Lacking Natural Simplicity

Some time ago I linked to a description of the POISE DMS-Plus, a Data Management System that I've used extensively since the mid 1980s. Last time I knew it was owned by Jenzabar. Since Jenzabar's description of it (as part of Jenzabar PX) has vanished from their web pages I thought it would be useful to have a description of it on my blog.

I'll include Jenzabar's description of the POISE DMS-Plus here:

DMS-Plus

Each Jenzabar PX Administrative application module is built on Jenzabar PX DMS-Plus®, a relational database management system (RDBMS) using fourth-generation (4GL) language. DMS-Plus includes a powerful report-writer, an information-retrieval system using multiple keys, screen generation, and has multiple user levels: the operator, intermediate user, advanced user, and application developers. It also includes an Administrative Systems Query (ASQ or ask) language that is designed to make the software easier to use for administrative personnel. ASQ permits users to select, order, display, and print information through simple English-like commands. In addition to ASQ, another option for queries and reporting is QBF (Query By Form), a point and select method which leads users through the data files and query options.

The benefits of a DMS-Plus approach to administrative processing include ease of use by nontechnical personnel, a greater flexibility, and speed and simplicity when implementing changes. It includes a powerful data management and information retrieval facility with right-to-use source code available. DMS-Plus includes list processing as well as interfaces to stand-alone word processing systems. Future enhancement to DMS-Plus are provided as part of the ongoing subscription services. DMS-Plus is the common language for all types of users.

DMS-Plus Highlights

• DMS-Plus provides Report Generators for easy sorting, selecting, and printing along with Screen Generators for easy data entry, update, and review.

• DMS-Plus utilizes data manipulation for inserting, updating and deleting data in the database. In addition to data definition for adding new elements to the database, data control is also available for accessing data by persons who have been granted access.

• DMS-Plus has a menu processor, import/export capability via ASCII file formats, a WordPerfect interface, and SQL compliant.

• DMS-Plus is a user-oriented system that operates in a multi-user environment. Simple English-like language prompts and responses are used to interact with the system. By utilizing the flexibility of DMS-Plus, each user office can determine the information collected, can regulate work flow, can set its own schedules for updating files, and can produce its own reports without depending on Information System services.

• DMS-Plus supports multi-user access. These application systems use many of the same functions and techniques that provide continuity across user departments and also allows cross-utilization of personnel.

• Many applications for information management needs may be implemented through the software provided with little or no additional programming.

• DMS-Plus provides an integrated base of information for administrative data processing. Information introduced into the system may be carried forward into all appropriate areas without redundant entry.

While the above describes the POISE DMS-Plus as a “relational database management system”, it isn't really. It does not use a relational database. It comes from the time before relational databases.

To me, a database is a program that controls access to a collection of data - you can’t get to the data without asking the database program for it.

A data management system (the DMS in DMS-Plus) just lets you associate related data together, and then multiple programs can access the pieces, usually using a standard API implemented as a library.

Before the advent of Unix, files were often much more complicated that just a stream of bytes. For instance, under VMS (where I've used the POISE DMS-Plus the most) the Record Management Services (RMS) provides sequential, relative, and indexed file organization, and fixed-length and variable-length record formats, and allows you to access records within these files sequentially, directly by key value, directly by relative record number, or directly by record file address. It was common practice for programs to each use their own specific mix of these aspects of RMS, with the details directly coded into the program.

The advantage of the POISE DMS-Plus is it provided a suite of programs that worked together, allowing users to describe the format of the files they wanted, enter and modify data in those files, sort and select data, and produce nicely formatted reports, without having to write any code. The programs were interactive, issuing a series of prompts to the user for what they needed to do next, and were very easy to learn. If a secretary wanted to keep track of the inventory of their office, they would run DMS:DESCRIBE to describe the fields they wanted and allocate space for the file, then run DMS:ADD to add data, DMS:SEARCH to search and update it, DMS:SORT to sort and select data, and DMS:PRINT to produce a nicely formatted report. If they had to do a lot of data entry into the file, or it was large or complicated, they could run DMS:SCREEN to define a text user interface screen format for interactive user input, arranging the fields or subsets of the fields on the screen to make data entry easy, and then run DMS:SCOPE using that screen format to do data entry. (I keep wishing for something as easy to use as DMS:SCREEN in the area of database backed web applications, but haven't found it yet.)

POISE DMS-Plus files were relational in that it was possible for fields in one file to refer to values in fields in another file, similar to SQL JOINs and FOREIGN KEY constraints. So, for instance, it was possible to have a code field in one file, but on reports include the description of the code by pulling it from a code file, keyed by the code field.

All the POISE DMS-Plus programs used a documented API, the Support Procedure Library, which programmers could use to write programs that accessed POISE DMS-Plus files for applications that required sophisticated, special purpose processing. Furthermore, they documented the file structures they used, so it was possible to write programs that accessed DMS-Plus files directly, if necessary.

POISE, by the way, stood for “People Oriented Interactive Software for Education”, and it probably had its greatest success in systems for education administration, but it was a general purpose tool and I have worked on projects using it for many other areas, including local and state government, real estate sales management, oil and gas royalty management, and many others.