Establishing a graphics standard

In 1987, Old Dominion University began its “Office Support” initiative. Though the crux of its efforts was to spread microcomputers and LANs throughout the administrative campus (student labs had already been taken care of), a vital component involved software. The realization that more access to equipment could lead to requested support for a wider variety of application software was frightening. Computing Services could not possibly support all the software available in the market and do it right. For this reason, the decision was made to establish standards and to rely upon the standards previously established for the public labs: WordPerfect, Lotus 123, and dBASE III+. Again, these were only standards; the customers could purchase any word processing, spreadsheet, or database package they wanted. And if they had problems Computer Services would try to help, but the degree of expertise would not be as high, nor would site licensing be sought, or short courses offered, on anything but the standard products. Standards could be changed of course, and application areas could be added.

By late 1988, our customers began to request graphics support. Computer Services decided to evaluate the market and establish a standard before people began to purchase their own preferences, and decided to concentrate on presentation graphics. There were other areas of interest, particularly CAD and animation, but these were more department specific. Presentation graphics could be used by any department.

A key element to the success of this project would be that the users be able to do the work in their own offices. This meant that our solution had to work on their current environment (the University is mostly DOS- based) and that it could not be cost prohibitive. The more costly aspects of graphics are the output devices. To meet this need, a centrally located Resource Center would be established. People whose departments could not justify the expense of a color printer, a film recorder, or a postscript (black and white) printer could use the equipment at the Central Facility with supplies being their only cost. Another requirement was that the product be compatible with the word processing/desktop publishing packages. Though Computer Services had not yet chosen a standard for desktop publishing, PageMaker, Ventura Publisher, and WordPerfect were all used widely enough to require consideration for graphic interface support.

Several of the industry-standard magazines and periodicals were reviewed and what seemed to be the top three software packages were chosen for evaluation: in the order in which they were rated by the NCGA shootout, these were Zenographics' Pixie, Ashton Tate's Draw Applause, and Software Publishing Corporation's Harvard Graphics. Harvard Graphics had been the number one choice for the previous two years, so the fact that it had dropped to third place aroused curiosity. Additional equipment is also needed for presentation graphics and a film recorder and a printer were suggested as a beginning. Also, since the final output is extremely important in this area, the project team (of two) agreed to do the work on low level computers (8088 and slow '286) so that the money could be used to buy the best output devices funding would allow. For the evaluation, funding for the three software packages, a film recorder, and a color printer was requested and received.

The film recorder purchased was the Matrix PCR. Without exception, every article we read rated it well over the others in quality. Though it was quite a bit more expensive than the others, it was considered a University resource instead of a departmental one, and deemed cost justified. The printer issue was more difficult. There was no clear- cut choice among the reviewers and there were no local vendors to visit for evaluation purposes. Graphics was not big in our area at that time; it still isn't. A few phone calls were made to the D.C. area to investigate what those heavily into graphics were using. The choice was narrowed down to the Tektronix 4693D and the CalComp Plotmaster. Though the quality of the Tektronix was preferred, the Plotmaster was chosen. The output was good, and the price, at a 40% educational discount if purchased directly from Calcomp, was the deciding factor.

Once the installation was over, the PCR was a dream. The colors and resolution were great, and, once set up, the program which allowed batch use of the system was easy to use. Since the PCR was consistently the top-rated film recorder, any package that supports output to a film recorder seems to support the Matrix (PCR). Printed output was not as easy. Though the advertisements claimed that the Plotmaster worked with Harvard Graphics, Ashton Tate, and Zenographics products, it did not work with Draw Applause or Pixie, the new kids on the block. (The claim was not false; the Plotmaster was supported by other products in the Zenographics and Ashton Tate line of graphics. We had simply made an incorrect assumption.)

Evaluation of the software was more involved. The project team realized that the best way to evaluate the products was not by following the “training” section of the manuals, but by finding a “real, live application” for it. Not one, but three presentations were due within three weeks of the arrival of the equipment, so software evaluation commenced quickly. Pressed for time, Pixie was used almost exclusively after the first few days. It was found to be the easiest to learn and use. Draw Applause was used to get a few special effects slides such as rotating text.

After this project, the concern arose that perhaps the time element had had too much of an impact on the decision. After all, each package had its strengths - something the others could not do. The project of producing the University's “Statistical Profile” manual was then undertaken. Previously, this annual document had relied on graphic output from Lotus 123, so there was little concern about being able to improve it.

