like an anti-aging therapy that can't treat dementia. And so it is also the case that physically realized games are more likely to play similarly in the past and present - sports have changed, Chess has changed, and both have seen reshaping via tech, but neither one unrecognizably so in the way that video games communicate phenomena with orders of magnitude more detail today versus 30 years ago.
The question, "which is better to favor, events or artifacts?" is a worthy one to ponder. Is it better to have fresh coffee available at all times, or a perfect coffee date once a month? Is it better to be able to listen to a recording 1000 times, or to attend a single live practice session?
What technology has done, beyond the basic needs aspects of Maslow's hierarchy, is let us pick and choose those options more often with fewer compromises. Its biggest failing tends to be in only solving a surface part of the issue: For example, having Facebook lets you message your friends and view their postings in some ambient fashion, without direct engagement. But it doesn't organize the party or send the invitations or prepare for guests or otherwise do the real work of a friend relationship. And so in that sense Facebook is a friend artifact that produces few useful friend events, and converts all those events that occur within its domain into proprietary artifacts of the company. But these artifacts aren't exactly worthy of emulating either. Anyone who controls them is probably sitting on blackmail.
> Imagine if we had a perfect replica of every conversation ever shared in a given location. Perfectly preserved lists of exactly who lived there, what their jobs and names were, and enough information about them to build a fairly detailed image of both the individual people and the culture they belong in.
Hopefully long after the people involved are dead…
I'm deeply worried that this generation's experiences are not preservable in the ways the last one's were.
Regardless of how you may feel about hardware quirks, etc., you can re-experience Prince of Persia today in a way that you will never be able to experience, say, Hearthstone the day after Blizzard decides to shut down the servers.
Hell, even right now you can't do it fully. Modern-day games have irrevocable change as a fundamental game mechanic. You can't even experience Hearthstone as it was _last year_ (with its ban list, standard rotation, etc). Not on ladder, not with your friends, not even in some offline form. This generation's gaming experiences are ephemeral by design. Nobody seems to care now but what a strange feeling it will be in 30 years that media from 60 years ago will be completely accessible, but from 30 years ago it won't be.
(I'm using Hearthstone as an example but similar principles apply to almost every successful mainstream game of this generation)
Why do we have such an attachment to preserving these these things? It's really only of interest for nostalgia's sake or for academics.
I think you can't actually re-experience Prince of Persia today because part of the experience is being in that historical context.
Hearthstone is only a game. Even if you could re-create the game, you can't re-create the feeling of queuing up with tens of thousands of other people queuing up for a new experience for the first time.
On the one hand, the Hearthstone game client and server may not be able to be preserved outside of Activision's vaults, but on the other hand people are producing more content than ever before, photos, videos, text, it's a deluge and you're free to capture and scrape as much as your budget can afford.
Maybe you're using hyperbole but I feel like there are better things to choose to be deeply worried about.
Even if it's for academics, the academics 30 years from now won't be able to study the game. Five years ago, I examined BASIC for the Atari 2600. I'm far from an academic, but even so, I was still blown away with that was done with 4K ROM and 128 bytes of RAM. Yes, the emulation was annoying (I think it would have been slightly easier on real hardware) but I still found it to be a positive experience.
To be sure, I could have had first hand experience with this, as I do remember seeing this cartridge advertised when it was initially released, but I never got the chance back then. And in retrospect, I think it's for the best, as I now have the background and experience now to know just how impressive a feat it really is.
So just because you don't feel a need to revisit games (or programs) of today 30 years from now (or use programs from 30 years ago) doesn't mean apply to everyone.
Academics 30 years from now won't be able to study the complete inside view of the trump presidency. Academics 30 years from now won't be able to study the complete inside view of what happened in wuhan during the covid outbreak. Academics will never be privy to the private thoughts of xi jinping.
Academics will never be able to experience hearthstone. They will only have thousands of hours of twitch or YouTube videos instead. This deeply worries me.
