We can read a thousand-year-old manuscript, yet archivists cannot decipher some materials that are less than 20 years old. —Gould and Varlamoff (2000:46).
Although technology is a key element in digital preservation, we believe it isn’t the greatest inhibitor – the lack of organizational will and way is. Despite the increasing evidence documenting the fragility and ubiquity of digital content, cultural repositories have been slow to respond to the need to safeguard digital heritage materials. —Kenney et al (2005)
At the 2007 Game Developer’s Conference, a panel of five speakers from industry and academia proposed a Digital Game Canon, consisting of ten games of historical and cultural significance “that everyone should play”. The games were: “SpaceWar!”, “Star Raiders”, “Zork I: The Great Underground Empire”, “Tetris”, “Sim City”, “Super Mario Brothers 3”, “Civilisation I/II”, “DOOM”, “Sensible World of Soccer”, and “Warcraft I/II/III” (Game Preservation SIG). The Canon had a mixed reception, as several posts to a thread on the Digital Games Research Association’s ‘Gamesnetwork’ list show. Initially unrelated to the GDC panel, list members were responding to a request for suggestions for a Game Studies canon; the Digital Game Canon event was introduced to the discussion part way through. Some offered their own list of the most significant game titles ever: Jonas Heide Smith volunteered the Danish Game Researcher Network’s canon (2007). Others demurred to the wisdom of creating a new canon: Barry Atkins asked whether “we really need a ‘best of’ for every academic category”, a response with which I had some sympathy (Gamesnetwork). Nevertheless, as a tactic, the Digital Game Canon was astute: the panel discussion received widespread coverage, including stories in the New York Times and Der Spiegel, and people began discussing the historical significance of games in a way they previously had not. Yet while this GDC session was a watershed, it is worth pausing to consider what this means. It means that 10 titles, chosen by five panelists (Matteo Bittanti, Christopher Grant, Henry Lowood, Steve Meretzky, and Warren Spector), will receive some attention (they are preservation targets in the Library of Congress funded project “Preserving Virtual Worlds”). It is notable that most of these titles were hits, and that, apart from “Tetris”, they were all either products of the United States, United Kingdom, or Japan. With the exception of “Spacewar!”, they are all well known internationally by virtue of their commercial success. This is understandable. However, as it is common to ask regarding canons: what has been left out? What other digital games, and indeed other types of software, might we consider to be culturally and historically significant such that they also warrant preservation? This is a question that needs to be asked repeatedly; the answers to it will, of course, vary depending upon the respondent’s perspective. In this article, my concern is with local software, specifically locally produced computer game software.
The question about local game production has yet to be asked in a scholarly context. Digital game history has usually been told with a focus on the U.S. or Japan, assuming the uniformity of products and reception worldwide (Herz, Kent and Schwartz, Sellers). While early digital games were a global phenomenon, the 1970s and 80s also saw games produced and consumed in specifically local contexts. Remarkably little is known about the local aspects of digital games history. In contrast to the high profile games targeted by the Digital Game Canon, in local contexts it is not always clear what software even existed; consequently, identifying priorities for preservation becomes a challenge.
I became interested in the preservation of local game software during the course of my research into the history of digital games in New Zealand. In 2004, I became aware of a significant quantity of locally written game software from the 1980s. Apart from a few specialist collectors, no one seemed to be aware of this, and no institutions were collecting it. Such a situation – where private collectors are effectively acting as custodians and amateur archivists of the nation’s digital cultural heritage, while institutions charged with collecting, preserving and making a nation’s creative output accessible overlook software – is not unique to New Zealand. While the work of collectors is to be applauded, as hobbyists, they have limited resources for undertaking digital preservation work. These circumstances prompted me to assemble a multidisciplinary research team that would be capable of bringing a range of expertise – legal, technical, and media-historical – to bear on these titles’ preservation. The project that is discussed here was conducted with the intention of advancing the software preservation agenda, in the hope that some of the early digital heritage of New Zealand – content that is unique internationally – will endure.
This article reports on our NZTronix team initiatives regarding this early computer game software. It begins by briefly introducing the game preservation landscape, before outlining the case for the preservation of local game software. The legal, technical, and media historical challenges addressed by the team in a pilot software preservation project are discussed next, and the mixed successes we enjoyed. The final section considers some of the key concerns in digital preservation discourse, including authenticity, fidelity, and variation, highlighting some of the points where digital game preservation diverges from other preservation initiatives.
Game Preservation: Digital decay The need to urgently preserve early digital games is an internationally recognised one (for summaries see Wilkinson (2005), Lowood (2004), and DiGA (2004)). Briefly, the main issues that threaten digital games’ longevity are: that content exists on degrading media (eg. magnetic tapes, floppy disks) and relies on obsolete operating systems; that the lifespan of hardware is limited because of physical and chemical processes within the chips (DiGA); and that copyright restricts the ability to undertake preservation work. As Wilkinson notes, the wide variety of gaming systems, the different technologies used, and the hundreds or even thousands of games produced for many of them makes this “truly a huge task”. These problems, of outdated hardware, deteriorating software, and intellectual property restrictions make accessing, preserving and distributing early digital games difficult; to do this while remaining within the law is even harder.
Many of the challenges of game preservation are common to other complex digital media objects.1 Consequently, technical interventions to preserve games are likely to also hold benefits for other interactive media, such as new media art. It is curious, then, how little cross-fertilisation there has been, to date, between digital game preservationists and other digital preservation communities such as librarians, media arts conservators, and museum professionals. The work of the vintage game community receives several mentions in the transcript from the 2004 Variable Media Network symposium “Echoes of Art: Emulation as a Preservation Strategy”; however, alongside their admiration for the work of game fans, the Network perpetuates the erroneous impression that the problems of game preservation have largely been solved. This misunderstanding appears to be reasonably widespread (see also Campbell-Kelly, 2003: 272). While the work of many fans, collectors and preservationists is commendable, and playable versions of many games currently exist, many challenges still remain. The legality of downloading most games on the internet is dubious, at best; the quality of preservation work undertaken and the longevity of game titles is also a pressing issue. Bruce Sterling pithily summarises the problems common to the preservation of born digital artifacts in the Variable Media anthology Permanence Through Change: “Bits have no archival medium. We haven’t invented one yet…We have no way to archive bits that we know will be readable in even 50 years. Tape demagnetizes. CDs delaminate. Networks go down.” (2003:20)
There are a number of game repositories and preservation projects currently underway, many in the United States. These include: the Software Preservation Society (formerly Classic Amiga Preservation Society); the Stephen M. Cabrinety collection and the “How they Got Game” project, both at Stanford University; the University of Texas Videogame Archive; the Strong National Museum of Play; and the “Preserving Virtual Worlds” project. In Europe, the work of the Berlin Computerspiele Museum is important, as is that of the related organization DiGA. Media arts conservation projects like the Canadian DOCAM Research Alliance are also worth mentioning here, given the similar challenges faced by those working to preserve games and digital media art.
Game Preservation: Local game software It is worth pausing to answer the question as to why local software matters? Why should it also be preserved? Why is preserving the titles in an international game canon not sufficient? There are a number of answers to this. Games were one of the key ways in which people became familiar with the then new technology of computers. Playing games allowed users to build confidence with computers. Particularly on early machines, playing a game required a user to first write or type in the game code. This authorial involvement points to a likely diversity in the game software of different cultures. While games generally are significant historical artifacts from the path to our becoming digital, the diversity of local softwares will be important for future historical work on how games and computers came to the rest of the world (that is, beyond the U.S. and Japan).
