Technical
2012
Move servers to virtual machine (VM) infrastructure - The three arXiv server machines are nearing end-of-life and Cornell central IT is aggressively moving to virtual machine (VM) infrastructure. We are thus working to transition the from the current servers to virtual servers managed by IT@Cornell by the end of 2013. As part of this transition we will formalize the update and maintenance process for these machines, and remove access by non-CUL or IT@Cornell staff.
Admin and moderation system improvements - The arXiv admins and moderators face an ever increasing workload and have had to work-around a number of bugs and awkwardness in the arXiv interfaces. We are focusing development effort July-September 2012 on working through the backlog of issues to improve admin and moderator efficiency. Tasks include: better layout of admin queues; improved linking between admin UI screens; fix locks on user changes at different stages of submission; improving messages/help for users to reduce number of questions they ask; more consistent handling of non-ASCII characters. After September 2012 work on such fixes will continue at a slower pace to deal with pressing issues.
2013
Invenio for display and access - It was decided in 2010 to move the display and access functionality of arXiv to the Invenio platform developed at CERN. The primary goal of this move it to reduce the maintenance burden of in-house code and share the effort of developing new features with other organizations using this open-source software. The move to Invenio will facilitate improved collaboration with our partners at NASA ADS and INSPIRE. Work to data (Aug 2012) on this move has progressed slowly because of unforeseen difficulties and staffing shortage. We are currently re-planning this work with the intention of redirecting a significant fraction of the development team's effort to Invenio in Fall 2012. The goal is to deploy this system in 2013 and at that time enable horizontal scaling of the arXiv web user-interface to cope with our ever-increasing traffic.
Security and login, email privacy - There are significant flaws in arXiv's security and we are lucky that we have not been targeted. Issues include: all password entry and authenticated interactions should occur via https; domain based cookies should not be sent to mirror sites; and user email addresses should be more carefully protected.
2014 (If there are any 2014 goals we know of, let's put them in this roadmap as 'planning work')
Alerting system - The email alerting system remains extremely popular but the mechanisms for subscription management are extremely outdated. Users should be able to see and control their subscriptions from their user account page.
User Support and Moderation
Strengthen physics moderation - We need a sustainability effort aimed at strengthening physics moderation. This will involve finding additional moderators and seeking new leadership within the arXiv physics community. Over the years, physics moderation has grown substantially dependent on Paul Ginsparg's involvement. With Paul's expressed desire to cease this routine involvement, we need to attract new people, explore new strategies, and build up the supporting infrastructure for physics moderation. This effort can in turn inform other subject areas with regard to effective and sustainable moderation practices.
New, better, and improved moderation tools - We need better and newer tools to support arXiv moderation processes. These processes have been supported in much the same way for 10-15 years. We can gain efficiencies and reduce effort and costs by applying newer technologies.
Review arXiv policy documentation and user support guides - arXiv has a substantial amount of policy documentation, which has grown organically over the years. Its organization and presentation needs review and updating. User help pages need an even more thorough review. Many of the user help pages are out-of-date or entirely obsolete. These materials need substantial reorganization and clean up.
Governance
2012
Create SAB bylaws - David Ruddy will discuss the draft bylaws with the SAB during their September meeting to seek input. SAB will also assess the idea of appointing a scientific leader — a part-time position to provide intellectual leadership from the perspective of the scientific community.
Create MAB bylaws - By early October'12, we will complete and announce the MAB bylaws, which describe the composition and operation of the group.
Form the first MAB - We identified a small election committee (sub-group of the interim MAB) to assist us in organizing and running an election to elect members for the first MAB. We will invite nominations from the member organizations,SAB, and the Simons Foundation.
Form the first MAB - Make MAB appointments in January'13
Finalize SAB bylaws - The current SAB members has started to review the draft bylaws during the last quarter of 2012 and aim to finalize them in early 2013.One of the outstanding tasks is assessing the idea of appointing a scientific leader — a part-time position to provide intellectual leadership from the perspective of the scientific community.
Test and refine the operation of the new governance model -The arXiv principles aims to clarify the authority, responsibilities, and constraints of CUL, MAB, and SAB. Ironing out problems and developing a working system may require some time to test and observe the inner operation of the governance model. For instance, there may be overlaps and tension, due to role ambiguity or personality conflicts. Currently there is not a clear mechanism to settle potential conflicts among the three chambers - CUL, MAB, SAB. CUL will engage SAB and MAB in proactively identifying and addressing problems. The first strategy to reduce/avoid tension is having in place a roadmap for arXiv so that everyone knows the CUL team's priorities and goals. Both boards need to understand the vision, priorities, and challenges to be able to contribute to the arXiv's governance.
Create a working relationship between SAB and MAB - The advisory groups need to regularly exchange information in order to contribute to each other's agendas in a meaningful and useful manner. For instance, CUL will provide joint briefings(reports) to SAB and MAB to highlight common interest areas and complementary perspectives. Having ex officio member representation in each group will also faciliate information sharing and developing a common understanding of the respective goals.
Further refine the benefits of being a member for participating institutions - During the governance planning meetings, several ideas emerged as potential free services for members; however, implementing these features may require significant staff time. Therefore these services need to be considered in the context of arXiv's current maintenance and development priorities. After the MAB is formed, in order to better understand demand for services, CUL will invite ideas/proposals and review them with MAB and SAB. Some of the ideas to be considered include: submission-based data when arXiv's metadata structure is ready; institutional repository bulk download. Member institutions are also very keen on informing their scientists and researchers about arXiv's business model (e.g.,arXiv LibGuide to share with their faculty and students in related disciplines). Often, we are advised to make it easier to find business info on the arXiv webpage (one idea proposed is to put links from abstracts or submission forms to 'sustainability' page).
