[Mirrored from: http://www.aas.org/Epubs/webinfo/r961118.htm]
Latest developments in AAS Electronic Publishing
A Report to the AAS Publication Board
Peter B. Boyce, 11/18/96
Integrated Digital Electronic Astronomical Resource
The most important feature of the AAS electronic publishing effort has turned out to
be the close connection with the Astrophysics Data System database of abstracts, and the concurrent use of a standardized article identifier to facilitate the automation of links to references and between articles. Agreements either in principle or in fact now exist between the major astronomical publishers and database groups.
Astronomy is well ahead of other disciplines in fostering interoperability between journals, bibliographic archives and databases. We have, in fact, created a working prototype digital library.
To bring attention to the need for standardized naming and interoperability conventions, we are proposing to officially launch the integrated resource at the Toronto AAS meeting. We are proposing to name the system Urania, after the muse of astronomy. Various organizations have been asked to be among the informal group which contribute to Urania. A separate draft document which might serve as a press release is attached. A search of the WWW reveals a Urania Observatory in Germany and a Urania amateur astronomy club in Italy. There should be minimal conflict of names.
ApJ and AJ
- All set for on-line by January. Final debugging of the SGML is in progress. Sample articles will be available at the Pub Board meeting. PDF versions of all issues since November, 1996, are available.
- The ApJ Letters is now served from UC Press. Next year we will implement a plan for electronic publication of the HTML version ten working days after acceptance by the editor. Paper publication and the PDF version will follow about a month later, after typesetting and pagination. This fast process requires additional work and extra attention which probably will not be cost effective for a major journal.
- PDF available starting with Jan 1996. We will make that available on line by February. AIP will provide SGML by the end of March, 1997.
AIP has agreed to furnish files to UCP for distribution through the AAS distribution mechanism used for the ApJ. They have verbally agreed to produce SGML and HTML files with links to references by the end of March, 1997. As of 11/11/96 we are still waiting for prices from AIP.
We expect there will be a period of time-consuming adjustments as the details get straightened out.
As a side issue the AIP uses a different flavor of SGML (ISO 12083 math) from that used by UCP (AAP math). The different SGML should not impact the user, but attention will have to be paid to the details of implementing the HTML translation. Because the AIP SGML is produced after the Xyvision typesetting (which has tags, but not inherently enforced uniformity) there can be some variation in the Xyvision output which introduces problems when translating back into SGML. We see a few problems with the SGML for the Letters which comes from Cadmus (who also use Xyvision typesetting). Careful attention to uniformity of the input and checking of the output has made the ApJ Letters a high quality electronic publication. The same will have to be done for the AJ. There is nothing insurmountable, just time consuming.
Discussions with CDS(Europe) are nearly complete. We expect to test the site before Dec 1996 and implement the full mirror as soon as the letter of agreement is signed. CDS mirrors the ADS abstracts.
Referencing from multiple possible sites to the ADS abstract database at multiple sites will requires all links to be made through a name resolver. This has been implemented.
CSIRO is interested in hosting a mirror site to serve the far east. We have an agreement in principle. The Australian mirror can be operational as soon as they can buy the additional disk drives.
The editors of A&A, PASP, and Icarus are very interested in having archives which can operate together with the AAS journals. Blackwell (publisher for MNRAS) is also interested, but RAS has been cautious in their approach to electronic publishing. Elsevier's New Astronomy has appeared and has published a small number of articles. The production time has been longer than the ApJ for several of the articles, but this will certainly shorten as they develop their production system.
Nevertheless, it points up the difficulty of electronic production and emphasizes the success of the ApJ which has produced 300 Letters and 700 ApJ articles in the time New Astronomy has produced seven.
Major discussions with A&A (and Springer) have been held in late June.
The main journal plans to be on line with an SGML version as rapidly as they can solve the production problems, now scheduled for sometime in 1997. The A&A supplements are on line with a PDF version, and expect to have an SGML version ready shortly. When the A&A goes on line we will have > 85% of astronomical literature on line in some form or other. Our goal is to make the different electronic journals operate seamlessly together for the users. We don't have to do everything exactly the same way internally. The A&A editor (Jim Lequeux) has established the requirement that the A&A references link to the ADS and to the ApJ directly, just as the eApJL works. He is also interested in the UCP hosting a US mirror site for A&A and A&A Supplements. We believe this will be a fruitful collaboration.
We have continued to work with the library and publishing communities to make them aware of our electronic publishing program. The object is to raise the awareness within these communities of the important user features of the eApJL. Our goal is to get the library community to understand the superiority of the AAS approach and to encourage them to pressure the publishers to produce more effective electronic journals.
