Holemans |  About  |  News  |  Hidden Art  |  History  |  Newsletter  |  Donate

Virtual Translator Workbench - Holemans

1981 - Virtual Translator's Workbench (Wycliffe/SIL/JAARS)

Summary of the Virtual Translator's Workbench for Wycliffe/SIL/JAARS '(1981-83).

In 1981, we discovered that Wycliffe Bible Translators was using PDP-11s with RT-11 operating system. I had used this in college and was using it in my work at that time. This work was done by SIL, which handles the descriptive and applied linguistics work of WBT. They developed a library of specialized computer software to assist in linguistic field work. Prior to the commercial introduction of battery-powered portable computers that could be easily taken into the field, SIL field workers used VT-103 minicomputers. This included specialized computer software to assist in linguistic field work such as the Direct Translator Support (DTS) software.

The Unix operating system, which is what Linux is based upon, was used as the inspiration in 1980 for the Virtual Operating System project at Lawrence Berkeley National Laboratory. Unix has a philosophy which emphasizes building simple, short, clear, modular, and extensible code that can be easily maintained and repurposed by developers other than its creators. This philosophy favors composability as opposed to monolithic design. I wished to encourage SIL to take advantage of this idea.

My attempts at this were not advancing. The first and biggest objection was performance. To address this, we proposed that we be allowed to buy one of their computer systems upon which we would port the VOS to RT-11. Though not without effort and many bumps, they made an exception to allow us to buy one of their computers and work remotely from California. At that time they were coding in Pascal. There was a free NBS Pascal complier for RT-11 from DECUS.

The plan was to first port the code to use the RT-11 operating system specific features and the NBS Pascal oddities. After this functioned, the idea was to determine a proto-type collection of translator-needed features, attempting to glean features from the VOS code. Then, code enough of these virtual machine routines to then implement a few translator tools. This would be followed by a collection of examples using these tools.

There were "workbench" variants of Unix from Bell Labs such as the Programmer's Workbench (PWB) and more closely related to a Translator's Workbench, the Writer's Workbench (wwb).

As a side effect, the tools would be more portable to any future systems. At the time of the publication of the ACM article on the VOS, there were five OSes supported and many more followed. As it turns out, several different systems would be in Wycliffe's future (MS-DOS/MacOS/Linux/Android).

After doing part-time work on this, they switched to using C. So, the project restarted. The need to port the code to Pascal was dropped. I started to learn C. Before I got to the point of designing or coding any translator features, I went to a Wycliffe computer conference at the JAARS campus in Waxhaw. While at the conference, I met GerardJan Vinkesteijn, who was a Unix fanatic. This meant that there was an on-site person who completely understood the philosophy.

With this knowledge, we stopped the project. To make the most use of the system we had bought, we then donated the computer to a Wycliffe translator in Papua New Guinea.

Holemans |  About  |  News  |  Hidden Art  |  History  |  Newsletter  |  Donate
julianlocals.com/history-1981_twb Page updated: Thu, 21Oct'21 12:19am EDT