New Zealand’s software history is remarkable for the numbers of people – both companies and hobbyists – who were involved in writing software, and the numbers of locally published titles (Wheeler and Davidson). While software was rarely developed in total isolation, local structural conditions kept influences from abroad in check. As well as being geographically remote, trade restrictions further moderated influences from the US and elsewhere (see, for instance, Swalwell and Davidson, forthcoming), while government incentives for the electronics industry supported the development of local games production. Inquiring into local conditions of production and reception usefully complexifies existing, taken-for-granted histories of game and software development. It also reveals hidden networks of friendships and influence (Oppenheimer).2 For instance, a Director of the Auckland-based Kitronix company, Ralph Stevenson, allegedly corresponded with the founder of Atari, Nolan Bushnell (circa 1970s), before developing his own version of a television ping pong game. Les Kenyon is another New Zealand based protagonist with links to international game history. After emigrating from the U.K. in 1980, Kenyon set up his business, Grandstand, locally. The ex-Hanimex salesman would go on to be a significant name in Australian and New Zealand games history, as the importer of the Sega SC3000 microcomputer, as well as the distributor of early handheld games in the U.K.3
Despite the rarity of many local game titles and their increasing fragility, they are unlikely to be the (direct) beneficiaries of large preservation projects such as those mentioned above. If local software is to be preserved, local people will need to do it themselves. In the absence of purpose built media museums (Jűrgen Claus cited in Lowood 2004:8), existing cultural institutions are probably best placed to undertake software preservation work. Yet despite an increased interest in digital heritage, games seldom register on their institutional radars. Partly this is because ‘born digital’ forms of heritage continue to be sidelined by the “frenzy to digitize” material artifacts (Gould and Varlamoff, 146; Cameron and Kenderdine); partly, a form of amnesia is operative, albeit one that is occasionally interrupted by nostalgic reminiscence (Swalwell 2007b). Underlying both responses is an acceptance of the ephemerality of electronic media and the inevitability of their decline. Getting comfortable with the imminent loss of cultural heritage materials takes some effort; it does not come easily to curators, archivists, librarians, and others whose professional lives revolve around heritage preservation. Such resignation is the result, I believe, of a widespread perception that software preservation is “too hard” and that “there are too many problems even to be bothered beginning”. This mindset is a major impediment. Contrary to perceptions, Kenney et al state that it is not technology but “the lack of organizational will and way” that is the greatest inhibitor to digital preservation.
Despite continuing to overlook digital games, existing cultural institutions and archives are well positioned to undertake software preservation work. They often enjoy legislative exemption from the prohibition against making copies of works, an activity that is part and parcel of software preservation. That cultural institutions have a significant and decidedly more hands-on role to play in software – specifically game – preservation should not be viewed as an impost, but as an opportunity. They have much to gain, not the least of which is the opportunity to reach new audiences. At a minimum, institutions need to use their position to coordinate preservation efforts by interested and talented collectors, researchers, industry and other participants.
The New Zealand situation In 2004, when I began my historical research, inquiries to the professional archival and collections community via the Museums Aotearoa listserv revealed that, with only a few minor exceptions, no cultural institutions were collecting digital game hardware or software. At that time, no one seemed to have any expertise in software archiving, or plans to develop this. I have written elsewhere of how digital games fell between the cracks of what different cultural institutions felt they should be collecting (Swalwell 2007b). Since my initial inquiries, things have improved, albeit very slowly. Now, there is at least an evolving discourse about digital heritage, even if there is still no action being taken on software. In 2004, the Government released its draft Digital Strategy, followed shortly thereafter by a Digital Content Strategy, which contained provision for a National Digital Heritage Archive (NDHA), to be set up by the National Library of New Zealand. Digital Strategy 2.0 has just been launched. The cultural and historic importance of digital or electronic publications had previously been recognised by the National Library Act (2003), which facilitates legal deposit for electronic documents. The NDHA is intended to keep such material alive over time, in a trusted digital repository. The Public Records Act (2005) similarly recognises the significance of electronic records to the nation’s history, though primarily for ensuring good governance (see also Mitchell 2008: 63). Despite these developments, it is still the case that no institutions have responsibility for collecting and/or archiving early software. The situation is similar for digital art, as Lissa Mitchell notes in a recent essay (2008).
NZTronix Aware that digital preservation work would require the talents of a multidisciplinary team, I set about trying to interest people with the necessary expertise at my university in becoming involved in such a project. Gradually, over a period of months, a team of people from a range of disciplinary backgrounds formed, stabilising in 2006 with a group of five researchers: computer scientists Ian Welch and Stuart Marshall, Susan Corbett, an intellectual property lawyer, Rachel Lilburn, an archivist, and myself. Collaborating with game collectors, our team proposed to bring our legal, computer science, archival, and digital humanities expertise to bear on some of the challenges currently preventing conservation of this content.
We decided to work towards a pilot preservation project on software for the Sega SC3000 microcomputer. There were three main reasons for this decision. First, collectors’ efforts have ensured that good documentation exists. Private game collectors are well organised in New Zealand, running a network at www.nzgc.co.nz. Two collectors with a particular interest in software written for the Sega SC3000, Aaron Wheeler and Michael Davidson, have assembled a list of locally published software, giving us comparably more information about the software written and published for this computer. The (2008) list shows some 300 titles of tape software, not counting cartridges, disks, or typed-in programs. Second, this quantity of software makes this system historically significant for New Zealand. Grandstand published a good portion of the software themselves, a reasonably common practice in the 1980s, intended to stimulate the market (Campbell-Kelly 2003:224). The Sega’s limited distribution in other English-speaking countries puts the company’s decision to publish software in context: if the Sega was to be a serious competitor of Commodore and other systems, it was imperative that there be software available for users to purchase. It is further known that the system supported a (Grandstand initiated) magazine, Sega Computer, for a number of years; a 1986 issue listed the contacts for fifteen User Groups across the length and breadth of New Zealand, pointing to a community of ex-users. Third, the software for the Sega has not received any archiving attention to date (as opposed to, say, the Software Preservation Society’s work with Amiga software). Wheeler has digitised a significant number of software titles from magnetic tape, creating .wav files and ensuring that these are able to be loaded onto the original hardware; these are, however, not currently able to be run in emulators.
In 2007, our institution, Victoria University of Wellington, granted funds for a pilot project focused on conserving early examples of this Sega software. The overarching project aims were: to conserve select examples of early local computer game software; to make this reinvigorated digital content available to the wider community, with the permission of relevant rights holders; to demystify the process of software archiving, demonstrating that it is possible to intervene to enhance the longevity of ‘born digital’ content; and to raise awareness of the nation’s unique digital cultural heritage and the need to preserve this. There were two substantive parts to the project, a Sega conservation pilot and an Early Software Database. These were designed to develop and trial solutions to three of the main challenges of software preservation: the legal, the technical, and the informational/media historical. I will detail each of these in turn.
Sega Conservation Pilot The aim of the Sega conservation pilot was to emulate and port one software title to a contemporary mobile device, with the permission of rights holders. This would involve: obtaining permissions from all relevant intellectual property rights holders; determining how to best document, store, and emulate the software; and how to conserve the user-experience with the software. To address the legal challenges first, our plan was to locate the owners of intellectual property in Sega game titles, and to secure non-exclusive, non-commercial licences to enable the use, copying and dissemination of titles. Corbett drew up a licence for this purpose (2007c). Using Wheeler and Davidson’s list of Sega SC3000 software, and with Lilburn’s advice on archiving practices, Corbett oversaw the pursuit of information that would determine ownership of assets: searches were undertaken through Companies Office records, telephone directories and the internet, and an appeal for information placed in the newsletter of the New Zealand Computer Society. A key objective was finding out exactly why Grandstand was struck off the Companies Office register in June 2000, as it was thought this would provide some indication as to what may have happened to the remaining assets and therefore any copyrights Grandstand may have owned in the software they published. Contact was made with the former Director of Grandstand, and contact attempted with all the accountants and solicitors who acted for the company when it was wound up. Unfortunately no one could recall anything about the copyright.
We had originally been hoping to trace all IP in three game titles and were targeting the titles published by three companies (Grandstand, Scorpion Software/Flexisoft, and Poseidon). However, we were not able to obtain a single licence agreement this way. The challenge is not so much in finding the people who wrote or published titles, though this can be difficult, more than twenty years later. Nor, interestingly, is gaining their support for a preservation project particularly difficult in most cases. The hardest part is determining whether they still legally own copyright. In many cases, the way in which a company was wound up means that ownership of IP is not able to be determined. The other two companies were not even registered with the Companies Office, meaning the paper trail went cold early on. Thus, despite knowing who published a title and receiving their in-principle support for our project, the paperwork does not conclusively support a determination of intellectual property ownership. This was the case in regards to the many titles published by Grandstand; these are, in effect, in a legal limbo.
