[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Reply to: [list | sender only]
Re: Towards a new version of the World Database of Crystallographers
- To: Multiple recipients of list <epc-l@iucr.org>
- Subject: Re: Towards a new version of the World Database of Crystallographers
- From: Yves Epelboin <Yves.Epelboin@lmcp.jussieu.fr>
- Date: Wed, 4 Feb 1998 09:31:14 GMT
Before going into details I believe there are a number of basic questions which must be answered. - 1 - Contents of the database From Brian's message I understand that he wants to create a unique database both for the use of WDC and for the journals. Let us examine this point first. Facts: ------ + some people publishing in the journals are not listed in WDC and should be. + Some people listed in WDC do not publish in the journals. + The address of a scientist as listed in WDC may differ from his/her address in a paper These points can be easily solved if the data base contains: + postal AND full address for the person as listed in WDC + postal address for publication PLUS some other reference. For instance assume that Howard moves for six months in my lab and that we publish a paper together. His institution address does not change but on the paper he will give a reference to my lab. This reference is not only made from the postal address but contains some additional information such as: URA009 CNRS, Universités P.M. Curie and D. Diderot because my institution is under the responsibility of these three institutions. This lengthy reference does not appear in WDC. ===> It means that the database should contain + a permanent institution reference + a postal address as now listed in WDC + an institution reference + a postal address FOR each paper or that this second part may change from one paper to the next one. Keywords: -------- Each individual is listed with a list of keywords to describe his/her interests. The journals require that the authors give a list of keywords for each paper. It would be a good idea to merge both BUT: + Assume a newcomer in the database who is registered for the first time because he is a co-author of a paper. Should his/her interests be reduced to the list of keywords corresponding to the publication? + In other words the list of keywords defining a publication is not the list of keywords defining the interests of a person. In most cases the intersection is not zero unless for occasional collaborations on subjects of minor interest which will not be listed in WDC. Entries: -------- Who should be listed in the database? The straight answer is all individuals either listed in WDC or/and publishing papers. There are a few questions: + How about occasional collaborations? Should we add in WDC a scientist who does not really belong to the community, is not too much interested in it but has been a side collaborator for a paper? + How about students who publish papers and then move outside the university. They would be added in WDC just in time when they are leaving the community and WDC would contain a number of false references. - 2 - Updates and new entries I am now speaking of updates and new entries coming from a scientist, not from the fact that somebody is publishing a paper. In other words I am making reference to the problem of updating WDC. In fact I am repeating some of my questions from a previous mail. - 2 - Validity of entries For the moment national editors are responsible for qualifying a person to be listed in WDC. I know that this is not perfect and they are some curious references however most of the entries seem to be valid. If we skip the national level who will be able to decide that a request has a meaning or not. Who, in Chester, may have the idea that: Yves Eepelboin, laboratoire de Patatologie biologique, 5 allée du Parc, 94200 Ivry is a nonsense ( the mail address is good: it is my private address!). Any french scientist knows that there is no crystallography lab in Ivry. At a time where the Web is open to everybody I foresee a number of difficulties and I am afraid that very rapidly WDC will contain a number of unqualified entries. Altogether most national editors have done a good job and regional editors too. They could replace national editors when they are not acting and, if not the best, are still reliable filters. At the meeting I had with them in Seattle, it appeared very clearly that most of them are willing to continue their job. I am very reluctant in concentrating the updates in Chester, not only because of the workload as Peter is afraid, but also because Chester does not seem to be the most qualified place to validate the entries. I refuse to move from that view before consulting the National Committees. If they agree that this responsibility may be moved to Chester, who will be sole responsible to decide who is listed in a given country I will agree. We cannot change the process without this agreement. - 3 - Technical aspects User interface -------------- The experience with the Web interface for US and France was rather good. However there have been complaints that the update process was dull: why should an already registered scientist fill again all the fields. Provision must be made to show the actual contents of an entry for modification. Managing the database --------------------- The database is becoming a real relational database and should be manipulated with professional tools. I do not know anything of the YARD database. All I am aware is that today developments are not made anymore using cgi-bin: - lake of security - too slow Today developments are made using a "three-third" system (I am not sure of the words in English): + Web interfaces: not the same for consulting and updating and depending on the level of responsibility of the user + a database + an intermediate engine directly coupled from the Web interface to the database. The engine takes care of security problems as well. Mirrors ------- For efficiency purpose I agree that the mirrors should be used to consult the database. However the database system (YARD + engine + interfaces) can not be distributed easily: + variety of platforms + software price + difficulties to install thus WDC has to be an export of the Chester database in a standard which can be installed on each mirror. However we must be aware that this may need a rather important work and that the mirrors managers are volunteers. We may have some difficulties. Should we not consult them before adopting a standard? Also for efficiency reasons requests for updating should be run from the mirrors. This adds a difficulty. There is much more to say but I will stop today on these basic points. I am not sure we can completly solve this problem by e-mail only. With best regards, Yves
Reply to: [list | sender only]
- Prev by Date: Re: Towards a new version of the World Database of Crystallographers
- Next by Date: Re: Towards a new version of the World Database of Crystallographers
- Prev by thread: Re: Towards a new version of the World Database of Crystallographers
- Next by thread: Re: Towards a new version of the World Database of Crystallographers
- Index(es):