Strengthen Research Associations Through Exit-Strategies for IT Services

Research relies heavily on, and is advanced by, the IT infrastructure and services provided. If critical infrastructure and services become unavailable unexpectedly, these need to be replaced given the time frame of a possible discontinuation, otherwise unavailability might have a strong negative impact on the overall research service in the research association. By identifying such infrastructure and services a research association can mitigate the risk by evaluating alternative offerings to the research communities early enough. For critical infrastructure and services there need to be replacements in order to enable research associations to overcome obstacles posed, for example, by companies that decide to exploit vendor lock-ins of their user communities. Exit-strategies for IT services make research more robust since other services can replace critical services if the necessity arises. In times of geopolitical uncertainties and crises this becomes an even more important asset and valuable quality. The platform Helmholtz Federated IT Services (HIFIS) offers IT infrastructure and services as well as Research Software Engineering (RSE) trainings and consulting services to the Helmholtz Association. Consequently, we have a strong interest in operating services that are less prone to such uncertainties. In many Helmholtz centres GitLab is such a critical service. Thus, alternatives need to be evaluated and set up. In HIFIS we decided to develop a concrete exit-plan to migrate to Forgejo, which is completely free and open-source, as an alternative to GitLab that follows the open-core model with additional Enterprise features instead. By exploiting existing vendor lock-ins, GitLab Inc. might make respective business decisions which hinder us from offering GitLab in the free community edition to the Helmholtz Association in the future. An exit-strategy expands the landscape of offered code hosting platforms by a substitution that can be enabled on demand. In this case-study we would like to show-case that considerations about exit-strategies lead to more robust offers to the research communities. By doing so, services should be selected with a focus on open standards and APIs from the very beginning to allow a migration to other services. In our evaluation of alternatives we came to the conclusion that Forgejo is a good choice because it is completely community-driven, based on open standards and licensed under a copyleft license. We will share our experiences about this case-study and invite feedback from interested communities on use cases and other aspects.