While some may feel that securing the probable owner’s support for a preservation project is good enough, we adopted a cautious position as having a watertight legal basis for software preservation is very important, particularly to institutions who, we anticipate, will be the ones undertaking this work in the future. Without this, there is always the chance that another party will come forward and claim ownership of intellectual property. Our inability to ascertain ownership highlights a significant hurdle to preservation of digital content (see also Corbett 2007a, 2008). Like many other jurisdictions, New Zealand needs an orphan works clause, a point that Corbett made in her representations to the Commerce Select Committee in its 2007 review of the Copyright Amendment Bill (Corbett 2007b). No such provision was introduced in the final legislation (the Copyright (New Technologies) Amendment Act 2008). Corbett’s documented attempts to establish ownership have been lodged in a new VUW Library collection devoted to Early New Zealand Computing, should such an amendment open the way to software preservation where an owner cannot be located.
Less orthodox methods produced some successes in terms of concluding licences with IP owners. Media historical research (detailed below) resulted in contact with an author of a 1980s Sega SC3000 game. A research assistant also searched local computing magazines from the period, looking for mentions of self-published or hobbyist software titles. Such archival methods, coupled with ‘word of mouth’ referrals from the public, resulted in licence agreements with two individuals: Warwick Stubbs, for his game “Astro-vill”, and John Perry for his game “Harbour”. Significantly, neither of these games had been commercialized, which meant determining ownership was comparably straightforward. While the code for “Harbour” was originally published in a 1984 issue of Computer Input magazine – which carried a standard clause that all code submitted became the property of the editor – the editor has no objections to our use of the code for our project, and the 25 year publisher’s copyright is due to expire in 2009.
While these legal issues were being sorted through, Welch and Marshall were working on the technical challenges. Our proposal, as above, was to emulate and port an early software title to a contemporary mobile device. Emulation involves getting software to run on contemporary hardware by using a program (an emulator) that simulates the function of historic hardware. To “port” a piece of software involves getting it to run on a system it was not originally written for. Emulators currently provide the main way for software whose hardware is obsolete to run (emulation of hardware). Though there are Sega emulators in existence (most notably MEKA), the software digitized by Wheeler is not currently able to be run in these, because it is not in a binary format that can be read by the emulators.
We were planning to both emulate and port a software title so as to experiment with emulation, as well as to make the revived software available for people to play on a mobile device. Porting to a mobile was thought to suit the type of content on which the project focused, as well as having benefits for the broader project. Retro games are, in theory, well suited to mobiles: the minimal detail in early game graphics is not seen as a lack when played on a small screen.3 Making historic content available for download to mobile devices would also be useful, we felt, in raising the profile of early games and awareness of their plight. Not only would people be able to play the game, they would also, in a sense, be able to carry around a piece of digital heritage with them.
One of the acknowledged problems with emulation is that emulators require upgrading each time a new operating system is released, to ensure compatibility. A critical question, then – and one with which we were concerned – is how to implement an emulator that resists obsolescence and is portable. Theoretically, this problem could be circumvented if emulators were implemented using a virtual machine: that way, when new hardware or operating systems were released, only the virtual machine would need upgrading, rather than each individual emulator. It was thought that Java would be a good choice for a virtual machine because it is a commercially supported, relatively mature language with a strong community of users, with good portability across platforms. This idea of adding a ‘layer’ mediating between emulator and contemporary device would be one we revisited as we modified our strategy.
The legal and the technical threads of our project were always closely interlinked, as was evident from our aim to emulate and port a software title legally. However, up to this point it had been possible for the two research strands to proceed parallel to each other. Their interrelation was to become more pronounced. Not only were the challenges of determining IP ownership causing delays to the technical part of the project, but when Welch and Marshall began probing how the Sega software worked, they discovered a removable cartridge, labelled “Sega BASIC Level IIIB”. This reads programs written in BASIC and runs them; anyone wanting to run early BASIC games or applications in an emulator would need to have access to both the title and this BASIC cartridge. The IP associated with the Basic cartridge – which was clearly labeled with Sega® logos – was not something we had previously considered. Corbett concluded that it would be protected. Despite attempts to contact Sega management in Japan, with Japanese game preservation colleagues acting as intermediaries, we were not able to request permission to use this. Partly in response to their survey of emulators, and partly in light of needing further permissions, Welch and Marshall decided to try a different approach. They decided to develop their own code translator for the source code.
From Emulation to Translation While writing code translators is not a widely understood activity, it has long been a recognized strategy in computer science: the cfront compiler was widely used in the past to translate C++ code into C code, and programming language translators is how founders of Microsoft, Bill Gates and Paul Allen, started out (Campbell-Kelly 204-5). Given their previous thinking that Java would be a good choice for a virtual machine, Welch and Marshall began to work on a code translator using Java. Together with Honours student, Vipul Delwadia, they developed a prototype cross compiler capable of compiling a subset of BASIC (the Sega Genesis assembly language) into Java bytecode, for execution on a mobile phone. (The change of platforms to Sega Genesis was due to delays in getting permission to work on SC3000 games.) Delwadia calls the converter “SGJ”, short for Sega Genesis Java. While the prototype is currently only working for Sega Genesis BASIC, versions for Sega SC3000 BASIC and Commodore 64 BASIC are in development. The prototype tool is available under an open source licence on the NZTronix website.
An overview of the translation process can be seen in Figure 1. A compiler called BasiEgaXorz was used which takes programs written in its custom dialect of BASIC and compiles them into ROMs which can run on the Sega Genesis (Mega Drive). These will run on both the original console and emulators alike. SGJ then takes the assembly source code and translates it into Java source code (Delwadia 2007). A demonstration program called “99 Bottles” was borrowed (from a website that hosts versions of the song “99 Bottles of Beer” written in different programming languages),5 and translated with the SGJ prototype tool. “99 Bottles” was chosen because it nicely combined a number of key functions being implemented, including sprites and tiles. “99 Bottles” has been successfully run on both a Palm Treo mobile phone and a Nokia N95 phone using SGJ (see Figure 2). While inserting a code translator between source code and contemporary device is an elegant solution, we acknowledge that the translation to Java may only be part of a longer term solution. It does not provide a full answer to the challenges of software preservation. (Such a ‘magic bullet’ would probably not be desirable; having a range of approaches within the preservation community is a strength.)
Several extensions to the SGJ prototype are envisaged and work was begun on these at the end of 2007. Joypad handling was implemented, and the prototype tested with a few sample programs demonstrating up/down/left/right/fire/etc. This is significant because “99 Bottles” is a linear program, whereas games are interactive programs. The objective is to adapt the prototype to translate the Sega SC3000 BASIC code so it can be applied to the two Sega SC3000 games for which we have obtained licences. This work was unfortunately not able to be completed by the end of the grant. It will be followed up at a later stage.6
I first saw “99 bottles” functioning on the Nokia during one of our team’s videoconferences. At the time I was simply pleased to see it working. My aesthetic response was delayed until later that night, when I found myself repeatedly reliving the look and feel of the blocky, 1980s-style graphics moving around on the small screen. The incongruity of retro graphics on a small LCD was exhilarating. We embrace the strangeness of this update, rather than attempting to disguise our intervention, as I explain in the final section.
Media Historical Challenges The list of titles that Wheeler and Davidson have amassed in their years of collecting points to a remarkable – and hitherto unknown – level of local software development for the Sega. This begs the question: what software was developed locally for other computer systems? At the very least, my research indicated that there were high levels of hobbyist activity around the Commodore, Amiga, and TRS-80 systems in the 1980s. I set out to discover what ‘New Zealand software’7 was written for other early computer systems.
While on the face of it, this was unconnected to the legal and technical aspects of the project, this media historical research was intended to lay the foundations for future software collecting and preservation. Local cultural institutions, beginning to show an interest in software, need guidance regarding the titles that are most historically significant, so they can prioritise their efforts (Zabolitzsky 2002:8). Yet no one has a full grasp of the ‘universe’ of locally written software, and there is a lack of records. I undertook to scope both the range and breadth of software that was written locally for microcomputers during the 1980s and 90s. Information would be collected on all types of software (early art projects, utilities, business software, games, etc), written for all makes of microcomputer. During 2007, I developed and prototyped a method for gathering information on New Zealand software from the community, over the web. Community and fan knowledge are often overlooked primary sources. I asked people to contribute what they knew about early software as a way of tapping into this knowledge.
