[Forwarding from the AmSci OA Forum. --Peter Suber] Posted on behalf of Subbiah Arunachalam, Leslie Chan and Barbara Kirsop Regarding comments to the list on the costs of setting up and maintaining an Institutional Repository (IR), we agree that figures are being publicised about the costs without distinguishing whether these are solely for the institute's published research output or for a high-cost Institute-wide or Campus-wide resource for all purposes. So when issuing cost statements it is critical to explain precisely what is being set up, otherwise figures are meaningless. Because of these confusions, we are concerned that the very high figures quoted can deter smaller institutions and developing countries from setting up open access archives. To off-set such concerns, we know that over 70 repositories for research publications have already been set up in the developing world and we have evidence of the low costs associated with these. Only yesterday a colleague from India responded to a question we asked, saying '....organisations can do with a server just costing about 40 K (rupees) and use a broad band connectivity. This connectivity has become very cheap recently due the launch of Broadband Services by BSNL/MTNL/Private Cos. Running the server is also not a problem for us. We are already paid for other work. It just takes fraction of our time to maintain the server and handle the approval of uploaded articles. In the same manner, Indian Universities / Research Institutes already have infrastructure in form of libraries and librarians....' And we have had other similar reports from developing country organisations. Cost should not be used as an argument to dissuade organisations from setting up an eprints server. As Arthur says, costs are trivial compared with the overall institutional budgets and can generally be absorbed. If further examples of costs are to be reported, the purpose of the repository and the overall research budget of the institution should be stated at the same time. Leslie Chan, Subbiah Arunachalam, Barbara Kirsop Electronic Publishing Trust for Development [Forwarding from the AmSci OA Forum. --Peter Suber] Posted on behalf of Subbiah Arunachalam, Leslie Chan and Barbara Kirsop Regarding comments to the list on the costs of setting up and maintaining an Institutional Repository (IR), we agree that figures are being publicised about the costs without distinguishing whether these are solely for the institute's published research output or for a high-cost Institute-wide or Campus-wide resource for all purposes. So when issuing cost statements it is critical to explain precisely what is being set up, otherwise figures are meaningless. Because of these confusions, we are concerned that the very high figures quoted can deter smaller institutions and developing countries from setting up open access archives. To off-set such concerns, we know that over 70 repositories for research publications have already been set up in the developing world and we have evidence of the low costs associated with these. Only yesterday a colleague from India responded to a question we asked, saying '....organisations can do with a server just costing about 40 K (rupees) and use a broad band connectivity. This connectivity has become very cheap recently due the launch of Broadband Services by BSNL/MTNL/Private Cos. Running the server is also not a problem for us. We are already paid for other work. It just takes fraction of our time to maintain the server and handle the approval of uploaded articles. In the same manner, Indian Universities / Research Institutes already have infrastructure in form of libraries and librarians....' And we have had other similar reports from developing country organisations. Cost should not be used as an argument to dissuade organisations from setting up an eprints server. As Arthur says, costs are trivial compared with the overall institutional budgets and can generally be absorbed. If further examples of costs are to be reported, the purpose of the repository and the overall research budget of the institution should be stated at the same time. Leslie Chan, Subbiah Arunachalam, Barbara Kirsop Electronic Publishing Trust for Development