The Electronic Astrophysical Journal Letters offers both WWW client and e-mail gateway access. While the WWW is an important and widely used mechanism for electronic distribution of research results, it is not the only medium used by the astronomical community. The AAS designed the electronic journal around a core of SGML-tagged documents to be able to quickly and easily take advantage of as many distribution mechanisms as possible. SGML is fairly easily convertible into a variety of formats, including HTML. Moreover, as an ISO standard in publishing storage formats, SGML allows for the translation of files into formats and distribution mechanisms not yet developed.
While the HTML-based pages are currently the most popular point of entry for users of the journal, it is important to note that they are not the only point of entry into this electronic journal. The AAS has also developed an e-mail gateway for users who do not have access to Mosaic or other HTML browsers. To conduct a search of the Journal using plain e-mail, send the keywords you wish to search on to: firstname.lastname@example.org. You will receive a listing of ``hits" back within a few moments. You can then submit a second ``query" listing the articles you wish to retrieve, and the full text of the article will be returned to you via e-mail.
A stand-alone client, which is based on the ANSI Z39.50 protocol and is accessible through astronomical software packages such as IRAF, is also available.
A user can select between browsing the journal, starting with the Table of Contents, or choosing to conduct a search. Two separate query pages are available, allowing the choice of ranked or boolean searches. The underlying search engine was developed by the Clearinghouse for information Discovery and Retrieval (CNIDR), for use with the U.S. Patent Office's database. It is based on a Z39.50 Version 2 client/server package, and uses an indexing and retrieval engine CNIDR has written. The Z39.50 code, and a simpler version of the index/retrieval code is in the current Isite distribution available from ftp.cnidr.org.
Matching documents, or ``hits" will be returned in order of weighted ranking based on their relevance to the query. To further refine your searches, you can also de-emphasize words that are of less importance. Articles that contain these words will not be eliminated from hits; they are simply given less weight and appear further down on your list of returns.
Searches can be conducted on full text of the article or on any one of the following options: abstract, authors, reference, titles or keywords. These options are available for ranked and boolean searches, and for de-emphasizing items in the text as well, refining searches substantially.
You can also control the format of the items that you retrieve. If you are interested in receiving only the titles of the articles that match your query, select ``TI" as your retrieval type. If you want to see where your words fall within the article, select ``KWIC" - keywords in context. The third option, ``B," for brief record, allows you to bring back the specific section of an article where your keywords are located.
The full text of the items retrieved in your search can be used to conduct another full search of the database to bring up papers similar to the one you have chosen to select, further refining a literature search.
The user can choose to view an article fully formatted in HTML (equations, special characters and greek letters will appear formatted in the text). Or the TeX version of the document or its subsequent dvi file can be viewed. Users also have the option to view full Postscript (and in the near future, PDF) versions of the articles. (Note: these are quite large in size and may take some time to load). Users can also print locally any one of these formats.
Users can navigate not only between articles in the journal database, but also to outside references. The journal is linked to the NASA Astrophysical Data System (ADS) abstracts, by clicking on the icon in front of the reference item. Documents which are not located in one of these two databases can still be accessed.
Uniform Resource Locators (URL's) are included in the journal for documents that reside in distributed databases. However, URL's tend to be fairly fragile - they can break, and they can move. To ensure as a robust a system of remote document location as possible, the AAS is constructing a a database of URN's (Uniform Resource Names) that resolve URL's into the current location of the document. Rather than relying on only one of an objects attributes as its identifier (as a URL does with an object's location), a URN contains directions to a server to execute or resolve a search for an opaque string within the URN. For example, a URN could use a Bibcode as its identifying opaque string. If a document changes locations, the new location needs only to be noted once - in the URN database - for all references to it to remain active.
Links to images, figures and graphs are available, as in-line thumbnail of the full images, which can be expanded into full displays. Images are scanned and stored in the database with the text of the articles. Plans are underway to construct links to the underlying data used to create graphs and tables from the relevant data storage sites.