Launched in August 2007, the “Early New Zealand Software Database” consists of a web form, database, search tool and backend interface. Developed by Sarah McKenzie, the site is written in PHP and utilises a MySQL database to store information submitted by users. Information is collected in 9 fields: Title, Year, Author/Developer, Publisher, System, Description, Notes, with the two remaining fields requiring the name and email address of the person contributing the information (these details are not publicly visible). Title and Description are the only mandatory substantive fields. This is to encourage those who only know, or can only recall, partial details to contribute these. A “Do you know more about this entry?” link invites others to help complete entries. There is a facility for uploading files and source code, where this is possible, and nominating Creative Commons licences for software and other assets, again, where appropriate. A media release and subsequent interview on Radio New Zealand National helped to raise awareness of, and elicit the first contributions to, the Database. At the time of writing, the Database has entries on 47 titles. While this number is modest, I expect that its existence will continue to attract interest and that this number will grow over time. In the medium term, it should provide a basis for developing criteria against which the historic significance of particular software genres, titles and systems can be evaluated, enabling targeted acquisition and preservation.
These titles help to assemble a richly textured picture of software writing activity in 1980s and 90s New Zealand, ranging across commercial, educational, artistic and hobbyist spheres. This is unusual: many histories of software are written from a single perspective, such as media arts, games, or business history, and attend only to software in that ‘class’. For example, a business history such as Campbell-Kelly’s From Airline Reservations to Sonic the Hedgehog focuses on companies and titles that became industry ‘leaders’ (and those that went under in the process), with little on titles that did not achieve significant market penetration. The only written account of the New Zealand software industry that I have been able to locate from the period is also business focused: commissioned by the Department of Scientific and Industrial Research, the focus is firmly on those industries capable of earning export dollars for New Zealand’s cash strapped economy (Kaiser, 1985). By contrast, the Database includes information on: shareware (eg. Joseph White’s “TRI” (1993)); ports (eg. the series of titles Sean Fausett and Nick Westgate ported from the BBC to the Apple II for Softime (NZ) Limited, published by Sherston Software (UK)); titles that were locally oriented or produced in low volume (eg. the game “Maze”, one of 6-8 games contained on a cassette for the ZX81, written by John Barton of Timaru; and the “Scheduling System”, by Alec Utting, originally written for North Harbour Netball). Titles that were commercially successful include “Lockie’s Dispensing Software”, for New Zealand and Australian Pharmacies, of which more than $1 million worth was exported to Australia, and the games published by New Zealand’s Acid Software. Custom written training software for Medical and Dental students at the University of Otago is included, as are a variety of titles which were never published, either because they were hobbyist creations, written by juniors (eg. “Twist-a-Plot’s Friend” (1986) written by Michael Clark, Form 4, Katikati College), or designed as conceptual art pieces, such as Stephen McGregor’s “Incorrect Sums” and “Before the Day of Wrath” (c. 1984). In these, McGregor seeks to ‘humanise’ the computer by making it add incorrectly and read poetry.
Knowing about these software creations considerably enriches our understanding of the early software scene, and the ways in which people came to incorporate microcomputers into their everyday lives. While most of them did not make a financial splash, these titles fill out a picture of the uses that people found and/or conceived for computers. This is valuable information, supporting finer grade historical analyses. While business software is clearly important, a lot more work has been done on this; as a result, it is better remembered and, one would think, these titles are more likely to be documented and therefore preserved (though Zabolitzsky suggests this is not always the case). The uses invented for computers in the home or community context, by contrast, are generally not well remembered. Indeed, the wide range of uses to which early home computers were being put was probably not well appreciated in the 1980s: the precise way in which the sporting competition draw was arrived at, for instance, likely remained opaque to most people.
Variation and fidelity In this final section, I reflect on our team’s preservation initiatives, particularly considering the status of preserved software compared to its ‘original’. I focus on the issues of ‘authenticity’, losses and gains, and the limited usefulness of the art historical notion of an original. As a network that has been actively theorizing what it is to emulate media art, the Variable Media initiative is an important reference point. While our material and approach differ, the debates that have been aired and publicized through the network have contributed to the development of discourses about media preservation. As discussed, our team abandoned (technical) emulation in favour of a strategy of translation. Emulation is, however, the term used by the Variable Media Network to refer to “recreating the look and feel of a specific thing, of a specific object or installation, by other means” (Depocas et al, 2003:125). In this, more inclusive sense, our strategy can be considered as an emulation, and I will refer to it as such in the remainder of the article.
The fidelity of an emulation is a key issue in digital preservation discourse: how an emulation differs from its original, the acceptability of any differences, and whether these are in keeping with an artist’s intentions are central questions asked when emulating media art. Such questions, and the assumptions that go with them, are not easily translatable to game preservation. Inasmuch as emulation is concerned with re-creating an original, it is assumed that it is possible and meaningful to identify one. This art historical notion is, however, problematic in relation to digital games. The term’s linguistic centrality, and some of the problems it gives rise to, are evident in Graeme Weinbren’s comments on his and Roberta Friedman’s artwork, “The Erl King,” at the 2004 “Echoes of Art” symposium. More than once, Weinbren asserts that there is no original (in that each installation of the work is a variation), yet he also refers to how it was in the original (Variable Media Network). Notwithstanding these shared difficulties, the notion of an original is particularly problematic and inappropriate for a discourse on games preservation.
A logic of variation is at the core of both the production and the experience of 1980s digital games. Variation was the norm. There have long been different ways of experiencing a game (the release of arcade titles for home consoles is but one example of this). Added to this, the variety of home computer systems meant that many different versions of the ‘same’ game existed, that is, games that appeared the same, but which differed at the level of code. Early game producers were no respecters of originality, and ports and clones were a central feature of the games business, enabling games written for one system to also be sold to users of other systems. Users also wrote their own versions of popular game titles (Swalwell 2008). Such ports and clones might be thought of as ‘copies with a difference’, except that this again implies an original.8 While the mass production of games and the circulation of remarkably similar features and gameplay in early games suggest simulacral qualities, significant variation also exists amongst these same artifacts. As hardware (and probably software too) ages and decays, it begins – somewhat perversely – to take on individual qualities. Each material artifact has its own ‘provenance’ (for want of a better word) in its history of care and abuse – whether console, arcade machine or early computer. These factors suggest an ambivalent relation between games and art preservation.
Differences between an emulation and an original are frequently discussed in terms of ‘losses’ or ‘gains’. These are often a source of anxiety, particularly where obsolete technology is being emulated: losses occasion lament, while gains are unwelcome. The discussion of degrees of fidelity by “Echoes of Art” delegates – including Jill Sterrett, Carol Stringari and Pip Laurenson – usefully complexifies this binary, by acknowledging that the criteria for successful emulations will vary with different projects or pieces. As Jon Ippolito notes: I think we have to fight the fantasy that we will be able to have everything; that we will have the original experience, the original cultural context, the original equipment. We have to choose for each thing what we are most interested in. Different people and different institutions may make different choices. (Variable Media Network)
Our team’s criteria for success were centred on raising awareness about preserving the born digital, in part by making reinvigorated content available to the community. Fidelity of the preserved software was important to us (particularly at the level of the code), as was conserving the user experience with the software. Emulating the look and feel of the original hardware was not attempted because of the decision to distribute to mobiles. (While an adaptation that worked with the original hardware is certainly conceivable, our approach derives from the recognition that sometimes, simply preserving games is not enough: conservation measures are required.) The look and feel of the original hardware is the major ‘loss’ in our project. One of the gains is in the different aesthetics of the Nokia mobile phone display, but this is not necessarily unwelcome. Indeed, it raises the larger question of whether some gains add something of interest or value to an emulation. Clearly the experience of playing a game on a mobile is very different from playing the same game on an early home computer or console (and this is a subject on which we intend conducting further, specific research.)9 While it might be unpopular with purists, I suggest the contemporary element provides a productive incongruity, adding a source of ‘friction’ to the contemporary user’s experience of historic content, a reminder that all is not as it once was.10
As an experiment in digital games conservation, our project departs at certain points from existing preservation wisdom. We believe that being able to play early digital content mitigates the loss of some aspects of the original game experience. Bringing early content to a contemporary platform fulfils our aims of publicizing the plight of early local game software, and of showing that the challenges of digital game preservation are surmountable. In closing, I will touch on two points that stem from this.