Another goal is to convince the publishers of journals relevant to astronomy to adopt a few simple standards for naming articles and for responding to queries automatically, so that the ApJ will work seamlessly with the other electronic journals. We have been remarkably successful in this latter activity.
Peter Boyce (with other Electronic Publishing Development Team members) has written articles for Physics Today, Computers in Physics, Vistas in Astronomy and Serials Review all peer reviewed. He has made presentations at several universities, the Special Libraries Association, a UNESCO/ICSU Press workshop and the European Science Foundation. Evan Owens has made presentations at the AAUP, SGML '96 and other places. Chris Biemesderfer has presented our program at a Library of Congress workshop. We have all talked informally to several other societies and to publishers such as Elsevier Science, Springer Verlag, Highwire Press and, of course, AIP and Cadmus. A more complete list is available on the AAS Electronic Publishing Projects Web page.
Copies of some of PBB's papers can be found through his homepage at http://www.aas.org/~pboyce (Sorry for the large graphic there. I'll make a thumbnail as soon as I have time)
We have signed a cooperative agreement with NIST who will provide $150,000 to support our electronic publishing program. We will share our results and provide them with some advice on setting up their own program for their own NIST Journal of Research. The money will be used to speed up development of authoring tools and systems for linking to and interoperation with other journals.
Third year of funding ($200K) is in hand. Grant ends September 1997.
Illinois Digital Library Initiative
Funded by NSF for four years at $1M per year, this large project is concentrating on electronic publishing and distribution in the physical sciences. The AAS will join this effort as a test of systems which allow interoperation (searching, retrieval, etc.) of a remote depository of electronic manuscripts. The AAS is the only organization with a sufficiently well developed independent archive to be in a position to offer a remote archive of SGML manuscripts. The DLI project is developing specialized user interface tools for accessing their internal SGML archive. The tools to access the remote AAS archive are not yet being worked on, although a prototype has been developed using a standard Z39.50 query. Because Z39.50 does not yet support user authentication procedures, we no longer provide a Z39.50 gateway into our archive.
The APS and AIP are contributing their manuscripts to the internal DLI archive, as are a number of other publishers. We have placed the provision of the ApJ manuscripts to the DLI project at a low priority, and have not yet provided them with our archive.
Other electronic journals which charged extra for the electronic version have very low subscription numbers. The typical number is 200 subscribers. In the case of the AIP and APS who initially started out using OCLC to distribute their journals, part of the difficulty can be traced to the difficult and complex OCLC license (a six month negotiation period for the electronic license was common).
Since OCLC has stopped supporting electronic distribution using their proprietary Guidon system, and will also terminate distribution using HTML, AIP and APS will be doing their own distribution. We made the right choice in 1992 to do our own development.
AIP has been trying to persuade Softquad to implement a Web browser which can directly render SGML and complicated math. Softquad fears there will not be a sufficient market to recoup their development costs. Therefore they are trying to put together an agreement with a number of publishers to fund the development of such a browser. They asked each publisher to contribute $75,000 to support the effort. For this fee, each publisher will get the "standard" and "light" versions which they can distribute for 18 months. Softquad keeps the rights to sell the "pro"version. After 18 months, the publishers will get the "standard" and "light" versions at a discount. The participating publishers will participate in developing the specification for the browser. The specifications will be made public, and will be brought to the attention to the various standards committees.
In the opinion of Peter Boyce, Evan Owens and Chris Biemesderfer, the development of a completely separate, proprietary browser is not consistent with the AAS philosophy to avoid proprietary software. We do not believe the cost is commensurate with the benefits we might obtain by becoming part of the coalition. Now that the browser wars have narrowed the market to two players, the development of a separate browser which our readers would have to pay for does not seem to us to be the appropriate path to take. We have not joined the coalition.
Future Directions for Development
New developments have slowed over the last six months as we have concentrated on outreach efforts, development of a reasonable financial model and preparing to bring the main ApJ on line.
In the coming year we will focus upon:
These items comprise a larger effort than we can expect to achieve. In addition, the UC Press will have to spend considerable effort in three operational areas:
- expanding interoperability and encouraging the adoption of standards by other publishers
- developing adjuncts to MS Word and Word Perfect for authoring articles for AAS journals
- reducing time to publication
- adding features - machine readable data tables, video clips, etc.
- continuing outreach efforts to the library and publishing community.
An outline of a development plan for the next three years is presented in another document.
- Clean up and finish streamlining the production process for the full ApJ.
- Work with the new editorial office to achieve a smooth transition by September.
- Accelerate the ApJ Letters production schedule as described above.