The Ephemeral Web and the Case for Proactive Archiving

Technical Summary: The Ephemeral Web and the Case for Proactive Archiving

Problem Statement
The paper addresses the structural ephemerality of the web, challenging the assumption that digital content is a durable record. Despite the perception of permanence, web content is fragile due to domain changes, redesigns, institutional relocations, and political instability. The author highlights that "link decay" is a systemic issue, citing Pew Research Center data indicating that 38% of webpages from 2013 were inaccessible by October 2023. The specific catalyst for this work was the rapid transformation of the Pakistan Embassy International School and College Tehran (PEISCT), which underwent a domain change, leadership turnover, and a physical relocation within two years of the author's graduation. These changes resulted in the loss of the institution's visual history and earlier content, illustrating how website redesigns often function as "quiet historical erasure" where continuity of access masks discontinuity of memory.

Methodology
To address the loss of the school's digital history, the author designed and deployed a lightweight, automated archival system. The methodology involved the following technical components:

• Architecture: The system was implemented in Python and deployed via GitHub Actions within a public repository. It operates as a symbiotic system, using GitHub for automation infrastructure and the Internet Archive's Wayback Machine for preservation storage.
• Link Discovery Logic: The archiver utilizes a recursive breadth-first search starting from a seed URL. It extracts quoted internal links from page sources.
  • Filtering: The system filters out CSS and JavaScript resources to avoid unnecessary archival bloat.
  • Inclusion: It specifically includes media and resource paths (e.g., files within wp-content) as these often contain historically valuable photographs and PDFs.
• Execution Strategy:
  • Runtime Constraints: Recognizing that GitHub Actions imposes a six-hour maximum per run, the system is capped at approximately 3 hours and 55 minutes.
  • Randomization: To prevent the system from repeatedly archiving the same early sections of the site while neglecting later pages (a common issue with fixed-order processing), the discovered links are shuffled before each submission run. This ensures that over multiple executions, different parts of the site receive roughly equal archival attention.
  • Submission: The system submits URLs to the Wayback Machine's save endpoint.
• Deployment: The workflow was scheduled to run six times daily.

Key Results

• Operational Scale: Between mid-January and mid-April 2026, the system executed approximately 580 times, consuming nearly 2,500 aggregate hours of execution time.
• Feasibility: The project demonstrated that proactive archiving is achievable using widely available, low-cost infrastructure without the need for custom servers, paid cloud orchestration, or institutional backing.
• Vulnerability Discovery: A critical operational lesson was the discovery that the preservation infrastructure itself is ephemeral. GitHub automatically disables scheduled workflows in public repositories after 60 days of inactivity. The author discovered the system had been disabled 12 days after the fact, revealing that archival systems require their own continuity monitoring (e.g., heartbeat checks) to survive.

Key Contributions

1. Proof of Concept for Automated Archiving: The paper provides a concrete demonstration that a single individual can maintain a persistent preservation process using open-source tools and public infrastructure.
2. Identification of "Contingent Liveness": The author argues that digital fragility extends beyond dead links to include the weakening of maintenance conditions. A site may remain visible to users but become unmaintainable due to internet shutdowns, filtering, or dependence on foreign services (e.g., WordPress infrastructure in a restricted environment).
3. The Case for Proactive Archiving: The paper argues that archiving should not be a reactive, niche activity performed by specialists after content is lost. Instead, it should be a proactive, integrated part of website maintenance. The author suggests that preservation should be tied to meaningful change (e.g., redesigns, domain shifts) rather than relying on repeated, redundant archival runs by outsiders.
4. Intergenerational Responsibility: The work frames digital preservation as a mechanism for communities to maintain continuity with their past, allowing future generations to understand institutional history that would otherwise be flattened by current transitions.

Significance and Claims
The paper claims that the ephemerality of the web is a structural condition, not an exception. Consequently, the current model of "saving after loss" is inefficient and unsustainable. The significance of this work lies in its argument that if digital societies wish to preserve institutional memory and public history, archiving must evolve from an ad-hoc workaround into a standard component of website maintenance.

The author acknowledges limitations, noting that the system is a lightweight prototype focused on a single domain and does not verify the rendering quality of archived pages. Furthermore, the paper does not resolve legal or ethical questions regarding archival reuse. However, the core contribution remains the demonstration that proactive, automated preservation is technically feasible and necessary to counter the structural fragility of the digital record. As the author concludes, without such measures, future generations lose not just files, but the ability to understand their own history and the institutions that shaped them.