Alright, you caught me. I didn't properly scale my language, foolishly assuming it was obvious from context.
I'm "deeply worried" about it the same way someone might be "deeply worried" that the series finale of their favorite show is going to suck, not the same way someone might be "deeply worried" that COVID-19 is going to wipe out millions of people.
We have solutions for this but the companies don't want to do them. Get them to open source the game after a set amount of time. Get the full version control history. There is no other way. Otherwise the data will just rot away on some hard drive buried in a data center when the servers get shut down.
Emulation is a fascinating topic on many dimensions. One that I find particularly interesting is that it seems like more primitive systems are exponentially more "sharable" than small ones - I can fit a NES emulator with the entire NES library in a hundred megabytes or so. The entire library of the PS3 would be many many terabytes, and not only do we not even have good PS3 emulation, it's unclear if we'll never get it. While any programmer can code a functional NES emulator in a weekend, it is not the case for these more sophisticated platforms.
I wouldn't be surprised that, if humanity is still around and computing a few hundred years from now, people will still be playing SNES ROMs. PS5 games though might be entirely forgotten (especially given online components, endless patches, etc.).
The corollary of that is that if you want your software to be preserved by history, your best bet is probably to package it as a NES/SNES/GBA ROM or something like that.
When CDs became game media, content size exploded, since it was cheap.
> While any programmer can code a functional NES emulator in a weekend, it is not the case for these more sophisticated platforms.
Any programmer can code an NROM-compatible NES emulator in a weekend that will play Super Mario Bros., Duck Hunt, and a few other 8K or 16K games. The plethora of mappers that exists for NES turn the NES into a complex platform even if you don't account for perfect graphic and sound accuracy.
PS3 is harder to emulate because it had that Cell BE, but PS4 and XBone have gone the way of arcade hardware - it's all basically a PC now, and will probably remain so simply because it's cheaper to have one real platform. Emulation in the future may disappear because it won't matter--it's all a PC anyway so you might as well just play the PC version of the game, and if none exists, you're going to port it, not emulate.
> But regardless of these details, software has two main qualities that give it this advantage in terms of preservation: It's interactive It's deterministic
No sure if we are talking of the same thing here but a
Software is deterministic, but the randomness in a Game makes it non-deterministic, even if the emulation is deterministic. See, for exampe, a loot in a dungeon crawler. It will be different everytime.
The randomness in games currently available boils down to picking from a set of very large numbers and executing a limited number of code paths. That is to say, there are no non-deterministic video games, because all outcomes must be programmed.
A problem that can crop up with emulators is a lack of random seed values between runs, leading to, for example, a dungeon crawler with the same loot every time.
32 comments
[ 4.6 ms ] story [ 75.3 ms ] threadE.g., not only do emulators introduce lag, also modern systems tend to be laggier than vintage ones. Graphics taylored to the specific attributes of a display technology won't convey the original feeling on modern hardware. Or, interacting with a game using modern controllers, by D-pad or thumb stick, which are operated by fine motor control, feels much different from using joysticks that involve the entire body, by this establishing a rhythm. (This is also true the other way round, e.g., Pong, meant to be controlled by fine motor controls using a paddle, is half the fun using the brute means of a digital joystick, either jammed up or down.)
This may sum up to the effect of the user entirely missing the point of a game. (There are some VCS games, I always deemed to be noteworthy, which are generally rated low in modern reviews. I had a hard time understanding why reviewers would miss the rather obvious qualities of those games – until I tried them using a D-pad controller.)
P.S.: This doesn't mean that emulators aren't worth it (to the contrary, it's great to have them), but it may be worth keeping in mind that you are not necessarily having the original experience.
Some emulators have run-ahead now, which removes internal processing frames from games. Combined with an adaptive sync monitor, 1000hz USB polling, and other tricks, emulators can actually respond to inputs faster than the original gaming consoles did on CRTs. It's pretty much the closest thing we have to magic.
