We consider that the establishment of an archive constitutes a commitment, both to the authors who deposit their writings in the archive, and to the scientific community which should be able to access it reliably. By its very nature, this commitment can not be made by a single researcher alone; it has to be made by an authoritative level within the host university.
At present (January, 1996), this process of formal commitment has not been concluded, and the present page describes the commitments that we hope to be able to make, rather than an actual commitment today.
We distinguish the following aspects of the commitment.
Persistence: once archived, a document shall continue to be available. This commitment is important in order that archived documents shall be citable in scientific publications. It does not make particular sense to cite a document that the reader may not be able to retrieve.
There is a main commitment and a few exceptions with respect to persistence. The main commitment is that every document which is placed in this electronic archive will stay there, and will continue to be electronically available at no charge except possibly transmission charges. The main exception is that if the copyright of an archived document is transferred to a commercial publisher (for example a scientific journal), then the full text of the document may be withdrawn and replaced by a reference to that new publication, if so requested by the author or jointly by the authors.
Note, in particular, that the document can not be withdrawn by the author under other circumstances than those just mentioned, since the continued access to an archived document is considered to be an important interest of the scientific community.
There is an additional exception if the author has awarded competing copyrights in some other ways, so that keeping the document on-line would cause legal problems. We do not want to get entangled in that. However, in any case, a paper copy of the document is placed in Linköping University Library at the time the document is first entered into the archive, and this copy will always be available by normal library practices.
Integrity: no one, not even the author(s), shall be able to change an archived document. This commitment is also important, together with persistence, in order that archived documents shall be citable in scientific publications. It is also important for the author if he or she is ever going to quote the existence of an archived document as evidence that he/she was the first authors of a particular research result.
Note that in order for this commitment to be effective and credible, it is not sufficient that authors or others are actually unable to change the contents of the documents; it must also be convincingly evident for the rest of the world that this is so. We foresee a combination of administrative and technical measures (possibly including public-key encryption) in order to guarantee integrity in this sense. For the time being, during the present experimental period, we do not have these mechanisms in place. The following temporary mechanism is used: all archived documents will also be formally registered as incoming documents by Linköping University, which means that an archive paper copy is date-stamped and retained, and that anyone at any time is entitled to read and inspect that paper copy.
In addition, in the particular case of entries in the electronic abstracts bulletin, we commit that when these have been entered in the definite, safe form of the archive, they will be manually double-checked against the officially archived paper copy. (For technnical reports and technical notes it will not be realistic to make such a double-check).
Openness: there must be fair and publicly stated rules for the possibility of entering information into the archives. This requirement goes without saying. At the same time, it is clear that there must be some restrictions as to openness, both for reasons of cost, and in the readers' interest of retaining the topicality of the archive.
For the ECSTER archive, the following commitments are made. The Technical Report Repository will include reports which are submitted by member research groups, until further notice, and provided that these reports have reasonable size, and that they can reasonably be argued to belong to the chosen subject area. Initially the subject area will be restricted to Reasoning about actions and change.
If it should ever happen that an author whose work has been excluded from the archive because it was considered as outside the current topic, that this author is dissatisfied with the decision and wishes to challenge it, then he or she is at least entitled to an electronic link to a URL of his/her choice where the case can be stated. (This applies for each researcher who is employed by a member research group, and where both the researcher and the group leader requests it). Of course, we expect this to happen very rarely if at all, but think that it is important to state the ground rules in advance, even for quite exceptional situations.
Notice, in particular, that the ECSTER archive is in its entirety an unrefereed structure. Entry in the ECSTER archive is therefore analogous to "publishing" an article as a departmental technical report, and we presume that it does not preclude later publication in a refereed conference proceedings or journal.
paper.ps The full paper in postscript format paper.html The full paper in HTML format abstract.ps The abstract in postscript format, English language abstract.html The abstract in html format abstract.txt The abstract in raw ASCII format, English language abstract-f.ps The abstract in postscript format, French language abstract-g.ps The abstract in postscript format, German language bibentry.bib A Bibtex entry for the present paperand similarly by extrapolation (abstract-g.html, etc). Of course, only a subset of these need to be submitted; there is no need to submit the paper in both postscript and html, for example. Files in raw ASCII format should use the full 8-bit character set even for "national" characters.
Note that the file containing the full paper should actually be called paper.ps or paper.html. The identity of this paper compared to others is presumed to be made by the directory name. The gzip compression scheme may be used, in which case the file names are postfixed by .gz in the usual way.
When the contribution has been set up in this way, please send an ordinary E-mail message to ECSTER, and we will arrange to fetch it and to include it in the ECSTER archive if eligible.
If a member research group has its publications organized electronically by a particular scheme which is different from the one just described, we may be able to set up our fetching procedure so that it understands your conventions. Please send us a note.