Tuesday, June 30, 2009
StarText People: Larry Groebe
(Note: This is one in a continuing series tracing the history of StarText. Read it from the beginning by using the Archive Post links on the right)
Interview with Larry Groebe
Note: I met Larry Groebe in the early 1980s, almost two years before he became a programmer for StarText. Like others that would work for us, he was a subscriber first. Larry was instrumental to our growth made many contributions, both as a subscriber and an employee. So much so I invited Larry to co-author a book on StarText, which could still be in our future. This interview is from the late 1990s.
Our first meeting was at Rave Computers in Dallas. You seemed passionate about doing videotex even in those days. What was the attraction?
Groebe: I'm not sure how much of this will be directly relevant to your story, but it WILL help you to understand how I was really "primed" for StarText. It also touches on the state of the art in computing and the online world prior to StarText. After all, since years in the computer business work like dog years - that makes 1984 something approaching a
hundred years ago! (and some days, yeah, it DOES feel that long ago.)
I'd been involved with online computing, after a fashion, since back in the spring of 1972. (!) It was in high school - the spring semester of 9th grade - that I was introduced to computers. Back then they called the concept "time sharing," a computing term which I suspect most people today have never heard of. The idea was that since were computers were (obviously) so large and expensive, that people should share them. One big computer could be used by many folks at once if we all shared time on it. It took some clever programming to allow allow more than one person access at once, but the results were worth it. (Among others, DEC made a great success of time-sharing systems.)
In the early 70s a few educators banded together in various regions around the country to make time available on computers. Dallas was part of Region 10; our region's computer was a room-filling Hewlett Packard 2000C tucked away in a building in Richardson. At the other end - my high school - we had a printing Teletype with acoustic coupler with which to reach it. In 1972, I could pick up the phone, dial in, place the handset down on the rubber cups of the coupler, and log on. Literally type in an ID and password and be connected to a computer. At the same time as other people! What a rush! What a feeling of power! What a chance for a ninth-grader! My own computer!
Of course, the computer was pretty much a blank slate. There were a few programs already written for it, but the main goal was to write your own. The sky was the limit; whatever you could think of doing. Actually, the sky wasn't the limit - the printing KSR-33 Teletype was highly limiting, as was the connection speed. Our acoustic coupler connected at a miserly 110 baud - yeah, 10 whole characters per second. Which was about as fast as the Teletype could print, anyhow. So you had your constraints. Nonetheless, my friends and I persevered, and within a year had developed an entire hacker culture in miniature, so powerful was the lure of having a computer at your fingertips.
We wrote all sorts of programs; games mostly, and things that would qualify as simulations if they had possessed any accuracy or detail. We also attempted our first "hacks", finding way to gain access to unauthorized accounts as well as ways to crash the computer.
The fact that we were dialing IN to the computer, and that others were doing the same, wasn't lost on us. At one point along the way I wrote the world's first online chat program. Or at least the first online chat program that I'd ever heard of. With it, I succeeded in holding a two-way conversation with my sister at her school several miles away. I felt a distant kinship with Marconi and Bell that day.
In the spring of 1975, my senior year, our Teletype was hauled away and replaced by a Cathode Ray Tube - a CRT. (also known in those days as a "glass Teletype".) At the same time we were upgraded to a 300 baud modem - a mind-numbing speed of 30 characters per second - as fast as any of us could read! New possibilities were upon us. Aside from the fact that the computer was in another part of the city, we were truly experiencing personal computing, well before the first Altair or Apple.
College nearly undid the whole good experience. Most of Trinity University's computing degree consisted of learning IBM mainframes and their (soon-to-be-extinct) punch-card environment. Despite the fact that the first qualifiably personal computers had appeared in the preceding months, the computer science faculty had no comprehension of their potential, seeing them as a useless toys. Thus, my major stayed in broadcasting and film. Only a book called Computer Lib, published by Ted Nelson, fired my imagination - but it did it spectacularly well. Nelson imagined machines which no one had yet built but which COULD be built; machines which linked text from document to document in a web of pages. He called them "Hypertexts."
