The paper presents work performed in a collaboration between the department of Information Studies and the Humanities Research Institute at the University of Sheffield. In it we investigate, problems caused by the use of platform specific software by humanities computing projects. The methodology for the project was that of case studies focusing on projects in the history of Natural history and science: The Hartlib Project (University of Sheffield) The Darwin Correspondence Project, (University of Cambridge) The Robert Boyle project (Birkbeck College, London), The Mueller Correspondence Project and The Newton Project (Imperial College. London). All projects had encountered problems with the use of platform specific formats, and had decided to use SGML as a way of providing a platform independent format that would also help to preserve the material independent of hardware platforms. This was influenced either by advice from other scholars, contact with humanities computing centres, or with national advisory bodies. In particular, projects considered adopting the DTD developed by the Text Encoding Initiative. Despite this enthusiasm a number of problems were encountered with the type of material to be encoded. All but one project therefore chose not to adopt it in its original form. The paper argues that problems were caused by researchers working in isolation from each other. This can cause the replication of problems that others have already worked on, and expertise not being shared with different projects. Documentation was often not kept up to date because those working on projects have no time to do this. The paper contends that there is a role for outside reportage in the collection and presentation of documentation and its dissemination throughout the humanities computing community.