DYABOLA, or the DYnamically Accumulating dataBase on Objects and Literature about Antiquity, is a software package for the DOS operating system that allows the user to access and search up to three important archaeological bibliographies: those of the DAI in Rome (for works from 1956 onwards, dealing with archaeology in general) and Madrid (from 1990 onwards, specialising in Iberian archaeology), and that of the Römisch-Germanisches Komission in Frankfurt (from the mid-1980’s onwards, specialising in scientific and theoretical approaches as well as the pre- and early history of Europe). The archaeological aspect of these databases needs to be stressed; works deemed to lie outside this field, such as purely philological works, are not included. The data comes on a CD-ROM; the user can however choose to copy the data to a hard drive, and work off that instead (as we do here at McMaster). The makers are continually updating the databases, with a new CD being shipped in the spring of each year.
This review was occasioned by the release of DYABOLA version 3.1, which we received in late spring 1996. The program is substantially the same as version 3.0, reviewed by the present author in the BMCR for September 1995 ( BMCR 95.9.27). An in-depth discussion of the program was given there; to avoid repetition, I refer the reader to that review for detailed information about the program’s functions and operations.
The makers have updated the platform requirements: at least a 386 running MS-DOS version 4.0 or higher, with 640 KB RAM, of which 540 KB is available for DYABOLA, 300 MB free on the hard drive (if all the databases are to be installed there), and a graphic standard EGA, VGA, or SVGA card is needed. The only printers supported are those of the EPSON LQ series, or ones fully compatible with them. A CD-ROM drive is also necessary. A mouse is not supported. DYABOLA can also be configured to run on a Novell NetWare network. There is no Macintosh version.
A fast machine is clearly advantageous. At McMaster we have been using a Dell Dimension XPS P90—a Pentium based machine running at 90 Mhz, with 16MB RAM—and have found the performance to be highly satisfactory. In terms of hard drive capacity, 300 MB should be sufficient for a few more years; at present, just under 203 MB is used, with the DAI Rome database taking up the lion’s share (145 MB). The yearly updates, of which we have received two, have so far averaged about 10 MB each.
As noted above, this version is much the same as the previous one. Perhaps the most obvious change is in the installation procedure, which has now been completely automated. A series of screens asks the user for setup information: whether or not to copy the databases to the hard drive; which database(s) to install; drive letters for the floppy, CD-ROM, and hard drives; which of the supported languages (English, French, German, Italian) to use; and which printer port to use. During this procedure DYABOLA will check that there is sufficient free space on the hard drive; if there is not, the user will be warned. DYABOLA will also advise the user to ensure that the number of files and buffers in the config.sys file is set at 30 for each; if not, DYABOLA will make the change itself if the user so desires. After these preliminaries have been completed, DYABOLA will be installed. An on-screen notice warns that this could take up to 60 minutes, but here the speed of the user’s system again comes into play. In our case, installation of the program and all three databases took less than 15 minutes.
This is a great improvement on the previous procedure, in which the user had to manually create the directory structure needed by DYABOLA and then copy the program and data files to the appropriate locations. As before, if the user is installing an update, any local data files must be backed-up before the installation, as they will be deleted. Now though, DYABOLA does warn the user that such a back-up is necessary. A log file is also created by the installation; if DYABOLA does not work, the makers ask that copies of this file, the autoexec.bat file, and the config.sys file be sent to them for analysis. In general, the installation procedure is now much more professional and user-friendly.
The parameters menu now has two more options; neither is explained in the manual. The ‘show subjects immediately’ option results in the subject heading(s) under which DYABOLA has classified an entry being automatically displayed as a result file is reviewed. Presumably, if an entry is an article from a collection, for instance a festschrift or conference proceedings, the ‘include parent title in result’ option should result in the entry for the collection being automatically included in the result file. In practice, I have not been able to discover how this option actually works.
Another change has been made in the way searches are defined. Previously, chronological limits could be specified by the user before the database was searched. Now, the whole database is automatically searched each time, and any chronological limits must be applied to the result file through the ‘Tools’ menu. Doing so will replace the original result file with the newer version, so if the original might be needed, a copy of it must first be made. It is hard to say whether this change is for the better or not; on the one hand, there is now the potential to always have at hand all the entries for a particular search, but this can lead to very large result files, take up more time, and create some extra work for the user, especially if only a specific chronological range was wanted in the first place.
As in the previous version, if, when the user is reviewing a result file, s/he comes across an article from a collection, there is the opportunity to manually call up the entry for the collection and add it to the file, after which it is again possible to call up a list of contents for the collection and then add any of these to the result file. Now however, when the display results function is exited, a dialogue box is supposed to appear on-screen, asking the user whether to make these additions a permanent part of the result file. I have found that this box does not always appear, and when it does not, the additions usually will not be accepted.