First, bringing back early games software and making it accessible for a wide audience is a useful step in building consensus and support for software preservation. Recalling discussions on software preservation in the U.S. from as early as 1986, Kaplan writes that the potential value of an archive was a central issue: early on, “participants simply could not identify a solid user base of any justifiable proportion” for preserved software (Kaplan 2002:2). Games have a definite advantage over other types of software here: they are popular, and playing them is pleasurable. This potentially makes them an attractive target for preservation efforts. Making early games available for a new generation to play is, then, likely to have both indirect and direct benefits for software preservation more generally.
Second, distributing examples of revived historic software effectively makes heritage something that is common. Ubiquitous mobile devices and networks offer cultural institutions a readymade platform for distributing and democratizing (at least some) digital cultural heritage material. Rather than this having to remain within the walls of the physical institution, anyone who wants to can carry around a piece of digital cultural heritage with them. Making historic software available for download allows users to develop their own relations with this content, facilitating, in Walter Benjamin’s words, “a unique experience with the past” (1940:254).
1. Some dimensions of the problem are also specific to games. For instance, elsewhere, I have argued that games need to be included in discourses on the historical changes in the visual, rather than quarantined from other forms of art and culture, so that games’ novelty is actually noticed, and appropriately valued (Swalwell 2007b). 2. The New Zealand case study highlights the diversity that exists across other national or regional contexts, such as the former East Germany, Brazil, Turkey, Sweden, and Russia. Some known examples of local game artifacts include: the Brazilian text adventure “Amazonia” (1983), in which “the goal is to survive an air accident and escape from the dangers of the Amazon jungle” (Game Brasilis); ex-Soviet era arcade games being resuscitated by Russian students (Zaitchik, 2007); and the 1992 Turkish game “Hancer” (The Dagger), a strategy game offering insights into Turkish culture of the time (Yilmaz and Cagiltay 2005). 3. Swalwell and Loyer (2006) provides a detailed account of New Zealand’s game history. 4. One possible issue is with the small size of sprites relative to the background. One of our licensors saw this as a potential problem: “the sprites…on the Sega are pretty small, so the planes etc on the game are tiny compared to the background – very dated! You might need to have them expanded if you want to get it to work on a small screen such as a phone.” (private communication, 17/3/08) 5. For background on this humorous endeavour, see Schade. 6. Related to this, they now have a working, secondhand SC3000 computer, and have run “Astro-vill” on this. 7. This term is intended simply to describe production location. I am not suggesting there is any inherent ‘New Zealandness’ to the software. 8. The notion of originality may be appropriate in well documented cases of innovation, for instance, by game auteurs, but this is not my concern here. 9. Marshall intends comparing the two experiences of hardware and software, in terms of the look and feel. He will also create video documentation of the software playing on the original hardware and on the mobile. Delays owing to changes in the project’s strategy meant this work has not yet been undertaken. It is worth noting that a contemporary player will never have the same experience as a 1980s player as, apart from anything else, the player has changed (Variable Media Network). There is simply not the same awe and excitement as characterized early encounters. This is not to say one should not strive for authenticity, but that there are limits to achieving this. 10. In general terms, though, it is worth considering whether and how users know that their experience of content differs from that which would have been available on the original hardware. Video documentation (which then itself requires preservation) would provide a basis for this.
Works cited Benjamin, W. (1992 [1940]) ‘Theses on the Philosophy of History’, in H. Arendt, ed., Illuminations, Harry Zohn, trans., pp. 245-255, London: Fontana. Cameron, F. and S. Kenderdine (eds) Theorizing Digital Cultural Heritage: A Critical Discourse, Cambridge, MA: MIT Press. Campbell-Kelly, M. (2003) From Airline Reservations to Sonic the Hedgehog: A history of the software industry, Cambridge, MA: MIT Press. Corbett, S. (2008) ‘What if Object Code had been Excluded from Protection as a Literary Work in Copyright Law? A New Zealand Perspective’, Michigan State Law Review, Spring, issue 1, pp. 173-197. Corbett, S. (2007a) ‘Digital Heritage: Legal Barriers to Conserving New Zealand’s Early Video Games,’ New Zealand Business Law Quarterly, vol. 13, no. 5: 48-71. Corbett, S. (2007b) ‘Submission on the Copyright (New Technologies and Performers’ Rights) Amendment Bill’, Commerce Select Committee (NZ Parliament), 15 February, available online at http://nztronix.org.nz/publications.php. Corbett, S. (2007c) “Copyright Licence”, available at http://nztronix.org.nz/publications.php Delwadia, V. (2007) ‘SGJ: Sega Genesis Java’, Victoria University of Wellington, unpublished Honours thesis, available online at http://homepages.mcs.vuw.ac.nz/~vipul/final-report.pdf Depocas, A., J. Ippolito, C. Jones (eds) (2003) Permanence Through Change: The Variable Media Approach, New York, Montreal: The Solomon R. Guggenheim Foundation/The Daniel Langlois Foundation for Art, Science, and Technology, Montreal. Available online at http://www.variablemedia.net/e/preserving/html/var_pub_index.html DiGA (2004) ‘Short Statement Concerning the ‘Referentenentwurf für ein zweites Gesetz zur Regelung des Urheberrechts in der Informationsgesellschaft’ (Consultant’s Draft for a Second Law Governing Copyright Law in the Information Society) with Regard to Complex Digital Cultural Artefacts (Computer Games)’, The Digital Game Archive. Available online at http://www.digitalgamearchive.org/projects.php Digital Strategy: A Draft New Zealand Digital Strategy for Consultation (2004) Wellington: New Zealand Government. Digital Strategy 2.0 (2008) available at http://www.digitalstrategy.govt.nz/Digital-Strategy-2/ Game Brasilis: Catálogo de jogos eletrônicos brasileiros, Senac: São Paulo, 2003. Game Preservation SIG (2008) ‘Digital Game Canon’, http://www.igda.org/wiki/Game_Preservation_SIG/Digital_Game_Canon Gamesnetwork list, archives available to subscribers only at https://listserv.uta.fi Gielens, J. (2000) Electronic Plastic, Berlin: Verlag, also available www.handhelden.com/_book Gould, S. and M-T Varlamoff (2000) ‘Digital disappearance’, Connexions: The UNESCO Courier (October): 46-7. Heide Smith, J. (2007) “A canon of significant games”, January 2, http://jonassmith.dk/weblog/a-canon-of-significant-games/ Herz, J.C. (1997) Joystick Nation: How Videogames Ate Our Quarters, Won Our Hearts, and Rewired Our Minds, London: Little, Brown & Co. Kaiser, M.S. (1985) ‘A New Industry for New Zealand’, Looking Back to Tomorrow: Reflections on twenty-five years of computers in New Zealand, pp. 147-162, Wellington: NZ Computer Society. Kaplan, E. (2002) ‘A Response to ‘Preserving Software: Why and How’’, Iterations: An Interdisciplinary Journal of Software History, vol. 1 (September 13): 1-3, available online at http://www.cbi.umn.edu/iterations/kaplan.html Kenney, A.R., N.Y. McGovern, R. Entlich, W.R. Kehoe, E. Olsen, E. Buckley (last updated December 2007) ‘Digital Preservation Management: Implementing Short-Term Strategies for Long-Term Problems’, http://www.library.cornell.edu/iris/tutorial/dpm/introduction.html Kent, S.L. and F.D. Schwartz (1997) “Super Mario Nation,” American Heritage, vol. 48, no. 5, (September): 65-74. Lowood, H. (2004) ‘Playing History with Games: Steps towards historical archives of computer gaming,’ paper presented at the Electronic Media Group, Annual Meeting of the American Institute for Conservation of Historic and Artistic Works, Portland, Oregon, June 14, available online at http://aic.stanford.edu/sg/emg/library/pdf/lowood/Lowood-EMG2004.pdf. Mitchell, L. (2008) ‘New Data: Curating and archiving digital art in New Zealand’, in Stella Brennan and Su Ballard (eds) Aotearoa Digital Arts Reader, pp. 62-68, Auckland: Clouds/ADA. No author given (1984) ‘Sega’s Young Programmers: Today New Zealand, tomorrow the world’, Sega Computer, (1st issue) p. 21. Oppenheimer, R. (2005) ‘Regional Media Arts Histories and their Contributions to Electronic Arts’, Leonardo Electronic Almanac, 