Here's a video demonstration I made: https://www.youtube.com/watch?v=1AvOa8yt6Vc
An an article of mine explaining the concept: https://byuu.net/input/run-ahead
For example, I know in that article you mention bsnes, but wouldn't it be exponentially more difficult to have that feature in something like a GameCube or N64 emulator (let alone something like a PS3 with its funky Cell architecture)?
The more modern your game system gets, the harder it is to perform runahead. But it's also required less and less. Newer consoles got more 'laggy' by default, with OSes, USB/Bluetooth input processing, triple buffering and so on. So runahead requirements have become less and less to recreate experiences.
As to you examples, the PS3 was meant to be played on LCD TVs much like our current setups. I don't know how RPCS3 compares to a real console, but I expect the difference isn't huge. Dolphin (GameCube) and Mupen (N64) can emulate their consoles so efficiently that you should be able to achieve 1 frame of runahead on somewhat powerful systems for many games.
Super Smash Bros Melee has a little bit of polling lag on console (the tournament standard) due to poor coding. By removing this lag (or rather, making it configurable), you can "hide" network latency behind this lag people are used to, and online play can feel like local play.
Like you said, magic!
Technically this is true of anything that can be represented digitally, but software's interactive nature gives it a much more profound ability to represent our culture and elicit a strong and personal reaction in those who experience it.
Based on this experience, I'd say, there's need to constantly update emulators to the current state of technical media representation. We are never on the safe side… This will only become trickier as any original hardware and/or media recordings of the software running are lost. Much will be lost in translation, since software has context.
Did this version actually emulate PDP-1 machine code, or was it a purely mechanical clone? Because if it's the latter, then that isn't an emulator.
Either way though, Spacewar is definitely one of those examples where the hardware directly impacts the experience. This is also why I am saddened by the fact that the only truly good CRT emulation I've personally witnessed is from Super Win The Game!, and since I believe that game is closed source, it unfortunately cannot be repurposed for other games or emulators. Unless you happen to know of a particularly good reshade implementation or something.
For a game like Super Metroid, the "CRT experience" essentially boils down to "a greater sense of atmosphere and an enhanced sense of depth in some of the sprites". Does this arguably improve the experience? Yes. But the core experience of Super Metroid is not lost if you can't experience it this way - I would argue Spacewar does lose a core part of its experience without that iconic display, and I absolutely believe we should make an effort to preserve that experience.
Generally, I like the media aspect and the material aspect of media, so I tend to put a bit of importance to this. Another aspect is the kind of artistic stroke in these media surfaces. E.g., Spacewar! owes much to the point plotting x/y display, combining both drawing and painted effects, e.g., for the explosions (these were known as painted displays for a reason). Maybe the most interesting video games in this respects are the early Atari coinops, which expose their internal construction in their compositions of counter-based video windows like Gothic architecture. (Something, which is missed as soon as there were bitmapoed graphics.)
We should try to preserve this kind of context. Not to the least, binary encoding is nothing without context – there's no clue to what this might be without external knowledge. (Is it a binary image, a music stream, a program, a data set? What is the compression algorithm used? What are the dependencies, on what kind of machine did it run? What was its purpose? Why was it worthwhile? How did it compare to other solutions? What were contemporary preferences?) There's actually an entire encyclopedia to go with every file. Preserving this will be a huge aspect of conservation, as digital artifacts pile up.