Upon graduation I entered the film business, mixing audio for Houston filmmakers. My best friend John Shelton, even more of a computer fanatic than I, was working his way to a degree up at computer heaven - MIT. During phone calls he fed me with tales of the wonders he had seen and used up there - a network of connected computers called the "Arpa-Net" and printers that used lasers to print letters.
John managed to get me a free account on a computer at MIT that could reached via the Arpanet, and I obtained a used TI Silent-700 terminal that printed on thermal fax-type paper. I was able to log into this network from Houston, and soon was hooked all over again. Here on this Arpanet was not only powerful computers, but a human community, and a sophisticated and interesting group they were. You could write electronic letters to them. They published daily electronic articles about science fiction and human-computer interaction. Some of the users were folks I had heard of, like Marvin Minsky. A couple MIT students were building some sort of interactive game called "Zork."
That was online computing in 1980. I lost my free account on the MIT computers after a few years and and had to leave the Arpanet community. When I rejoined it, it had been renamed the Internet.
In the fall of 1980 I returned to Dallas to join the Tracy-Locke advertising agency to run all their in-house audio-visual equipment. In 1981, personal computing was a hot topic (wasn't it just a year or two later that Time named the computer the "Man of the Year?"), and for the already-converted it wasn't hard to see that a revolution was in progress. A friend and I bought an Atari 400 but really coveted an Apple II. Tracy-Locke had just lost the TI electronics account but retained a single 99-4 computer, which I had ample chance to play with. When the agency pitched the Best Products account (a now-defunct retailer along the lines of Service Merchandise) I went along to run the slide show...and to present a vision of our forward-thinking qualities.
Using the TI 99-4 (and storing the program on cassette!) I programmed a demonstration of a futuristic Best Products online catalog which could be dialed up via modems. It involved a menu system that helped you locate and purchase the exact product you wanted, even telling you the amount in stock. I also wrote a two-page summary vision of electronic retailing possibilities (which I still have). I imagined that someday the yellow pages would list modem numbers along with voice numbers. I saw difficulties in selling highly visual items like diamonds, but imagined how space for descriptive copy would be unlimited and possibilities for programmed sales help to assist you in your purchasing.
I saw how sales could be linked to inventories in real-time and realized that the amount of customer tracking possibilities should be wonderfully enticing to any retailer.
Somewhere about this time (1981) Radio Shack's burgeoning TRS-80 business sparked them to partner with a recently-started online service called CompuServe. I became an early subscriber. Loved what I saw, although costs of $5/hour even at 300 baud limited my online time. If the costs could be brought down, this had possibilities. What kind of data COULDN'T be stored on a computer!
So by Christmas of 1982, I wised up and realized that computers were advancing far faster than the Dallas film business ever would. In fact, if I squinted, I could see that personal computers were in a position in society not unlike where movies were in the first decade of the century, or radio was in the 20s, or TV in 1950 - poised to become a media and cultural revolution. I could see myself there. So I went to work at Rave selling Commodore computers - the first computer-related job I could get. And one day, you walked in...
In short: I had seen what computers could do. I had seen what personal computers were starting to do. I understood what a modem was and felt in my gut the allure of online communications. As a pop-culture/media buff, I had studied the mass-media revolutions. I could see it all happening. Of course, my timetable was a little bit off, and I didn't realize at the time that StarText's isolation made it only a transitional medium...
Talk a little about your pre-employment involvement with StarText. You participated as a subscriber columnist. What did you write for us? And can you remember your old ID number?
