by Robert John Stevens, December 9, 2015
I have a great interest in this KSL article, Utah inventions: WordPerfect led the PC word processing market for almost 10 years, because I was fortunate to be a member of WordPerfect’s leading programming teams for eight years.
At the first and only WordPerfect reunion a few years ago André Peterson, who is now deceased, told me WordPerfect was the world’s best selling software and #1 on the charts for six straight years, even outselling Microsoft DOS. His memory was probably credible since he was in charge of all the trade shows where he actively disseminated such good news to customers.
I can’t imagine any programming job better than the one I had at WordPerfect during its glory days. Every workday was a thrill. We were all bright and young in our twenties and early thirties.
On the very day in May of 1990 when Microsoft Windows 3.0 released, along with the WYSIWYG (what you see is what you get) versions of Microsoft Word, Excel and PowerPoint, my team which was developing the WYSIWYG version of WordPerfect 5.1 for OS/2 became the Windows WordPerfect development team.
Within about 12-14 weeks we ported over our beta code to work on Windows but unfortunately, due to feature creep such as manager Eric Meyer’s insistence of implementing sculptured dialogs that appeared to look three dimensional (even though Word had only 2D dialogs), macros because as Eric told me our DOS WordPerfect users expected them (even though Word did not have macros), and developing a component (COM) version of our File Open, File Save and File Save As dialogs and integrating into them our QuickFinder search engine (another feature not found in Word), we added months of work and delayed the release of WordPerfect 5.1 for Windows until November 1991.
Previously I was on the team that ported the DOS version of WordPerfect 5.0 to OS/2 (a text-based word processor). If I remember correctly our team grew from an original six to twelve people by its release.
I also programmed on the OS/2 Shared Code team for WordPerfect 5.1 where I broke the Thesaurus and Speller out of WordPerfect, and developed the first Writing Tools API so Speller, Thesaurus, Rhymer and later Grammatik could exist independently and be used stand alone, but communicate with their host programs such as GroupWise and WordPerfect Presentation.
Deceased programmer and friend Steven M. Cannon took over the Windows Speller before I finished so I could focus on the Thesaurus for Windows where I invented and implemented three scrolling list boxes that shifted horizontally but left highlights and a hand icon as breadcrumbs marking previous words that had been looked up. Apple later used my interface or invented their own for the Mac OS X Finder.
After our programmers began to work on WordPerfect 6.0, I remained behind and with two other programmers fixed thousands of bugs and released the maintenance versions of WordPerfect 5.1 and later WordPerfect 5.2 for Windows (where I invented and first implemented Incremental Search). Despite the negative press for the buggy release of WordPerfect 5.1 for Windows, the press applauded 5.2 and by then we recouped about 33-36% market share for word processing.
Incremental Search is the most successful and most widely used technology to emerge from WordPerfect. It has become ubiquitous and used by billions of people every day.
Management began to lay employees off in 1994 and work became very stressful not knowing as without any notice fired employees were escorted out by security guards. After reading an article that recommended employees resign early to avoid the stress of a declining company, I resigned that month and took a severance which I used to develop Easy Letters and co-found WriteExpress Corporation with Dr. Melvin J. Luthy, a Linguistic Professor at BYU.
In May of 1995 Eric Meyers called and asked that I return to work at WordPerfect. I accepted his offer on condition Novell would give me the rights to Rhymer. They agreed and I began work on the Perfect Office integration team. I was there only a short time and then returned to the Shared Code team where I continued fixing bugs other programmers made in WordPerfect 6.0 for Windows and in the Perfect Office Installation software, until I resigned again in December of 1995 during another massive layoff, ironically receiving a second full benefits package for my eight years of employment.
Read the KSL article at Utah inventions: WordPerfect led the PC word processing market for almost 10 years.
by Robert John Stevens, August 6, 2013
Recently, Google announced their programming tests were not an accurate indicator to discover the best performers. Are our interview methods better than Google’s?
Having programmed with hundreds of programmers, there is nothing so degrading to me than to be asked to take a programming test, especially when I receive an email link to an online test before meeting face-to-face with the employer.
If you too hate programming tests then you may also have difficulty focusing on them.
What if you agree to take an online programming test, get an “A” and still are rejected before given a chance to interview? That happened to me. You can imagine how I felt.
When did you last send your doctor a medical exam before employing his or her services? Would you dare to or do you just assume they are qualified?
Maybe it is time for programmers to follow the examples of doctors, frame their degrees, hang them in their offices and bring them to job interviews.
When companies throw tests at candidates, it suggests they either have only moderate interest in them or there is something intrinsically wrong with their interview or judgement process. It also suggests their programmers are having difficulty networking to attract great talent.
I sent my resume to Google twice in 1998 when they probably had less than six employees. I had an excellent history of programming for big names such as WordPerfect, Broderbund and InstallShield and had just spent six months researching search engines for WordCruncher. I also had completed my coursework for my M.S. degree in Computer Science from BYU.
But Google turned me down both times saying they only hired PhDs.
Why? Perhaps because the Google founders were both PhD candidates and they only wanted to hire programmers at their level.
Perhaps they knew A” people hire “A” people and sometimes “B” people. As companies grow, “B” people hire “B” and “C” people who then fill most of corporate America’s positions.
Perhaps the Google founders wanted to hire only “A” people for as long as possible.
The average programmer programs just 5-6 lines of code per day. A great programmer may program hundreds but probably not during debugging, maintenance and enhancing stages.
Great ideas, architectures and code take time to develop. Perfecting code takes an enormous amount of time and focus. Given these facts, how can the best people be discovered by requiring them to take programming tests?
Startups must be slow to hire and quick to fire. Rather than waste time interviewing, hire programmers for a day or even a few hours and work alongside them and you’ll know how good and capable they are.
One effective way to hire good technical people is described in “The Entrepreneur’s Manual: Business Start-Ups, Spin-Offs, and Innovative Management.” Author Richard White consulted with more than 600 startups and technology companies in Silicon Valley. Buy and read his book.