[1] https://spacewar.oversigma.com/html5/
[2] https://www.masswerk.at/spacewar
[3] https://www.masswerk.at/minskytron
Edit, regarding
> "For a game like Super Metroid, the "CRT experience" essentially boils down to "a greater sense of atmosphere and an enhanced sense of depth in some of the sprites". Does this arguably improve the experience? Yes. But the core experience of Super Metroid is not lost if you can't experience it this way"
If there's some to "the media is the message", well, this was originally based on a study on the effects of projected media and the value attributed to it regardless of the content. So, yes, the specific mode of projection and the mode of bodily sync might be well the better part of the experience. However, the Great North American Video Game Crash (appropriately spelled and pronounced GNAVGC) may be still an argument to the extent that the contents still matters, at least a bit. :-)
In all honesty I really have trouble buying the idea that a CRT is necessary to enjoy super metroid(or these old games in general). That's my favourite game, I grew up playing it on a snes, when we put the snes away after we got an n64, I kept playing it on emulators, the last time I played it I used an 8bitdo sn30 over Bluetooth on retroarch on my android phone, the game is the same. The atmosphere is the same, playing it felt the same, there was nothing 'missing' by not playing it on my snes on the old crappy TV in the hallway while my family walks back and forth in front of me.
Eight year old me would have been amazed and exctatic if he could have carried around and played the entire catalogue of 8,16 and 64 bit console games on a small computer that fits in my pocket using a wireless controller.
I've always thought of consoles as simply a means to play games. I love that there's hardware accurate emulators or even inaccurate emulators, that I have the ability to play these games on nearly any popular device that exists with graphical enhancements or using one of the many hacks or fan patches available. Being freed from the original hardware these games were developed for has allowed them to live on and be enjoyed in ways nobody back then even considered and for me that's far more awesome and exciting than trying desperately to recapture the exact look and feel of all these games.
Then there's the fact that a lot of early console artwork was designed to be output on a composite video cable instead of as separate RGB channels, so the crosstalk may need to be emulated to get the right color blending: http://www.chrismcovell.com/gotRGB/rgb_compare.html
(The reduced effective spatial resolution is often more of a detriment, so plenty of games do look much better in pure crisp RGB from an emulator.)
Now we could argue that, for example, Alexandre Dumas' novels were for the most originally published as a series of episodes, which doesn't affect us, who know them only as books. However, it does matter to understand what this meant for some of the text strategy in those novels. And, since we still understand what this kind of publication is, we have indeed a chance to understand what this means. On the other hand, if a medium is entirely gone, it may be real hard to reconstruct this kind of context and it's meaning. (E.g., speaking of the C64, if you never have seen RF output on a customary consumer color cRT TV set, it may be hard to understand why Commodore switched from the rather elegant font of the PET to those blocky characters. Was it just for fancy? Right, it must have been so, since the 1980s loved bold fonts, weren't they?)
Surely, these are minor details, but they may built up to an extent, were it becomes hard to understand the worth of abstract code. In some cases, like Spacewar!, the threshold may be lower, in other cases it may be more robust, but I wouldn't dare to bet future history on this. I think, we still may require constant awareness and constant involvement in adapting resources like emulators. ("Who cares, it will probably still run in 20 layers of emulation" may be not enough.) But then, it may be just me with my interest into those kinds of media types.
However, for a beginning, we should come up with a format of ROM or media images, which includes the original instructions/documentation in a standard format. This may actually help quite a bit. (Compare the notion of ET on the VCS as the "worst game of all times", which only came to being in the 2000s, quite a while after the fact, and was based mostly on the fact that the game was now played without the instructions that would have been included with every cartridge. As I recall it, the most disappointing game on the system was the rushed Atari Pac-Man.)
But on some level all artifacts are just echoes of the event and therefore are prone to decay and lose their meaning. Making it a digital artifact just allows some of those axes of decay to slow down, like an anti-aging therapy that can't treat dementia. And so it is also the case that physically realized games are more likely to play similarly in the past and present - sports have changed, Chess has changed, and both have seen reshaping via tech, but neither one unrecognizably so in the way that video games communicate phenomena with orders of magnitude more detail today versus 30 years ago.
The question, "which is better to favor, events or artifacts?" is a worthy one to ponder. Is it better to have fresh coffee available at all times, or a perfect coffee date once a month? Is it better to be able to listen to a recording 1000 times, or to attend a single live practice session?