Well, for the first few months of my StarText involvement I had several ID numbers, since I was using some of the free trial subscriptions you handed out at the Rave store. You had dropped by with trial subscriptions to seed to our customers. Since Rave stores were probably the most popular Texas home computer stores of their day, we could boost StarText's presence. (This was at the height of the early 80s home computer revolution - the era of Atari, TI, Timex, and Commodore. Rave was founded by a pair of brothers from South Africa - the Jacobsens - and rose to become the area's most sophisticated Commodore dealer, locked in a deadly battle with Videoland stores. Rave unfortunately managed it's finances quite poorly and filed for Chapter 11 in 1984. Errol Jacobsen later gathered up a few of his favorite Rave employees and started another company called "Soft Warehouse," which changed it's name in 1991 to become CompUSA. )
Anyway - I would hand out free StarText trials to subscribers who bought Vicmodems, but I also held a few back for myself and used them to get online. I think it took me three or four trial subscriptions before I decided to fork out my hard-earned cash for 3 months of real service at $7.95/month. My ID number, 2419, was assigned 4/21/1983, according to the piece of paper sitting in front of me. But before I got much used to that ID number, StarText underwent a major software revision and my ID became 26091. That's the one *I* remember fondly.
I was looking over some old issues of StarText INK to remind myself of the StarText system in those days and to remind myself of what I wrote! I think I must have produce at least 4 or 5 columns at different times.
The first was COMCHAT, the Commodore Chat. It happened, as I suppose so many subscriber columns did, because you had some hobby you were passionate about. I knew a thing or two about the Commodore 64 computer since I was selling them. That put me in a position to talk about them, review them, test software that came my way, and so on. Which is what I did in ComChat for a year or two. At the height of the Commodore 64's popularity my column consistently made it into the top 10 subscriber columns (if memory serves). I treated the column with the utmost respect - I couldn't have taken it any more seriously if I'd been paid to write it. It was done for free, but it was a connection to a real world of journalism that I'd left behind in college and which, up until this moment, I had not hoped to regain. (Well, I *did* have a couple of book ideas being pursued, but this was REAL)
I recall nights spent slaving over the keyboard and then sending the column at 300 baud to a special editorial email address, where the next day I'd have confirmation from the editors - real editors! - that the column was online. And over the next few days, I'd get email from readers who had read and enjoyed the column. I lived for that feedback. Authors sometimes say they feel they are living in a vacuum, tossing out words but not knowing what impact they make when they land. With StarText, that wasn't a problem. The feedback was there, swift and certain.
Sometime after COMCHAT, I started a second column for a second computer of mine which quickly became my favorite - MOD100 for the Tandy Model 100 "Laptop." You'll remember the computer.; for about five years it was THE journalists' workhorse. The Star-Telegram owned dozens of these notebook-sized ,self-contained wonders. The Model 100 really was a brilliant concept, and the last computer for which Bill Gates personally helped write the software. So this computer was well deserving of a column, and I started writing MODEL100 for it. Later, my Tandy friend Sanjay Gakhar took over much of the column. Eventually, with relatively few companies wrote software for the Mod100, we ran out of things to say in the column. By that time I was hosting a Model 100 user group at the Infomart (in Dallas).
After writing these two computer-specific columns and now being established as some sort of expert, I wanted a larger forum for my views, and from that need came the MICRO-SCOPE column. Of all the effort I put into writing for StarText, this was perhaps my favorite. This was due in part to the format - because Microscope was co-authored with Carrington Dixon. We evolved a unique "point-Counterpoint" style to it and a unique method of collaborative writing. We alternated weeks. One person would write the body of the column. Then he would email the column to the other writer who had no idea what he was about to read. This author would insert his parenthetical remarks (IN ALL CAPS) as a sort of verbal counterpoint to the main topic. Then he'd mail the column back to the original author to make any minor edits or a summary statement, before sending it to StarText. The result had a great give and take. We tackled all sorts of hot industry topics of the day like "What can we do with (name your favorite computer brand) bigots?", "Will the IBM PC take over?", and an annual list of predictions for the year ahead. In a file box somewhere I have most of these columns. I'll have to dig them out. We responded to our readers email in a second column called MICROMAIL.
I had a hand in the formation of another even more collaborative writing effort, the FORUM. I devised the structure of the FORUM based on what I'd seen on the Arpanet; that all responses be collected up each day and republished in a "digest" format. After helping get the thing underway, I rarely participated because the general direction and tone of the discussions weren't of interest.