13:5. Schade, O. (n.d.) ‘History’, 99 Bottles of Beer: One program in 1216 variations, http://99-bottles-of-beer.net/info.html Sellers, J. (2001) Arcade Fever: The Fan’s Guide to the Golden Age of Video Games, London, Philadelphia: Running Press. Sterling, B. (2003) ‘Digital Decay’, in A. Depocas, J. Ippolito and C. Jones (eds) Permanence Through Change: The Variable Media Approach, pp. 11-22, New York, Montreal: The Solomon R. Guggenheim Foundation/The Daniel Langlois Foundation for Art, Science, and Technology, available online at http://www.variablemedia.net/e/preserving/html/var_pub_index.html Swalwell, M. and M. Davidson (forthcoming) ‘Malzak’, in M. Bittanti and I. Bogost (eds) Ludologica Retro: Volume 1: Vintage Arcade (1971-1984), available online at http://nztronix.org.nz/malzak.php Swalwell, M. (2008) ‘1980s Home Coding: the art of amateur programming’, in Stella Brennan and Susan Ballard (eds), Aotearoa Digital Arts: New Media Reader, pp. 193-201, Auckland: Clouds/ADA. Swalwell, M. (2007a-) ‘Early New Zealand Software Database’, http://www.nztronix.org.nz/main.php Swalwell, M. (2007b) ‘The Remembering and the Forgetting of Early Digital Games: From novelty to detritus and back again’ Journal of Visual Culture, vol. 6, no. 2: 255-273. Swalwell, M. and Erik Loyer (2006) “Cast-offs from the Golden Age”, Vectors: Journal of Culture and Technology in a Dynamic Vernacular, issue 3, April, http://www.vectorsjournal.org Variable Media Network (2004) ‘Echoes of Art: Emulation as a preservation strategy’, symposium transcript, available at: http://www.variablemedia.net/e/echoes/morn_text.htm#contemp Wheeler, A. and M. Davidson (updated 2008), ‘SC3K Software List’, http://homepages.ihug.co.nz/~atari/SC3K08.html Wilkinson, K. (2005) ‘A Call to Arms: Computer and video game preservation projects unite’, International Game Developer’s Association, Game_Preservation SIG mailing list archives, October, http://six.pairlist.net/pipermail/game_preservation/2005-October/000047.html Yilmaz, E. and K. Cagiltay, ‘History of Digital Games in Turkey’, DiGRA Conference: ‘Changing Views: Worlds in Play’, 17-20 June 2005, DiGRA Proceedings, Available online http://www.digra.org/dl/ Zabolitzky, J.G. (2002) ‘Preserving Software: Why and How’, Iterations: An Interdisciplinary Journal of Software History, vol. 1 (September 13): 1-8, http://www.cbi.umn.edu/iterations/zabolitzky.pdf. Zaitchik, A. (2007) ‘Soviet-Era Arcade Games Crawl Out of Their Cold War Graves’, Wired, June, http://www.wired.com/gaming/hardware/news/2007/06/soviet_games
Melanie Swalwell
The final, definitive version of this article was published in a special issue of Convergence, vol. 15, no. 3, in 2009. It is available from Sage
Back to homepage
At the 2007 Game Developer’s Conference, a panel of five speakers from industry and academia proposed a Digital Game Canon, consisting of ten games of historical and cultural significance “that everyone should play”. The games were: “SpaceWar!”, “Star Raiders”, “Zork I: The Great Underground Empire”, “Tetris”, “Sim City”, “Super Mario Brothers 3”, “Civilisation I/II”, “DOOM”, “Sensible World of Soccer”, and “Warcraft I/II/III” (Game Preservation SIG). The Canon had a mixed reception, as several posts to a thread on the Digital Games Research Association’s ‘Gamesnetwork’ list show. Initially unrelated to the GDC panel, list members were responding to a request for suggestions for a Game Studies canon; the Digital Game Canon event was introduced to the discussion part way through. Some offered their own list of the most significant game titles ever: Jonas Heide Smith volunteered the Danish Game Researcher Network’s canon (2007). Others demurred to the wisdom of creating a new canon: Barry Atkins asked whether “we really need a ‘best of’ for every academic category”, a response with which I had some sympathy (Gamesnetwork). Nevertheless, as a tactic, the Digital Game Canon was astute: the panel discussion received widespread coverage, including stories in the New York Times and Der Spiegel, and people began discussing the historical significance of games in a way they previously had not. Yet while this GDC session was a watershed, it is worth pausing to consider what this means. It means that 10 titles, chosen by five panelists (Matteo Bittanti, Christopher Grant, Henry Lowood, Steve Meretzky, and Warren Spector), will receive some attention (they are preservation targets in the Library of Congress funded project “Preserving Virtual Worlds”). It is notable that most of these titles were hits, and that, apart from “Tetris”, they were all either products of the United States, United Kingdom, or Japan. With the exception of “Spacewar!”, they are all well known internationally by virtue of their commercial success. This is understandable. However, as it is common to ask regarding canons: what has been left out? What other digital games, and indeed other types of software, might we consider to be culturally and historically significant such that they also warrant preservation? This is a question that needs to be asked repeatedly; the answers to it will, of course, vary depending upon the respondent’s perspective. In this article, my concern is with local software, specifically locally produced computer game software.
The question about local game production has yet to be asked in a scholarly context. Digital game history has usually been told with a focus on the U.S. or Japan, assuming the uniformity of products and reception worldwide (Herz, Kent and Schwartz, Sellers). While early digital games were a global phenomenon, the 1970s and 80s also saw games produced and consumed in specifically local contexts. Remarkably little is known about the local aspects of digital games history. In contrast to the high profile games targeted by the Digital Game Canon, in local contexts it is not always clear what software even existed; consequently, identifying priorities for preservation becomes a challenge.
I became interested in the preservation of local game software during the course of my research into the history of digital games in New Zealand. In 2004, I became aware of a significant quantity of locally written game software from the 1980s. Apart from a few specialist collectors, no one seemed to be aware of this, and no institutions were collecting it. Such a situation – where private collectors are effectively acting as custodians and amateur archivists of the nation’s digital cultural heritage, while institutions charged with collecting, preserving and making a nation’s creative output accessible overlook software – is not unique to New Zealand. While the work of collectors is to be applauded, as hobbyists, they have limited resources for undertaking digital preservation work. These circumstances prompted me to assemble a multidisciplinary research team that would be capable of bringing a range of expertise – legal, technical, and media-historical – to bear on these titles’ preservation. The project that is discussed here was conducted with the intention of advancing the software preservation agenda, in the hope that some of the early digital heritage of New Zealand – content that is unique internationally – will endure.
This article reports on our NZTronix team initiatives regarding this early computer game software. It begins by briefly introducing the game preservation landscape, before outlining the case for the preservation of local game software. The legal, technical, and media historical challenges addressed by the team in a pilot software preservation project are discussed next, and the mixed successes we enjoyed. The final section considers some of the key concerns in digital preservation discourse, including authenticity, fidelity, and variation, highlighting some of the points where digital game preservation diverges from other preservation initiatives.
Game Preservation: Digital decay
The need to urgently preserve early digital games is an internationally recognised one (for summaries see Wilkinson (2005), Lowood (2004), and DiGA (2004)). Briefly, the main issues that threaten digital games’ longevity are: that content exists on degrading media (eg. magnetic tapes, floppy disks) and relies on obsolete operating systems; that the lifespan of hardware is limited because of physical and chemical processes within the chips (DiGA); and that copyright restricts the ability to undertake preservation work. As Wilkinson notes, the wide variety of gaming systems, the different technologies used, and the hundreds or even thousands of games produced for many of them makes this “truly a huge task”. These problems, of outdated hardware, deteriorating software, and intellectual property restrictions make accessing, preserving and distributing early digital games difficult; to do this while remaining within the law is even harder.
