The Rise, Fall and Survival of Ashton-Tate's dBASE
eweek.com
eweek.com
While Tate had a productive relationship with Ratliff and realized he was important to the companies success. Ed Esber fought with Ratliff. He famously told dBase's inventor that his contributions to the company were no important to the success of Ashton-Tate than the guy on the loading dock.
Ratliff quit and started Migent to compete against Ashton-Tate. Esber sued the young company over appropriation of trade secrets. You can trace the company's decline from this point. Ratliff introduced an innovative client server database called Emerald Bay but lacked the marketing chops to succeed. SQL was beginning to take off and when Microsoft introduced Access and made it easily accessible to the same customer base that dBase targeted it was the death knell.
Clipper was also abandoned by Computer Associates who bought it from the original Nantucket Corporation. There are however two open-source projects that are still being actively developed: Harbour (http://www.harbour-project.org/) and xHarbour (http://xharbour.org/). Both are being used by enterprises that still has xBase apps around.
I have fond memories of DBase, especially DBase III+. I remember when computers had "turbo" buttons and you could make a dog program run like lightning by simply hitting the button. I also remember compiling DBase programs in Clipper, and all sorts of other fun stuff. I worked for one guy that bought faster crystals to make his CPUs run better -- the very first PC overclocking that I ever saw.
Eventually, the home/small business database market went to Microsoft, who released MS Access. All of the "real" programmers said it was a toy and wouldn't deign to use it, but consumers bought it up and started creating their own apps. This was great news for me, because most of the time they created a mess and somebody had to come in and clean it all up. DBase was a big part of how I went from high school kid to independent programmer.
Good times.
dBase was very expensive, shooting itself in its own foot.
The danger of a 50000% profit or whatever Ashton-Tate accomplished, is a business culture builds up that requires 49999% profit margins to survive, when prices push down to normal levels via competition, collapse is inevitable.
Great historical context here, by the way: http://weblogs.foxite.com/joel_leach/2011/10/13/why-microsof...
I still miss it. I work with Delphi and later python (I like so much python) but still feel cheated by how much extra work is necessary.
I'm dreaming in a new language that mix python/foxpro and have sqlite integrate directly, so database and app development were as natural as with fox.
I still consider the Visual FoxPro REPL and debug window THE golden standard! NO even iPython is superior. It was so out-your-way and dam practical (you can use it directly from your own apps!, talking about easy solving on your running apps)
I still end up using VFP occasionally to port our older clients to our newer software.
Python/SQLite is almost as productive, you can do some sweet stuff with a few lines of code. Not as REPL, though. Throw in some pandas, IPython and some SQLAlchemy and your are rocking.
My first DB system was dBase III+. Eventually I ended up doing some work in Clipper Summer 87, followed by Clipper 5.x with some OO even!
I remember playing around with CA's Visual Objects "evolution" of Clipper, but never did anything with it.
By then, I was more interested into Turbo Pascal and Turbo C++ for Windows 3.x.
As anecdote, maybe shared by other HN members, I remember having some problems migrating from the dBase mindset of what a DB is, into the SQL model.
One nice thing about Clipper is that it had "code blocks" that was the first time I had contact with what we today call Closures. It had a syntax close to Ruby blocks: { |X| ... }
Clipper's constrained class system (de-constrained by a third party library) is still one of my favorite dev environments. Brian Marasca's ObjectDb library was an eye-opener and a great tool--thanks mate!
CA-Visual Objects, wow, talk about a missed opportunity. I co-wrote one of the first (and few, in the end) books about it during the beta phase. As the years went on I came to understand that with Clipper in the CA stable there was little chance of any different outcome but still a huge disappointment.
Of further note was the mention of FoxBase, which became FoxPro and is what we used to develop one of the first online offender databases for our state sheriff's department. For some reason using FoxPro seemed less like "programming" than using Clipper. (too much like Visual Basic)
Thanks for the walk down memory lane!