You'll have to ask Julie Barrett about the genesis of the STARSOAP - "As the CRT Scrolls." My memory is a little fuzzy on this one, although I remember it was a lot of fun to do. It was a mutual thing between the two of us. We brought in about four more writers to help write the weekly chapters, but after a few months they became less than enthused by the constant grind of putting out more plot, and they dropped out. Julie and I finished alone, and Julie did most of the final work. She was always the better fiction writer.
The basic premise involved two rival California families who had competing computer firms. The material ran just this side of soap opera parody (we hoped). The Malfatorres, an old Italian clan, had gotten into electronics as a side business when Prohibition hit their wine business. They now made more money from computers than from wine. They were full of old money and old-money politics and alcoholism. The other company, "Sigma Beta Corporation" was a bunch of young and hungry upstarts, who weren't above corporate espionage to win. CRT was heavy on intercompany love affairs (PG-rated) and technical skullduggery. We spent weeks building to a climactic moment when one company's prototype computer basically blew up in mid-demonstration on the floor of Comdex.
"As the CRT Scrolls" was HARD WORK! Writing our characters in and out of predicaments; coming to surprising climaxes each week and resolving them the next while starting new ones; this wasn't easy. We tried various approaches; writers alternating weeks and writers "owning" groups of characters with an editor weaving together the segments. I remember CRT as much for the work and the thought that went into it as for the actual output. I felt we were performing a grand experiment in this new medium - collaborative online fiction. It might lead somewhere.
Of course, it didn't. There have been attempts since - most recently "The Spot" on the Internet and Amazon's hiring of John Updike to start a chain novel. But it remains a novelty. And interactive computer fiction like "Zork" doesn't exist anymore either. Fiction publishing is such a tenuous activity these days - read within only proscribed formats - that no-one wants these. Or to put it more simply - "You're looking at a screen? Put PICTURES on it, not words."
I'm sure I left out others. I think I had a chapter of the StarNovel. I have a sheet of paper here that indicates I wrote "TRYIT." And of course, as a staff member I wrote all sorts of things - including, ultimately, a column in the Monday Business section of the Star-Telegram. With that, at last I was a legit writer. Funny, because when I first talked to you about someday coming to work for StarText, I imagined myself on the editorial side, not the technical side. Didn't think I could demonstrate enough programming skills, whereas I KNEW I had the writing and editing capabilities. And year after I left StarText I was hired as a writer by Soft Warehouse. It's an odd orbit.
Then were you hired by StarText. How long did you work here? Where has your career taken you since leaving StarText, including your current title and position
I was hired for StarText in May of 1985, after subscribing for close to two years. I'd been working for Tandy from the time Rave stores closed until StarText took me. And honestly, I'd been hinting repeatedly that I'd like to work for the service. I believe other local online services were also in development during this period. Something called "BISON" (the Belo Information Systems Online Network) was getting under way, but its content was weak and it vanished almost before it got started. No, StarText had the field to themselves. And for me, it wasn't even the local slant to the service: it was the flat-rate fee. Still, the local angle had it's appeal - none greater than the opportunity to leave Tandy and work at the venerable Star-Telegram. The Tandy job had been an excellent step after the closure of Rave. I was part of a team on the 16th floor testing unreleased Tandy computers for problems.
This was my first opportunity to stretch my programming skills. I ended up writing a low-level software test suite for the TRS-80 Model 16 Computer. The programs had to be written in assembly language to test basic computer functions like the disk operating system. I also worked on another computer planned as a sequel to the extraordinarily popular Model 100 - the Tandy 600. It was an attempt to cross the 100 with the MS-DOS laptops of the era, and introduce a new Microsoft product - Handheld DOS (HH-DOS) However, the computer was so slow and bulky, and the software so awkward to use, that after release the computer flopped. More promising was another Tandy model which was release while I worked there - the Tandy 1000. This was their first PC-clone. It flew off the store shelves and ultimately changed the company's computer directions. I also got my first look at another transforming product - a Microsoft DOS enhancement they called "Windows" which ran on a Tandy 2000. It played a mean game of "Othello" but little else at that stage.
Anyway, in the spring of 1985, it seems StarText was [in the market for a programmer.]