Many of the challenges of game preservation are common to other complex digital media objects.1 Consequently, technical interventions to preserve games are likely to also hold benefits for other interactive media, such as new media art. It is curious, then, how little cross-fertilisation there has been, to date, between digital game preservationists and other digital preservation communities such as librarians, media arts conservators, and museum professionals. The work of the vintage game community receives several mentions in the transcript from the 2004 Variable Media Network symposium “Echoes of Art: Emulation as a Preservation Strategy”; however, alongside their admiration for the work of game fans, the Network perpetuates the erroneous impression that the problems of game preservation have largely been solved. This misunderstanding appears to be reasonably widespread (see also Campbell-Kelly, 2003: 272). While the work of many fans, collectors and preservationists is commendable, and playable versions of many games currently exist, many challenges still remain. The legality of downloading most games on the internet is dubious, at best; the quality of preservation work undertaken and the longevity of game titles is also a pressing issue. Bruce Sterling pithily summarises the problems common to the preservation of born digital artifacts in the Variable Media anthology Permanence Through Change: “Bits have no archival medium. We haven’t invented one yet…We have no way to archive bits that we know will be readable in even 50 years. Tape demagnetizes. CDs delaminate. Networks go down.” (2003:20)
There are a number of game repositories and preservation projects currently underway, many in the United States. These include: the Software Preservation Society (formerly Classic Amiga Preservation Society); the Stephen M. Cabrinety collection and the “How they Got Game” project, both at Stanford University; the University of Texas Videogame Archive; the Strong National Museum of Play; and the “Preserving Virtual Worlds” project. In Europe, the work of the Berlin Computerspiele Museum is important, as is that of the related organization DiGA. Media arts conservation projects like the Canadian DOCAM Research Alliance are also worth mentioning here, given the similar challenges faced by those working to preserve games and digital media art.
Game Preservation: Local game software
It is worth pausing to answer the question as to why local software matters? Why should it also be preserved? Why is preserving the titles in an international game canon not sufficient? There are a number of answers to this. Games were one of the key ways in which people became familiar with the then new technology of computers. Playing games allowed users to build confidence with computers. Particularly on early machines, playing a game required a user to first write or type in the game code. This authorial involvement points to a likely diversity in the game software of different cultures. While games generally are significant historical artifacts from the path to our becoming digital, the diversity of local softwares will be important for future historical work on how games and computers came to the rest of the world (that is, beyond the U.S. and Japan).
New Zealand’s software history is remarkable for the numbers of people – both companies and hobbyists – who were involved in writing software, and the numbers of locally published titles (Wheeler and Davidson). While software was rarely developed in total isolation, local structural conditions kept influences from abroad in check. As well as being geographically remote, trade restrictions further moderated influences from the US and elsewhere (see, for instance, Swalwell and Davidson, forthcoming), while government incentives for the electronics industry supported the development of local games production. Inquiring into local conditions of production and reception usefully complexifies existing, taken-for-granted histories of game and software development. It also reveals hidden networks of friendships and influence (Oppenheimer).2 For instance, a Director of the Auckland-based Kitronix company, Ralph Stevenson, allegedly corresponded with the founder of Atari, Nolan Bushnell (circa 1970s), before developing his own version of a television ping pong game. Les Kenyon is another New Zealand based protagonist with links to international game history. After emigrating from the U.K. in 1980, Kenyon set up his business, Grandstand, locally. The ex-Hanimex salesman would go on to be a significant name in Australian and New Zealand games history, as the importer of the Sega SC3000 microcomputer, as well as the distributor of early handheld games in the U.K.3
Despite the rarity of many local game titles and their increasing fragility, they are unlikely to be the (direct) beneficiaries of large preservation projects such as those mentioned above. If local software is to be preserved, local people will need to do it themselves. In the absence of purpose built media museums (Jűrgen Claus cited in Lowood 2004:8), existing cultural institutions are probably best placed to undertake software preservation work. Yet despite an increased interest in digital heritage, games seldom register on their institutional radars. Partly this is because ‘born digital’ forms of heritage continue to be sidelined by the “frenzy to digitize” material artifacts (Gould and Varlamoff, 146; Cameron and Kenderdine); partly, a form of amnesia is operative, albeit one that is occasionally interrupted by nostalgic reminiscence (Swalwell 2007b). Underlying both responses is an acceptance of the ephemerality of electronic media and the inevitability of their decline. Getting comfortable with the imminent loss of cultural heritage materials takes some effort; it does not come easily to curators, archivists, librarians, and others whose professional lives revolve around heritage preservation. Such resignation is the result, I believe, of a widespread perception that software preservation is “too hard” and that “there are too many problems even to be bothered beginning”. This mindset is a major impediment. Contrary to perceptions, Kenney et al state that it is not technology but “the lack of organizational will and way” that is the greatest inhibitor to digital preservation.
Despite continuing to overlook digital games, existing cultural institutions and archives are well positioned to undertake software preservation work. They often enjoy legislative exemption from the prohibition against making copies of works, an activity that is part and parcel of software preservation. That cultural institutions have a significant and decidedly more hands-on role to play in software – specifically game – preservation should not be viewed as an impost, but as an opportunity. They have much to gain, not the least of which is the opportunity to reach new audiences. At a minimum, institutions need to use their position to coordinate preservation efforts by interested and talented collectors, researchers, industry and other participants.
The New Zealand situation
In 2004, when I began my historical research, inquiries to the professional archival and collections community via the Museums Aotearoa listserv revealed that, with only a few minor exceptions, no cultural institutions were collecting digital game hardware or software. At that time, no one seemed to have any expertise in software archiving, or plans to develop this. I have written elsewhere of how digital games fell between the cracks of what different cultural institutions felt they should be collecting (Swalwell 2007b). Since my initial inquiries, things have improved, albeit very slowly. Now, there is at least an evolving discourse about digital heritage, even if there is still no action being taken on software. In 2004, the Government released its draft Digital Strategy, followed shortly thereafter by a Digital Content Strategy, which contained provision for a National Digital Heritage Archive (NDHA), to be set up by the National Library of New Zealand. Digital Strategy 2.0 has just been launched. The cultural and historic importance of digital or electronic publications had previously been recognised by the National Library Act (2003), which facilitates legal deposit for electronic documents. The NDHA is intended to keep such material alive over time, in a trusted digital repository. The Public Records Act (2005) similarly recognises the significance of electronic records to the nation’s history, though primarily for ensuring good governance (see also Mitchell 2008: 63). Despite these developments, it is still the case that no institutions have responsibility for collecting and/or archiving early software. The situation is similar for digital art, as Lissa Mitchell notes in a recent essay (2008).
NZTronix
Aware that digital preservation work would require the talents of a multidisciplinary team, I set about trying to interest people with the necessary expertise at my university in becoming involved in such a project. Gradually, over a period of months, a team of people from a range of disciplinary backgrounds formed, stabilising in 2006 with a group of five researchers: computer scientists Ian Welch and Stuart Marshall, Susan Corbett, an intellectual property lawyer, Rachel Lilburn, an archivist, and myself. Collaborating with game collectors, our team proposed to bring our legal, computer science, archival, and digital humanities expertise to bear on some of the challenges currently preventing conservation of this content.
We decided to work towards a pilot preservation project on software for the Sega SC3000 microcomputer. There were three main reasons for this decision. First, collectors’ efforts have ensured that good documentation exists. Private game collectors are well organised in New Zealand, running a network at www.nzgc.co.nz. Two collectors with a particular interest in software written for the Sega SC3000, Aaron Wheeler and Michael Davidson, have assembled a list of locally published software, giving us comparably more information about the software written and published for this computer. The (2008) list shows some 300 titles of tape software, not counting cartridges, disks, or typed-in programs. Second, this quantity of software makes this system historically significant for New Zealand. Grandstand published a good portion of the software themselves, a reasonably common practice in the 1980s, intended to stimulate the market (Campbell-Kelly 2003:224). The Sega’s limited distribution in other English-speaking countries puts the company’s decision to publish software in context: if the Sega was to be a serious competitor of Commodore and other systems, it was imperative that there be software available for users to purchase. It is further known that the system supported a (Grandstand initiated) magazine, Sega Computer, for a number of years; a 1986 issue listed the contacts for fifteen User Groups across the length and breadth of New Zealand, pointing to a community of ex-users. Third, the software for the Sega has not received any archiving attention to date (as opposed to, say, the Software Preservation Society’s work with Amiga software). Wheeler has digitised a significant number of software titles from magnetic tape, creating .wav files and ensuring that these are able to be loaded onto the original hardware; these are, however, not currently able to be run in emulators.