The usefulness of the ‘export’ function has been improved greatly. Previously, only 100 entries at a time could be exported from a result file. Now, the amount has been increased to 300 (the amount given in the manual, not 400 as stated in the readme file that comes with the program). Given the number of references even a seemingly simple search can produce, this is a major step forward for users who routinely want to transfer their search results to another program. The program now also asks for the appropriate drive letter before exporting to a floppy; it still does not, however, check to ensure that there is actually a floppy in the drive before beginning the export process. If there is no floppy, the program will hang up and the computer must be rebooted. As before, exported data is written to a file called ‘dyabola.txt’, overwriting whatever was already in that file. Therefore, the user must again remember not to export a fresh set of references until the previous set has been copied to another file. Also as before, the DYABOLA character set is lost in the export process, so that some or all letters in certain languages, for instance Greek, will not come out properly. There seem to be no problems with French, German, Italian, and Spanish.
A last change to note is the fixing of a bug in the output process under the ‘print’ option. In this option, each entry is printed out as it appears on screen, along with all associated subject headings and references to reviews. In version 3.0 the reviews were not printing out; now, this print option works properly.
Although the program has been improved, there are some areas in which further advances could be made. The manual remains clear and helpful, and has been updated. However, it has not kept pace fully with the program. Some program features are not discussed—three of the parameter menu options are omitted. Some are different—the ‘search by local data’ option has been renamed to ‘publisher, body,
Next, DYABOLA remains a DOS program, although it can be run from Windows upon the creation of a suitable PIF file. A true Windows program would however be preferable. But even in its DOS form, the lack of support for a mouse is regrettable. Given the amount of selecting from menus that the program requires, many users would, I believe, find that a mouse would ease their work. Concerning the printing process, it remains very surprising that laser printers—the industry standard—are still not supported.
One of the previous bugs still exists: if a result file arranged in alphabetical order is scrolled through backwards, past the first entry in the file, then instead of next displaying the last entry in the file as would be expected, an entry from somewhere in the middle of the file is displayed and the entry-numbering becomes confused. As before though, scrolling forwards through the whole file will clear up this problem. Also as before, dialogue boxes tend to open up at random locations across the screen, rather than in a consistent spot. There also seems to be a new bug. In our copy at any rate the ‘search by periodicals’ function does not work properly. Rather than producing a complete list of the contents—articles and reviews—for the chosen periodical, only the reviews are obtained.
A user may well find that a work known to him/her has been inadvertently omitted from a database. Given the quantity of material dealt with, this is perhaps not surprising; we can hope that further updates will also continue to fill in the gaps that remain. What is somewhat more immediately irksome is the continued lack of consolidation of author names, so that, to give but one example, entries for K.M.D. Dunbabin can be found under Dunbabin K., Dunbabin K.M., and Dunbabin K.M.D. (as well as the misprint Dunabin K.M.D). Although in general all the variant forms of a name will be right next to one another in the author list, it is still bothersome to have to make two or more searches rather than just one. Names with accented letters are another trouble spot; the works of Y. Thebert, for example, can also be found under Thebert, Y. It might no longer be feasible, however, to substantially change the databases at this point, so these quirks will probably have to be endured.
In any case, it must be admitted that the few problems, though they may be irritating, do not invalidate the overall utility of the program. DYABOLA remains easy to learn and use, and on the whole, does its job quickly and well. In the two and a half years we have been running it at McMaster, it has been heavily used and has proven its worth to all of us who need regularly to compile bibliographies on archaeological topics. Improvements have been made, and we look forward to further upgrades.
I should also mention here a note that is displayed whenever the RGK database is accessed: due to financial constraints, their library can no longer afford to print a catalogue, nor can it keep up with purchasing all the relevant literature. Therefore they ask for offprints of articles on prehistoric and early Europe to be sent to them. With respect to books, series, and journals, they mention the possibility of some sort of exchange process. As they say, it is in everyone’s interest that their library remains as up-to-date as possible.
When we purchased it, DYABOLA cost 4122.50 DM for the access program and all three databases updated through 1993. The last full prices I have are for 1995: 4642.00 DM for the whole package. This year, the updates for the three databases cost a total of 430.00 DM. DYABOLA is available from: Projekt DYABOLA, Postfach 45 01 44, D-80901 München, Germany. Fax: +49/89/32 35 21 82. E-mail: info@dyabola.en.eunet.de.