From this project the choice of Pixie as the University standard was confirmed. It was the easiest to use, and it would easily perform all the tasks required to do the project. Draw Applause, though great for special effects, was deemed too complicated for the average chart, and definitely for the average user. And Harvard Graphics was too frustrating. It sometimes seemed that it could do nothing right. It could do 3D pies, and explode a slice, but it could not do both; i.e., it could not explode a 3D slice. It supported the Plotmaster, but only 7 of the 264 colors. Labels were limited to 12 characters. To make them longer, one had to go into Draw/Annotate which is not the simplest of processes.

Further research showed that another plus for Pixie was the Zenographics family of products. Pixie files could be directly imported to Mirage, which is a highly-rated professional product widely used by graphic artists. Mirage has been called the “lear jet” of graphics packages and is often used in evaluating and comparing output devices. With Mirage to do the special effects, Draw Applause was no longer needed. Mirage was found to be heavily used in the D.C. area and meetings were set up. The meetings with World Bank, USGS (U.S. Geological Services), and the Department of Energy were very helpful. We visited the graphics departments and were surprised that they all used DOS machines and Zenographics software. We asked each “Why not Macs or mainframe?” and each responded the same: mainframe output devices were not as good, and the Macintosh did not yet have good color output support.

One of the Pixie/Mirage benefits was that customers could create relatively fancy slides themselves with Pixie, yet if they wanted something even fancier, they could take their files to the graphic arts department where the professionals could add artistic embellishments with Mirage. This would be of benefit to both parties; the users would be getting what they wanted, and the graphic artists would spend less time with the more “mundane” aspects of the design.

Finally, Zenographics Metafile, (since then upgraded and renamed ImPort) a program which converts a variety of graphic formats to a format that can be used in Mirage, was deemed essential. With this product, images from SAS, Harvard Graphics, and many other packages, (even from software in non-DOS environments) have been brought into Mirage. Thus, enhancements and output services can still be provided to users with software other than the chosen standard.

The next step was marketing. Of great help to this project and particularly in this area, was the Associate Vice President for Computing and Communication Services. Not only did he make a commitment to find funding for the project, but during his meetings with faculty, staff, and administrators, Mr. Hamage briefed them on the current research project and either brought them by, or encouraged them to give us a call. Most of the time, we spent approximately 15 minutes teaching them Pixie and then let them create their own charts. On a few occasions, we created the entire presentations for them - usually these were for presentation to the ODU Board of Visitors. Finally, the time came to announce the standard to the University. A special presentation was prepared which used four concurrent slide projectors synchronized with a programmer (graciously loaned to the University by Norfolk and Southern Corporation) and background music. This show was presented first to the Vice Presidents and later to the Deans. They were quite pleased with the capabilities they saw as well as with the offerings of the program (the site license price and output capabilities with supplies as their only cost). Next, we ran an article on Pixie in our monthly bulletin and another about the Zenographics grant and site license in the newsletter, both to publicize the availability of the product.

The research continues. Recently a GPIB board replaced the MVP Star board (which comes from Matrix) for output to the PCR. The color is much richer: though the PCR is capable of 16 million colors, the MVP Star board is limited to 256 colors per file. Though more than 256 colors per image might seem excessive, the need becomes clear when comparing shaded (gradated) backgrounds or using bitmapped images with fleshtones. Additional equipment has been purchased, a color scanner, a targa board, and a high resolution monitor. These have allowed us to include color scanned (bitmapped) images as part of our graphics. Output to a VCR through Mirage using the targa board is the current project.

In summary, our recommendation for developing a graphics standard is as follows. First, realize the need for a standard (or, if you are fortunate enough to have enough staff for adequate support, standards). Next, based on hearsay and magazine evaluations, choose your top three or four favorite software packages. Then, play with the packages following their “tutorials,” but realize that the purpose of this is more to get acquainted with using the package than for a true evaluation/comparison. The next step is the evaluation. Find a true need for graphics in your school, and volunteer to take on the assignment. This is when the true strength of the package comes out - it is not how many things a package can do, but how well it can do them. Remember to look at types of charts, black and white as well as color capabilities, fonts, ease of learning, ease of use, compatibility with other software (both for importing and exporting purposes), and output support. And finally, market your decision. Use newsletter articles, short seminars, meetings with individuals or departments, and presentations. And remember that from now on, whenever possible, your presentations should take advantage of graphics.