Since my programming background was slim, and since StarText's technical problems were unique, Joe Donth gave me a test problem to work on. The outcome would determine in large measure whether I was hired. He looked at my code for the Tandy Model 16, then described to me what happens inside the VAX computer when a subscriber logged on. He asked me then to go home and "flowchart" that process. I guess Joe liked it. After work one day at Tandy he called me to come over to the Star-Telegram. Joe, myself, and Gerry Barker walked over a few blocks to this downtown hole-in-the-wall of a bar. It can't be there anymore, because when I visited, just that once, it seemed it had been there for decades. Dark, smoky, and long, like a diner turned sideways, it was a a classic, eastern-big-city, federal worked and newspaperman's rotgut bar. We slid into a dark booth and Joe delivered the news. I was in and StarText was going to lead the next communications revolution. As if I had any doubt. This was Joe Donth at his best, weaving a seductive tale of the possibilities of videotex and the online communications breakthrough that the newspapers would lead. And I would write that story - literally, in all the code that made it happen. When I left at 6:30 I was so pumped I ran the 8 blocks back to the Tandy Center in a time that an Olympic sprinter would have been hard to match.
So I walked in, anxious to go to work. I have a handwritten flowchart dated 5/30/85 which is titled "Version 4 UserLogin Module" and must be the first thing I started on. So much to learn! I knew StarText only as a user. Now I really had to get to know the staff and find out what life was like on the other end of the computer screen. In fact, there was the whole paper to learn about. Getting to know the Star-Telegram was a delightful experience. So much history and tradition for an amateur historian like myself. Every day I walked into a building that literally hummed and shook with the sound of the presses spinning out the product. And the front of the building carried the very real presence in the air of Amon Carter, the scion of the paper. That conference room of his, kept just so, like he was away on a goodwill tour and expected back shortly. Even the closets in that area were stuffed with mementos left over from the era; newspapers and awards and random bits from the Amon administrations. Even though the paper had recently been bought by Capital Cities, the history was there. Right there. up in the front room for instance, was the office that housed the original studio of the mighty WBAP radio. Not hard to think, walking into StarText, that we might be recreating a bit of the magic that surrounded those pioneering efforts.
Even the paper's editorial staff reeked (creaked?) of the honored authority of the newspaper tradition. You saw it in the long-term staff commitment; people like Jim Jones and Jerry Flemmons and Elston Brooks who had been around that newsroom FOREVER.
StarText's editors sat at a terminal at the end of the news room. The rest of the staff were wedged into offices in and around the data processing department on the second floor. StarText had one programmer; the other IS functions had half-a-dozen others. And then there were the computers: VAXes and PDP 11s hidden away from the world in there own bright, false-floor, fireproof computer room and guarded over by an ever-changing cadre of computer operators.
The VAX was, fortunately, a friendly computer, and my friend Paul Barrett had access to one so he gave me a crash course in operations. Nonetheless, I never strayed far from Digital’s VAX documentation, which consisted of enough 3-ring binders to fill a four foot bookshelf. And that was just the VAX operating system. The programs you wrote yourself, of course, and StarText was written in a language called DBL. This was an extremely obscure offshoot of an extremely obscure language called Dibol, and if I ever DID know why DBL was StarText's language, I've since forgotten it. DBL was escribed in a single 3-ring handbook. It seemed to be a mishmash of BASIC, FORTRAN and most other similar languages, so I picked up the rudiments fairly quickly. I'm in possession of a printout dated June 8, 1985 which is listed as the "Version 4 Login Program. " and carries a version number of 0.1 (!) I must have been able to do SOMETHING in 9 days.
