Like the shoemaker whose children go without shoes, I allowed the older propylon model to remain undocumented-and therefore unavailable to others-for an embarrassingly long time. When someone actually wanted the model recently, it became clear that I should stop putting off the chore. I set about properly documenting the model with a guilty conscience and a sense of relief that it was no longer possible to forestall the job.
I thought it worth commenting on the nature of the work for those who are similarly inclined to put off the inevitable.
First, the best news. At the conclusion of the process I felt a good deal of weight leave my shoulders and, consequently, a new lightness in my step. Guilt is a heavy load to bear. Furthermore, the process was not onerous and turned out actually to be helpful. It obliged me clean up the file, find some errors, and assure myself that any user would have a clean, full, relatively error-free model to work with.
Since I was documenting a CAD model rather than a database, the needs were different from those outlined in the article about documenting a database here in this issue of the Newsletter (Documentation for Databases: A Crucial Step). The first set of items listed there was needed in this instance as well, but the second set was specifically for databases. For a CAD model there are different requirements since the important job is to make sure all the layer names are clear and that a user can figure out their meanings. In addition, standard views that are stored as part of the file must be explained.
I used the CSA Layer Naming Convention (www.csanet.org/inftech/csalnc.html) to name the layers in the older propylon model; indeed, the convention was developed in response to the problems I first encountered in dealing effectively with CAD layers more than a decade ago. As a result, explaining the layer names was not so difficult. I only needed to explain the way I had applied the CSA Layer Naming Convention to the older propylon model and refer users to the web document explaining the convention. Had I named layers using mnemonic aids or some other shorthand, it would have been necessary to provide a name and a description of the material found on each of more than 150 layers in the model instead of explaining the logical scheme followed.
In the process of checking the layer names to be sure that I had properly specified all, I found, not surprisingly, that there were some incorrectly named layers in the model. In addition, there were some unnecessary layers (that are often impossible to remove in AutoCAD® for reasons I have never fully understood). I cleaned up the errors and did my best to remove the unused layers. The result is not just accurate documentation but a better model. While I suspected there would be some errors and consequent corrections, I did find a few more inconsistencies in the layer names than I had expected (and a spelling error in the copyright statement in the model!).
In the course of preparing the documentation, I realized that one issue could easily be overlooked: measurement precision. Some parts of the model were created with the aid of hand measurements made on site, others with photogrammetric measurements made in the CSA office from photographs taken on site, still others with the aid of W. B. Dinsmoor, Jr.'s text and drawings (portions either inaccessible to me or outside the range of my permit). As a result, some measurements were accurate to the nearest millimeter, others to the nearest centimeter, and still others to lower levels of precision. Those different levels of measurement precision had to be described. More problematic, potential users had to be given enough information to permit them to know which portions of the model could be relied upon for what levels of precision. In the end, I even included photographs to try to assist potential users. (1)
Despite having much of the work done for me by the CSA Layer Naming Convention, making sure that all 150+ layers of the model were clearly specified was time-consuming. It also took some time to write an introduction and to deal with the issue of model precision. In all, it may have taken a day or two of my time, and most of that was spent simply organizing the layer names, checking to be sure all names were proper, nd making the necessary changes in the CAD model.
Although I have previously sent the CAD file to a few people, I am now able to allow anyone to use the model with some confidence that a user familiar with AutoCAD may do so without difficulty. To check the documentation, as now available on the Archaeological Data Archive Project Web site, please see www.csanet.org/archive/csaarchive/greece/oprop/oldprop.html. From that Web page the model may also be downloaded.
-- Harrison Eiteljorg, II
To send comments or questions to the author, please see our email contacts page.
(1) When retrieving measurements from a CAD model, there is no way to tell which dimensions or coordinates are more precise and which ones less so. The system will provide information that seems to imply equal levels of precision for all data. Therefore, the user of such a model must be able to determine appropriate levels of precision independently. Return to body.
For other Newsletter articles concerning the use of electronic media in the humanities or the applications of CAD modelling in archaeology and architectural history, consult the Subject index.
Next Article: CSA/AIA Project Database
Table of Contents for the Winter, 1999 issue of the CSA Newsletter (Vol. XI, no. 3)
Table of Contents for all CSA Newsletter issues on the Web
Return to CSA Home Page