Sunday, April 22, 2012

The True Cost of Acquisition

I used to be a database administrator, and a fairly good one if I do say so myself1.

For years I did the job the hard way, then a short foray into the heady bleeding edge of the dawning client-server world in a small but growing company in Greenbelt Maryland (which is in Washington DC in the same way as Kingston upon Thames is in London) introduced me to a product called ERWin Desktop Edition2. It was a wonderful little tool that allowed me to array a database design as a diagram, and as I was doing it the software was sorting out the foreign keys 3 and letting me concentrate on the design itself.

It was an amazing piece of kit. I could (if I wanted) bind the various data fields in my database to Visual Basic controls so designers would be locked into my vision of how to manipulate the fields and not get creative, as they are wont to do. It came in two flavours: Visual Basic and PowerBuilder, the two most widely-used developer languages for the newly-emerging GUI-centric business world of Windows 95 and Windows NT4. Both flavours were made by Logicworks and I praised their product loud and widely because it was An Goode Thinge.

And it was cheap to own at $250.

It couldn't last of course. Computer Associates bought the company out and first order of business was to pull the Desktop Edition version of the software off the shelves leaving only the much higher cost Enterprise Edition. In time they reissued the product with many new features, but they never relaunched the Desktop Edition, and the entry price was around $4000 for the versions they did sell.

I recently had cause to need the product (my copy of the Desktop Edition will not run on Windows 7, sadly, coming as it does from the Windows 95 era) so I went to CA's website to scope out the latest situation vis-a-vis ERWin. There was a free Community Edition! Kudos to CA! I downloaded it and it was useful though it never delivered what it said it could on the box when it came to translating the designs into an actual database - the real point of such software is database administration on a day to day basis, not just the initial design, and to do that the tool needs to actually build the databases and read the design back. It was, however, only licensed for three months after which I would need to download it all over again.

And so it came to pass.

As I was looking at the site I noticed that there was a "Developer Edition" available. Now that sounded promising! I was a developer, sort of. Surely it would be a reasonably priced product - not $250 but perhaps not more than $450 and that might be thinkable.

Four thousand, seven hundred and some piffling number tens in small change. So, we must assume some pretty affluent developers associate in the Computer Associates, er, association. I find it difficult to see how a burgeoning market in pirated software isn't simply falling apart with these bargain basement prices making cracking the software not worth the effort.

And he best part is that experience shows that in about five years the software won't work with the operating system or the database software any more, which is where I came in.

  1. And I do
  2. It came in a variety of larger scale editions too
  3. Which in this case are what the "links" between tables consist of which are the nuts and bolts of any database based on tables and a big part of the reason for going to the trouble of making a database in the first place
  4. I'll not get involved in the continuing argument as to whether anyone should have gotten involved with these languages in the first place, the fact was they allowed rapid development of good-looking apps and in the hands of those who knew what they were doing both VB and PB were awesome


Dunx said...

Two points of perspective which might amuse you.

The first is that when I was at Oracle I was working on Designer/2000, back in the days when people put "2000" on products to make them seem futuristic. My job was to reverse engineer database structures, and ERWin was one of the tools we used as a benchmark - not to clone its features, but to make sure that the reverse engineering behaviour was at least equivalent.

But the second is that I've seen the same kind of price inflation applied to a product to its detriment.

Designer/2000 was (and still is, I believe) an enterprise product - it's built for teams of developers to share their system structures in a common repository - but someone had the nice idea of taking the reverse engineering bits and the diagramming bits and making a lightweight single-user tool.

"Let's make it free," someone said. "Then we don't have to waste money supporting it."

"Oh, hang on - someone might be willing to pay money for this," someone else said (looking no doubt at ERWin). "Let's charge $300 for it."

But the final price was $1,000. It didn't sell very well.

So if you ever happen to come across a copy of the demo CD for Oracle Database Designer, cherish it because it will never be reissued ever again.

Steve said...