Daily Archives: May 30, 2013

Old web recollections…

My colleague, Paul Jones, has gotten some great press recently regarding a bit of digital archaeology he did in finding a very early copy of Tim Berners-Lee’s demo web page developed for a hypertext conference in 1991. Paul worked on the academic side of UNC-CH’s computing house, while I was working on the administrative side. I was quite interested in some of the things that Paul was working on in the late 80’s, and envied the latitude he had to do cool stuff. Paul’s work inspired several of the things I subsequently did.

I was an IBM MVS systems programmer for many years, starting in the mid-80’s, as the lead CICS and VTAM programmer then and later as systems programming manager. (I’d also worked on both COBOL and distributed microcomputer, including a port of Kermit to CTOS). We typically focused on mundane things like money and grades in the administrative applications, though. However, I’d gotten interested in the concept of hypertext when I saw (sometime around 1989 or 1990) the DEC Videotext application. I’d just finished writing a VTAM application for terminal menus (actually finishing a design started by other colleagues who’d started the app but never finished) and realized that with a DECNet/SNA Gateway, we could link the VAX with the IBM mainframe and make the hypertext VAX app available on the IBM 3270 terminals. I convinced my boss to buy the DECNet/SNA Gateway, and I wrote a VTAM app to read the datastream from the gateway and interact with the 3270. It actually worked, but it never caught on with the users, and about the time it was stable, along came Berners-Lee with the Web, and the rest was history.

Thanks to Paul, I’d gotten interested in Gopher in 1992, and used it, in conjunction with WAIS, to help publish some administrative/Institutional Reasearch statistics (the “Fact Book”). We’d gotten an IBM AIX server (I called it Bullhead, named after a trout stream in Stone Mountain State Park) for Gopher/WAIS that was a textbed for these efforts as MVS was not a compatible platform at that time for this type of work.

By 1993, Paul was really getting excited about the Web, and after hearing him talk about it many times, and seeing the Mosaic browser and how cool it was to have text and pictures together ;-), I decided that I needed to try it out. So, on November 24th, 1993 (I remember because it was the Wednesday before Thanksgiving and thus a very slow day, and a good time to try such things) I downloaded a Web server source package, set up and ran the “make” and then set up my first web page. While I don’t remember exactly what I put on the very first one, I soon set up a page with a picture of the office building at 449 W. Franklin Street, along with descriptive text about the Administrative Data Processing department. I was working part time on a Masters’s in Comp Sci, and had just gotten to the place where I needed a thesis topic. I was fascinated by the the concept of the Common Gateway Interface (CGI) and decided to do my thesis on application development using the Web and CGI. I wrote the paper in 1994, and finished my degree in the spring of 1995. I tried to save an electronic version of the paper, but the postscript file seems to be corrupt (at least the tools I’ve tried wouldn’t read it). Maybe I can resurrect it some day.

We did a lot with CGI in the years that followed. We enabled TCP/IP on MVS and used CGI scripts on Bullhead to allow us to pull data from MVS databases. That lead to the creation of our first web-enabled class registration tool, Student Central, in the mid-to-late 1990’s as well as a variety of other data access tools. The commercialization of the Internet in 1995 opened the door to the marketplace for internet software, but that’s another story.

I wasn’t quite as close as Paul to the beginnings of the Web, but it was a heady and interesting time. This fall, I’ll have 20 years of web experience…I guess I’m getting old!