Continue with the membership drive - We are very encouraged with the 5-year pledges received so far. We want to increase the number of arXiv member institutions to create a large and international network of supporters. Another goal is to be able to reduce the institutional membership fees in the future (current annual institutional fees are in the $1,500-$3,000 range).
Develop reserve fund policies - The purpose of the arXiv reserve fund is to support unexpected expenses to ensure a sound business model. Currently, arXiv has a reserve fund of approximately $125,000 accumulated during the last two years due to unexpected staff vacancies and other savings, such as transition to virtual servers. The 2013-2017 budget projections includes a contingency line ($50,000-$100,000 per year); however, the Cumulative Operating Contingency balance does not factor in the potential expenditures that will be charged against this account in support of unforeseen expenses. We will discuss the desirability of establishing a minimum and maximum contingency fund levels. The financial projections assume that the membership fees will increase at an average rate of 2 % per year. Ideally, membership fees should decrease as the number of participating institutions increases. This is one of the reasons we would like to identify an ideal contingency amount so that we will know when to stop accumulating and start reducing membership fees. We need to develop policies about how contingency funds will be used and how the account will be structured (e.g., funds needed for closing the business versus development funds - also a part of the reserve funds to create an endowment & the interest can be use the interest for reducing membership fees).
Business Planning
Define a R&D agenda and seek external funds to advance arXiv - So far, CUL's sustainability planning efforts focused on arXiv's operational budget to support the core services and arXiv's strengths in order to stay mission-centric. One of the key goals ahead of us is to define a research agenda for arXiv. We'll seek input from the advisory boards, users/scientists at large, and information scientists. We will develop a methodology for making decisions about arXiv's R&D projects and partners. One of the challenges will be maintaining/developing arXiv as a distinct system vs. envisioning it as a part of a joint scholarly communication infrastructure (interoperability, research data, tracking funding sources, etc.). This is likely to be a potential tension area between the library community and scientists.
Continue the dialogue with publishers/societies - In celebration of the arXiv's 20th anniversary, on September 23, 2011 Cornell University Library (CUL) hosted a meeting at Cornell with the representatives from several publishers and societies that are interested in Cornell's sustainability planning efforts. The meeting report provides a synopsis of the discussion and recommends next steps for continuing this dialogue. We will resume this investigation and discuss the feasibility and desirability of establishing a research and innovation collaboration in support of arXiv. This effort is envisioned to entail a separate funding stream (created by participating publishers and societies) from the operational budget, which includes resources for the routine and core services currently provided by the arXiv team (including essential updates).
Define and communicate measures of success for arXiv- Scientific repositories' ecology continues to evolve in response to the changes in scholarly communication patterns, information policies, repository technologies, and global financial trends. During the last three years, arXiv has been going through a streamlining process and some goals continue to be works in progress (for instance, developing of an Invenio-based discovery and access interface, streamlining the moderation process, strengthening staffing through back-up arrangements, etc.). To address this dynamic process, one goal over the next several months is to create an assessment model to help CUL continue to fine-tune the sustainability model. An assessment plan will also help identify unforeseen developments and making course adjustments to the service and collaboration model. Working with the SAB, one of the goals of the transitional MAB will be to develop an assessment framework with three key components:
- Desired outcomes and success measures to measure progress (for instance, dynamics of the governance model, level of financial support, enhancements to arXiv, improvements to moderation system, etc.).
- Assessment model to gauge and report success based on the identified outcome measures.
- Plan a five-year review process to enable CUL to conduct a comprehensive self-evaluation with input from the Simons Foundation, MAB, SAB, and other key partners.
Communication
Put in place a systematic and regular communication strategy for MAB and SAB - Communication and follow-through are essential - don't promise anything to SAB or MAB members if you'll not be able to fulfill. Make sure to follow up - whether it is relaying progress or delays (or rejecting a recommendation). When you hear ideas and get suggestions from SAB and MAB and don't follow their advice, you need to at least explain why. List priorities and communicate them prior to meetings. Implicitly seek help in setting priorities. Be clear with your expectations (seeking input vs. help CUL make a decision by selecting one of the alternative strategies). Explain that CUL has limited resources and set expectations accordingly. If there are requests that Cornell cannot meet, someone (member) may be able of pay. Seek partnerships and other revenue streams.
Communication with users (scientists) - Share opinions received from scientists and users with SAB and MAB.How about seeking input from other scientists, arXiv users? What is working well for them that needs to be maintained? What are their unmet needs? What kinds of changes they want to see implemented? CUL needs to have a systematic way of gathering user feedback and this info ought to be shared with MAB and SAB (both groups are also important channels of input as they represent other groups). Will it be useful to encourage users to send feedback directly to SAB and MAB and make them aware of these advisory groups (through a mailing list)?
Economics of open access awareness building - Scientists know very little about how arXiv is run and how much it costs. Improve communication, it is important for scientists to know that open access is not free and even open access systems need to be carefully managed. Some scientists may think that they don't need libraries any longer. Be vocal about libraries' new roles. Use members to spread info to their scientists - there needs to be outreach to scientific community
Also make it known that arXiv has a QC strategy - many librarians don't know about it.