The USER program was the single largest piece of code I've ever worked on. With all the linked libraries, subroutines, et al, the printout, stacked, was two or three inches thick. I eventually got to know it well enough - and rewrote the whole thing enough - to literally visualize it in my head, the flow of the code as it looped and whirled through the program. But the USER program was only one of over 100 programs that helped StarText run. The software tended to cluster around key system features and functions:
AAEBOOK, AAELIST,AAETEXT, and the like were for converting the Grolier encyclopedia into online form. There were bunches of classified-ad-related programs. And billing-related programs like CREATESUBS, DOCUMENT, REPORTER, and the SCREEN program. USAGE, QUICK, QUICKER, and QUICKEST for reports. Dozens of others. A lot to keep track of, so it was great to have first Sheri Suggs and later Steve Smaller as help)
I stayed with StarText for 4 1/2 years. During that time we added an immense number of features, among them...
· Grolier's Encyclopedia
· The FilmVault
· Credit card billing
· Weekly stocks
· Changeable passwords
· Time -restricted accounts
· Federal Executive Board
· Online surveys
· Email enhancements
During that time I also worked on introducing the "Seven Days of StarText" package; hosted the monthly Infomart meetings; wrote TechTalk for StarText INK; wrote a weekly column in Tarrant Business, gave talks to user groups and appeared on local radio talk shows on behalf of the service.
1989 was a year of intense highs and lows. In March, I was honored as the second
Star-Telegram "Employee of the Month." The first person to receive the award was a maintenance worker who had been there 30 or 40 years; I was number two. Heady stuff. As was a trip to the "Videotex Industry Association" trade show in June, where I spoke of the success of our service, our plans for the future, and took notes on what others
were doing (none of whom, so far as I can tell, are much in the business today. Only the phone companies - and their product is even further removed from what it was then, than StarText's.)
By September 1989, however, Joe had I had wildly differing views of the direction to take StarText, and how and where to allocate resources. It took until September before I had the courage to quit. I consulted with everyone I could think of - even my father. It was one of the hardest things I had ever done. It was the giving up of a dream. But it became necessary.
So I left. Immediately spoke to U.S. Videotel in Houston whose Minitel -based service was showing promise. They couldn't figure out where I would fit. Neither could Jeremy Halbreich of The Dallas Morning News, who wasn't ready to get back into the online business (and indeed wasn't for another 6 years!). I even applied at Soft Warehouse as a salesman during the Christmas season, but was turned down. That shocked me. I was stuck doing minimal freelance computer work until a longshot application to a contract firm paid off. In December I was asked to go to work on a contract basis for IBM. The language was APL, which I'd studied in college and which was rare enough that they were willing to hire anyone for. The IBM job was a relaxing six-month vacation. There was a team of 11 people doing less programming together than I had maintained solo at StarText.
I didn't go back online much at StarText. I checked in occasionally, but there were too many memories.
In May of 1990 a friend from my old Rave days who had jumped to Soft Warehouse called me with word that there was an opening at that company again. As a writer. Surprise! And soon enough, I turned into a writer who did programming, just as at StarText I was a programmer who did writing. I spent seven years at Soft Warehouse, during which time the company became known as CompUSA and grew from 16 stores to nearly 100. Our marketing efforts played a large role in that growth. In particular, I had sole responsibility for creating the CompUSA catalogs that were so popular for many years.
But after seven years, CompUSA had grown into a large corporation with a matching large corporate bureaucracy. And we were repeating ourselves. My last major project at CompUSA was to create the company's web site, which went online in the fall of 1995. Much of the time was spent preparing the company for the project, since few believed in the possibilities of the online revolution. The investigation phase alone took nearly a year. The initial design of the actual web site took only a few months. CompUSA started realizing revenues from it almost immediately, and now, from what I understand, it contributes a nice piece of revenue.
After getting the CompUSA web site going, it was time for ME to go. So in the summer of 1996 I joined with a few other former CompUSA employees to start the Insider Creative advertising agency. A former CompUSA vice president, Ellen Miller, left in 1995 to start her own marketing consulting agency. In the summer of last year she was ready to expand and I became the first employee of Insider Creative. Our clients are high tech firms doing business in the retail industry; names as big as Microsoft, IBM, and Compaq. I'm one of two Creative Directors. I do most of the writing, and yes - I do some programming on the side...including our company web site. www.insidermarketing.com. It's not StarText, but it's not trying to be.
Posted by G Bark at 8:42 AM