What technology has done, beyond the basic needs aspects of Maslow's hierarchy, is let us pick and choose those options more often with fewer compromises. Its biggest failing tends to be in only solving a surface part of the issue: For example, having Facebook lets you message your friends and view their postings in some ambient fashion, without direct engagement. But it doesn't organize the party or send the invitations or prepare for guests or otherwise do the real work of a friend relationship. And so in that sense Facebook is a friend artifact that produces few useful friend events, and converts all those events that occur within its domain into proprietary artifacts of the company. But these artifacts aren't exactly worthy of emulating either. Anyone who controls them is probably sitting on blackmail.
Hopefully long after the people involved are dead…
Regardless of how you may feel about hardware quirks, etc., you can re-experience Prince of Persia today in a way that you will never be able to experience, say, Hearthstone the day after Blizzard decides to shut down the servers.
Hell, even right now you can't do it fully. Modern-day games have irrevocable change as a fundamental game mechanic. You can't even experience Hearthstone as it was _last year_ (with its ban list, standard rotation, etc). Not on ladder, not with your friends, not even in some offline form. This generation's gaming experiences are ephemeral by design. Nobody seems to care now but what a strange feeling it will be in 30 years that media from 60 years ago will be completely accessible, but from 30 years ago it won't be.
(I'm using Hearthstone as an example but similar principles apply to almost every successful mainstream game of this generation)
I think you can't actually re-experience Prince of Persia today because part of the experience is being in that historical context.
Hearthstone is only a game. Even if you could re-create the game, you can't re-create the feeling of queuing up with tens of thousands of other people queuing up for a new experience for the first time.
On the one hand, the Hearthstone game client and server may not be able to be preserved outside of Activision's vaults, but on the other hand people are producing more content than ever before, photos, videos, text, it's a deluge and you're free to capture and scrape as much as your budget can afford.
Maybe you're using hyperbole but I feel like there are better things to choose to be deeply worried about.
To be sure, I could have had first hand experience with this, as I do remember seeing this cartridge advertised when it was initially released, but I never got the chance back then. And in retrospect, I think it's for the best, as I now have the background and experience now to know just how impressive a feat it really is.
So just because you don't feel a need to revisit games (or programs) of today 30 years from now (or use programs from 30 years ago) doesn't mean apply to everyone.
[1] http://boston.conman.org/2015/06/16.1
Academics will never be able to experience hearthstone. They will only have thousands of hours of twitch or YouTube videos instead. This deeply worries me.
I'm "deeply worried" about it the same way someone might be "deeply worried" that the series finale of their favorite show is going to suck, not the same way someone might be "deeply worried" that COVID-19 is going to wipe out millions of people.
I wouldn't be surprised that, if humanity is still around and computing a few hundred years from now, people will still be playing SNES ROMs. PS5 games though might be entirely forgotten (especially given online components, endless patches, etc.).
The corollary of that is that if you want your software to be preserved by history, your best bet is probably to package it as a NES/SNES/GBA ROM or something like that.
When CDs became game media, content size exploded, since it was cheap.
> While any programmer can code a functional NES emulator in a weekend, it is not the case for these more sophisticated platforms.
Any programmer can code an NROM-compatible NES emulator in a weekend that will play Super Mario Bros., Duck Hunt, and a few other 8K or 16K games. The plethora of mappers that exists for NES turn the NES into a complex platform even if you don't account for perfect graphic and sound accuracy.
PS3 is harder to emulate because it had that Cell BE, but PS4 and XBone have gone the way of arcade hardware - it's all basically a PC now, and will probably remain so simply because it's cheaper to have one real platform. Emulation in the future may disappear because it won't matter--it's all a PC anyway so you might as well just play the PC version of the game, and if none exists, you're going to port it, not emulate.
No sure if we are talking of the same thing here but a Software is deterministic, but the randomness in a Game makes it non-deterministic, even if the emulation is deterministic. See, for exampe, a loot in a dungeon crawler. It will be different everytime.
A problem that can crop up with emulators is a lack of random seed values between runs, leading to, for example, a dungeon crawler with the same loot every time.