No abstract

We describe the aims and aspirations for the Digital Curation Centre (DCC), the UK response to the realisation that digital information is both essential and fragile. We recognise the equivalence of preservation as "interoperability with the future", asserting that digital curation is concerned with 'communication across time'. We see the DCC as having relevance for present day data curation and for continuing data access for generations to come. We describe the structure and plans of the DCC, designed to support these aspirations and based on a view of world class research being developed into curation services, all of which are underpinned by outreach to the broadest community.

As the scholarly communication system evolves to become natively web based, citations now commonly include hyperlinks to content that is issued on the web. The content at the end of those hyperlinks is subject to what has been termed 'reference rot': a link may break or the content at the end of the link may no longer represent what was first noted as significant. Reference rot threatens both the usability of what is published and the long-term integrity of the scholarly record. The aim of the Hiberlink project has been to focus on this problem and then to compile and analyse a large corpus of full-text publications in order to quantify the extent of reference rot. The results are now out, and the task has shifted to alerting publishers and libraries on what to do in order to ensure that published web-based references do not rot over time. This has implications for the integrity of the scholarly record and for authors of that record. Fortunately, the Hiberlink project has progressed further than originally envisaged and has recommended remedies aimed at alleviating reference rot. Reference rot in scholarly statement: threat and remedyBased on a breakout session presented at

No abstract

This paper describes (i) how we rewrote a piece of technical text to make it easier to understand, and (ii) how we attempted to measure our success at this task.The purpose of the paper is to chart the difficulties in which we found ourselves and to make the task easier for others who may wish to follow our path.There are numerous guidelines on how to produce readable writing, but, unfortunately, they are not consistent. Klare (unpublished), for example, reported wide areas of disagreement in the 156 suggestions that he found in 15 books (five of which were written for writers in general and ten for technical writers). (See Table 1.) Furthermore, most of the available guidelines do not take into account typographic design considerations. Exceptions here, of course, are the books by Hartley (1978,1980). The general procedures currently advocated by Hartley for redesigning text can be listed under three interrelated headings: Textual -Use the active voice. -Use simpler wording. -Either shorten sentences, or expand them into two or three simpler sentences. -Divide long paragraphs into short ones. -Number and list actions and procedures (and put them in temporal sequence). -When in difficulty think of how you would explain to a friend what you are trying to write. Write this down. Polish it. Typographical -Decide on the printed line length and typesize. Type the text with a matching number of characters per line. -Use units of line feed in proportion to separate out and to group units of text. (For example, separate headings from the text by using half a line space below the heading and one line space above it; separate paragraphs from each other by half a line space; start new sentences on a new line.) -Set the text unjustified (Le. with equal word spacing and ragged right-hand margin -as in normal typescript). -End each line at a sensible place syntactically (e.g. at the ends of clauses). Avoid word breaks (hyphenations) at line ends. -End each page at a sensible place (e.g. do not have the first line of a new paragraph as the last line of the page). -When printing use bold lower-case type (not capitals) for main headings. Procedural -Leave each revised draft for at least 12 hours. -Revise and simplify revised drafts. -Do not look back at the original text (except afterwards to check on ambiguities or points of meaning). -Ask colleagues to help simplify the revised draft (either by simplifying it themselves, or by pointing out where they might expect difficulties to occur). -Repeat as often as time allows.The procedural aspects of these guidelines arc only suggestions but they are particularly important. Problem words and phrases in one version seem to appear in two or three revisions before they finally disappear. There never seems to be a final solution.To see how these guidelines work in practice readers are now invited to compare Figures 1 and 2. These figures present the first page in its original and revised form from a document containing four pages of text. As noted above, the text in Figure 2 can still be ...