Legacy Interfaces and Services Offered Online by Major Companies
 |
||
|
Online services with unique interfaces that change regularly and through those changes provide a different experience and different content to their users. |
||
|
Digital Species: Social Media |
Trend in 2023:
|
Consensus Decision |
|
Added to List: 2019 |
Trend in 2024:
|
Previously: Practically Extinct |
|
Imminence of Action Immediate action necessary. Where detected should be stabilized and reported as a matter of urgency. |
Significance of Loss The loss of tools or services within this group would have a global impact. |
Effort to Preserve | Inevitability Loss seems inevitable. Loss has already occurred or is expected to occur before tools or techniques develop. |
|
Examples Interfaces to user-driven social platforms such as MySpace, Bebo, Facebook, Twitter/X, Reddit, LinkedIn, Sina Weibo, Flickr, and social platform API-driven services such as Botometer and Tweetdeck and many others, and superseded policies and terms governing access and re-use of platform data. |
||
|
‘Critically Endangered’ in the Presence of Good Practice Robust and extensive web archives with strong documentation of platform documentation (such as terms of service) over time; search algorithms (where available); ranking and personalization of interfaces; awareness of IPR and the presence of orphaned works. |
||
|
2023 Review This entry was added in 2019 to highlight the configuration of interfaces and, therefore, the ever-changing arrangement and presentation of content. Personalization means that the same query can produce quite different results to different users at the same time; the application of machine learning to behavioral surplus means the same may obtain different results at different points in time. That is over and above the rapid churn in the appearance of web interfaces. There is little appreciation of the implications for the use of online services and the potential for manipulations that arise. Moreover, the digital preservation community, which is historically concerned with data rather than interface, has only rudimentary tools to address this challenge. The 2021 Jury agreed but noted a trend towards greater risk due to security issues posed by hosting legacy technology software and services which have prompted disposal of content imminently without adequate review or selection. The 2022 Taskforce agreed these risks remain on the same basis as before (no change to the trend). The 2023 Council agreed with the Practically Extinct classification and noted an increase in imminence, recognizing that while the need for major efforts to prevent or reduce losses continues, it is now much more likely that loss of material has already occurred, and will continue to do so, by the time tools or techniques have been developed. |
||
|
2024 Interim Review The 2024 Council agreed these risks remain on the same basis as before, with no significant trend towards even greater or reduced risk (‘No change’ to trend). In light of the recent changes to API services by major platforms (e.g. Twitter/X, Reddit), they recommend maintaining the entry separate from ‘Consumer Social Media Free at the Point of Use’ to highlight the risk of loss of access to underlying data. Since 2023, a number of major global platforms (such as Twitter/X and Reddit) have begun restricting researcher access to platform APIs, making API services cost-prohibitive to academic and collecting institutions, or shutting them down completely (Davidson et al., 2023, Summers, 2023). With restrictions to this method of collection, data will be (and likely has already been) lost to the institutions responsible for maintaining access to research data and heritage collections. While this entry focuses on interfaces and the risk of losing the display and functionality characteristics crucial to interpreting social media content, it also encompasses the multi-modal access and terms under which social media platforms make content available. While the algorithms that serve content to a user’s feed or browser-based interpretations of interfaces make a major impact on how content might be interpreted, the terms and conditions, data services, and policy frameworks around platforms also provide crucial context. 2024 has seen increased awareness of social media as a manipulating factor in social life, such as congressional hearings in the US around social media impact on youth, pop culture books such as Cathy O’Neill’s publication The Shame Machine: Who Profits in the New Age of Humiliation (which includes social media amongst a host of other shame-creating industries), and a proposed warning label for teens on social media platforms from the surgeon general. New EU legislation, such as the Digital Services Act, also demonstrates wider awareness of the impact of social media and the need to hold platforms accountable through reliable records of content published by users as well as policies and service documentation. Though Twitter/X and Reddit, for example, still exist and user data has not disappeared at a platform level, access to that data in a way that supports computation access or archiving as scale has been shut down. Without access to data via API, research and analysis methods are severely hampered. Given the fast-paced, ephemeral nature of social media, this means research potential has already been lost and tools crucial for using social media data to support public policy and services have become obsolete. Additionally, the Council recommends further research as part of the next 2025 review, in particular, to highlight the particular risk of loss of legacy platform policies and T&Cs, which are required to interpret datasets or web crawls of social media at a certain point in time. This issue of lack of access to former policies has been raised in academic forums and this should be addressed more explicitly. They propose describing this issue and recommending 1) further research into existing archives of platform information and policy pages (such as in IA collections and LoC 'Business in America' collections) and 2) adding this issue as 'Aggravating Circumstances' for other entries or potentially creating a new entry (but preferably not). |
||
|
Additional Comments The interfaces alone have less impact if they are gone (since many research users are interested in the extracted data). However, it's an important distinction for us to make that we could end up preserving social media data as datasets in the long run, meaning that the look and feel (which serves a different sort of purpose) will be lost. A number of social media researchers in groups like the Association of Internet Researchers (AoIR) may care more about the data, but perhaps it is worth exploring a bit more about their interest in interfaces. Without the interfaces and the underlying software that enables social media platforms, it will be impossible to preserve the look and feel and even meaning of a large portion of content that depends on particular functionality or interface to be accurately or authentically interpreted, including for evidential uses, artworks, design research, and historical/qualitative research. The loss of these interfaces (or lack of any indication of robust documentation by platforms) means a significant gap in the cultural heritage of many communities and even entire nations. For example, some content creators on YouTube may lose access to their content and accounts due to copyright infringement claims or reports of inappropriate content, which may or may not be supportable. The risk of loss is higher if the content is not stored anywhere else. Though some mitigation methods are available through the platform, this issue may only affect a small number of accounts. Some of the content/iterations of these are likely preserved to an extent within existing web archives but not as targeted collection efforts. As we've seen with MySpace and other platforms where the platform producers decide to remove content or shut down with little notice, loss may be sudden. Replaying social media content from 2014 through modern interfaces poses a challenge to authenticity and reliability. If no recording or documentation for legacy interfaces has been preserved, it will not be possible to recreate older interfaces. You'd think the platform owners may preserve older versions, but these copies (if they exist) are not available to research or collecting institutions, and it would be worth engaging in a conversation about making them available for scholarly projects and cultural heritage. Some of this information is almost certainly lost already (some through deliberate disposal). The imminence of action depends on the type of institution. See also:
|
||
No Change