Help make sure my film about game preservation isn't stupid!


Also, @OneSecondBefore you might contact Kari Kraus or Rachel Donahue, the two people responsible for that paper above. My experience with them has been that they’re quite open and helpful with regards to what you’re working on.

(I’d also say they’re much more knowledgeable than I am on the subject.)


Much food for thought here, thanks @talbain! I was familiar with the Preserving Virtual Worlds project (it’s super cool), but I had not read that paper you linked. I’ll check it out!

Regarding the film’s imagery: I conducted the interview within the game itself, recording voice over a skype call. I was controlling the game in first-person view, with the subject (Issac)'s avatar visible as he speaks. Worlds surrounds the player’s view of the game with a very un-photogenic HUD that takes up much of the screen. I cropped this HUD out and replaced it with animated representations of myself and Issac, along with a “dialogue box” that contains an animated audio waveform that corresponds to the speech of whomever is talking. The goal is to make you feel like you’re watching someone speaking, even though you’re looking at a simple avatar.

Here are a couple screenshots that give you an idea of what it’ll look like. The first screenshot shows the arrangement I described (though the waveform part isn’t done yet).


I think if the intent is to show off the game world to lay people (unsure about intended audience and their technical/historical knowledge of games) and present Second Life as an ancestor, having some side-by-side imagery might help connect that more strongly, at least visually.


This interview was cut from a feature film about people using spaces in ways that weren’t intended. I had to cut it for pacing reasons but I liked it enough that I’m adapting it into this short film. It’s intended for a general audience. It’s primarily meant to be an entertaining documentary interview/character study with interesting imagery from a weird digital world. But it’s also meant to itself act as the kind of ethnographic document of a player culture that my narration discusses. And if I can provide a light introduction to a few concepts in game preservation then that’s great too.

I’m definitely coming around to the POV that the Second Life reference needs to go.


The interview itself focuses mainly on the current player community of Worlds, the story of a recent influx of newcomers, Issac’s history with it, his creations in it, and his efforts to preserve a record of it.


This comes across most strongly from what I’ve seen/read. Also, minor thing, but I think rather than saying a historian may have a different experience, you might just say that they will have one. I dunno what the future holds obviously. Maybe ARG will evolve into Holodecks and they can recreate 90s living rooms (that’d be pretty cool).


I’d also say don’t mention Second Life. At least don’t say “today’s” because Second Life is also pretty much dead.

I don’t agree the situation is comparable to the loss of silent films. The incredible cheapness of data storage and transfer has changed everything and works will rather rarely be “lost” per se anymore. The problem is rather the contextual information needed to properly understand or replay the work. This point actually seems to be the core of your film so I don’t see why you should play up this kind of alarmist and not very well-founded analogy.


I’d say that for the early web and early software generally, lost is an accurate term. For things around mid-2000s-ish and later, “effectively lost” would be more accurate. It is not as though they don’t exist, somewhere, but more that they a) cannot be used b) cannot be understood or c) do not have requisite context (the last one, in particular, is one archivists are hugely concerned about). Still, there is definitely a lot of information, even today, that is actually lost. I’ve even had a few developers ask me for source code because they broke something in a later build (that they couldn’t figure out how to fix) and didn’t have the previous versions to work from. Ironically, it’s typically after that first big loss of information that the organization in questions becomes a lot better about their data practices (though the big loss everyone talks about in game preservation is when Square Enix lost the source code to Kingdom Hearts and had to reverse engineer their own game).

I suspect, though have no objective proof, that these sorts of losses probably happen more often than any company is willing to admit.


Well OK, if you’re talking about private source code. I consider that part of the contextual information. Once something makes it out to the Internet and gets some amount of exposure, it’s past that. Linus Torvalds said back in the 90s he didn’t see the need to back up his source code because everybody was doing it for him.

Can you give an example of a specific piece of software from the 80s or 90s that you encountered at the time and was lost? Even in the 80s, data copying had already gotten a few orders of magnitude cheaper than it was earlier so I think that was already past the cutoff line for automatic/accidental public partial preservation.


I mean, Polybius might be an example.

But I will say that I couldn’t find explanations for basic DOS commands without the use of an engineering manual from the 1980s. I figured it would be easy to find those online, and while I found the commands, the explanation for how they work and what they can do was basically absent (or at least, written so poorly as to make them unusable).

Again, I consider process more important than product, and consider things like losing source code to be losing that software. You can reverse engineer and emulate it, but byuu (the person who created a cycle accurate emulator for the SNES) will tell you it takes a very long time to do so accurately and may actually be impossible for future software (or even software now), even for games, especially as games as a service becomes more normalized.

Loss of software is tied to some degree to hardware as well, and what you consider loss. Having access to software isn’t really the same as being able to use it - and use is what’s important. Having the binary in the right order isn’t enough (it’s why a lot of preservationists realize magnetic tape, even if it’s the safest medium for digital preservation, isn’t really viable in the long term, since it effectively makes the material completely unusable).

To use a weird metaphor, it would be like transferring a book into binary but then losing the understanding of how binary functions. You may have captured a book, in some sense, but you can’t use it.

(And all of this is a long form way of saying I can’t name specifically a piece of software that is lost in the sense you’re asking for.)


Right, whereas e.g. the poems or Sappho or most live TV from the 50s are almost totally lost. The archival problem to be solved has changed entirely and I think it should be openly acknowledged instead of making facile comparisons with the admittedly good intention of funding archiving but that ring false. If you want a historical analogy it might be a fading painting whose original colors can be restored after scientific analysis.


Yeah, this is a good analogy. It’s being proactive in our attempts at preservation rather than reactive.

I’d also say the thing that’s probably most likely to be preserved is pornography (at least, when it comes to surviving for future use).


just gotta say

that avatar is fecking incredible


Right?? One thing that’s very cool about Worlds is that players can create custom “avatar galleries” that exist as physical spaces in the game world, where you can view and select a custom avatar. This means that present-day explorers can wander upon old avatar galleries that people made in the 90’s, and equip those avatars. That’s how Issac found that one. It was created by a Worlds user from around 1996.


Bumping this thread so it doesn’t die. No huge updates to report, but I did almost accidentally overwrite half of the film yesterday! I accidentally dropped a clip onto my timeline in Premiere, replacing 5 minutes of footage. I didn’t notice and then saved and closed the project file. Fuck!

Turns out Premiere puts autosaves in a separate folder, so I was able to recover my work without any losses. Phew, thank goodness for idiot-proof failsafes.


You have now learned the importance of version control. And entirely as predicted, after almost losing something (not chastising you, just saying the story you presented is unbelievably common).


I remember someone posting a while ago that they were a digital preservation academic and that gamergate made grant writing a lot harder. Who was that? I think it was someone who’s not a regular poster. If anyone has a link to the actual post handy, I’ve been wanting to reread it.


It’s a topic of discussion in the above thread @BIGHEADMODE , though I’m not sure if this is the one you’re looking for.


Yeah, the thing is I do know better… I used version control when editing my feature. But I was cutting corners since this is a much more manageable short film. But it turns out version control is always worth it, even for the small stuff.


Thanks! I think I was misremembering the Twitter thread in the OP as an SB poster.