In 2007, our institution, Victoria University of Wellington, granted funds for a pilot project focused on conserving early examples of this Sega software. The overarching project aims were: to conserve select examples of early local computer game software; to make this reinvigorated digital content available to the wider community, with the permission of relevant rights holders; to demystify the process of software archiving, demonstrating that it is possible to intervene to enhance the longevity of ‘born digital’ content; and to raise awareness of the nation’s unique digital cultural heritage and the need to preserve this. There were two substantive parts to the project, a Sega conservation pilot and an Early Software Database. These were designed to develop and trial solutions to three of the main challenges of software preservation: the legal, the technical, and the informational/media historical. I will detail each of these in turn.
Sega Conservation Pilot
The aim of the Sega conservation pilot was to emulate and port one software title to a contemporary mobile device, with the permission of rights holders. This would involve: obtaining permissions from all relevant intellectual property rights holders; determining how to best document, store, and emulate the software; and how to conserve the user-experience with the software. To address the legal challenges first, our plan was to locate the owners of intellectual property in Sega game titles, and to secure non-exclusive, non-commercial licences to enable the use, copying and dissemination of titles. Corbett drew up a licence for this purpose (2007c). Using Wheeler and Davidson’s list of Sega SC3000 software, and with Lilburn’s advice on archiving practices, Corbett oversaw the pursuit of information that would determine ownership of assets: searches were undertaken through Companies Office records, telephone directories and the internet, and an appeal for information placed in the newsletter of the New Zealand Computer Society. A key objective was finding out exactly why Grandstand was struck off the Companies Office register in June 2000, as it was thought this would provide some indication as to what may have happened to the remaining assets and therefore any copyrights Grandstand may have owned in the software they published. Contact was made with the former Director of Grandstand, and contact attempted with all the accountants and solicitors who acted for the company when it was wound up. Unfortunately no one could recall anything about the copyright.
We had originally been hoping to trace all IP in three game titles and were targeting the titles published by three companies (Grandstand, Scorpion Software/Flexisoft, and Poseidon). However, we were not able to obtain a single licence agreement this way. The challenge is not so much in finding the people who wrote or published titles, though this can be difficult, more than twenty years later. Nor, interestingly, is gaining their support for a preservation project particularly difficult in most cases. The hardest part is determining whether they still legally own copyright. In many cases, the way in which a company was wound up means that ownership of IP is not able to be determined. The other two companies were not even registered with the Companies Office, meaning the paper trail went cold early on. Thus, despite knowing who published a title and receiving their in-principle support for our project, the paperwork does not conclusively support a determination of intellectual property ownership. This was the case in regards to the many titles published by Grandstand; these are, in effect, in a legal limbo.
While some may feel that securing the probable owner’s support for a preservation project is good enough, we adopted a cautious position as having a watertight legal basis for software preservation is very important, particularly to institutions who, we anticipate, will be the ones undertaking this work in the future. Without this, there is always the chance that another party will come forward and claim ownership of intellectual property. Our inability to ascertain ownership highlights a significant hurdle to preservation of digital content (see also Corbett 2007a, 2008). Like many other jurisdictions, New Zealand needs an orphan works clause, a point that Corbett made in her representations to the Commerce Select Committee in its 2007 review of the Copyright Amendment Bill (Corbett 2007b). No such provision was introduced in the final legislation (the Copyright (New Technologies) Amendment Act 2008). Corbett’s documented attempts to establish ownership have been lodged in a new VUW Library collection devoted to Early New Zealand Computing, should such an amendment open the way to software preservation where an owner cannot be located.
Less orthodox methods produced some successes in terms of concluding licences with IP owners. Media historical research (detailed below) resulted in contact with an author of a 1980s Sega SC3000 game. A research assistant also searched local computing magazines from the period, looking for mentions of self-published or hobbyist software titles. Such archival methods, coupled with ‘word of mouth’ referrals from the public, resulted in licence agreements with two individuals: Warwick Stubbs, for his game “Astro-vill”, and John Perry for his game “Harbour”. Significantly, neither of these games had been commercialized, which meant determining ownership was comparably straightforward. While the code for “Harbour” was originally published in a 1984 issue of Computer Input magazine – which carried a standard clause that all code submitted became the property of the editor – the editor has no objections to our use of the code for our project, and the 25 year publisher’s copyright is due to expire in 2009.
While these legal issues were being sorted through, Welch and Marshall were working on the technical challenges. Our proposal, as above, was to emulate and port an early software title to a contemporary mobile device. Emulation involves getting software to run on contemporary hardware by using a program (an emulator) that simulates the function of historic hardware. To “port” a piece of software involves getting it to run on a system it was not originally written for. Emulators currently provide the main way for software whose hardware is obsolete to run (emulation of hardware). Though there are Sega emulators in existence (most notably MEKA), the software digitized by Wheeler is not currently able to be run in these, because it is not in a binary format that can be read by the emulators.
We were planning to both emulate and port a software title so as to experiment with emulation, as well as to make the revived software available for people to play on a mobile device. Porting to a mobile was thought to suit the type of content on which the project focused, as well as having benefits for the broader project. Retro games are, in theory, well suited to mobiles: the minimal detail in early game graphics is not seen as a lack when played on a small screen.3 Making historic content available for download to mobile devices would also be useful, we felt, in raising the profile of early games and awareness of their plight. Not only would people be able to play the game, they would also, in a sense, be able to carry around a piece of digital heritage with them.
One of the acknowledged problems with emulation is that emulators require upgrading each time a new operating system is released, to ensure compatibility. A critical question, then – and one with which we were concerned – is how to implement an emulator that resists obsolescence and is portable. Theoretically, this problem could be circumvented if emulators were implemented using a virtual machine: that way, when new hardware or operating systems were released, only the virtual machine would need upgrading, rather than each individual emulator. It was thought that Java would be a good choice for a virtual machine because it is a commercially supported, relatively mature language with a strong community of users, with good portability across platforms. This idea of adding a ‘layer’ mediating between emulator and contemporary device would be one we revisited as we modified our strategy.
The legal and the technical threads of our project were always closely interlinked, as was evident from our aim to emulate and port a software title legally. However, up to this point it had been possible for the two research strands to proceed parallel to each other. Their interrelation was to become more pronounced. Not only were the challenges of determining IP ownership causing delays to the technical part of the project, but when Welch and Marshall began probing how the Sega software worked, they discovered a removable cartridge, labelled “Sega BASIC Level IIIB”. This reads programs written in BASIC and runs them; anyone wanting to run early BASIC games or applications in an emulator would need to have access to both the title and this BASIC cartridge. The IP associated with the Basic cartridge – which was clearly labeled with Sega® logos – was not something we had previously considered. Corbett concluded that it would be protected. Despite attempts to contact Sega management in Japan, with Japanese game preservation colleagues acting as intermediaries, we were not able to request permission to use this. Partly in response to their survey of emulators, and partly in light of needing further permissions, Welch and Marshall decided to try a different approach. They decided to develop their own code translator for the source code.
From Emulation to Translation
While writing code translators is not a widely understood activity, it has long been a recognized strategy in computer science: the cfront compiler was widely used in the past to translate C++ code into C code, and programming language translators is how founders of Microsoft, Bill Gates and Paul Allen, started out (Campbell-Kelly 204-5). Given their previous thinking that Java would be a good choice for a virtual machine, Welch and Marshall began to work on a code translator using Java. Together with Honours student, Vipul Delwadia, they developed a prototype cross compiler capable of compiling a subset of BASIC (the Sega Genesis assembly language) into Java bytecode, for execution on a mobile phone. (The change of platforms to Sega Genesis was due to delays in getting permission to work on SC3000 games.) Delwadia calls the converter “SGJ”, short for Sega Genesis Java. While the prototype is currently only working for Sega Genesis BASIC, versions for Sega SC3000 BASIC and Commodore 64 BASIC are in development. The prototype tool is available under an open source licence on the NZTronix website.