Home - Programming - Delphi: Where it went so, so wrong...

Delphi is more than a development environment, it is THE development environment, and I say that as a long time high priest of Object Pascal and all things that originated at Borland International.

So yea, I'm a fanboi.

But Borland missed a huge opportunity. Now I'm not going to bore you with a discussion that goes around and around on various forums. I'll just give this one point and then wisely run away, and it's this.

Borland marketing sucked, and development got distracted.

Yeah, the product should have sold itself, and it did for a while. It was the baddest IDE on the planet and brought true Rapid Application Development for Windows to the common man. And yeah, Borland should have fought harder to keep Andreesen and the other developers that Microsoft shrewdly stole away. In fact, it should have seen it coming. Years before, Microsoft hired away the engineers responsible for the DEC VAX operating system to create Windows NT, a ground-up rewrite of the Windows core OS that made Microsoft the giant that it is even today. DEC shriveled up, got bought by Compaq, and the DEC Alpha became a footnote in CPU history.

But at the end of the day, getting Borland products in the hands of college students and partnering with more companies to secure market penetration -- two of the simplest and most obvious things any tech company can do -- were failures. I can't really blame them entirely; Microsoft did its level best to beat Borland to the punch and lock them out of every educational district and every corporate development environment, but they weren't invincible, and Borland's survival relied on defending itself, both in market penetration and in the courts, from Microsoft's attacks.

And then there's development. Marketing can be fixed, but not supporting basic features that clients are demanding for years at a time is the original sin in software development. Unicode and cross-OS compilation were low hanging fruit. Kylix was interesting, poorly executed, and somewhat useless, but it should have been open sourced and maintained for street cred and eventual adoption once Linux got greater market share. Not that Fire Monkey isn't cool, but it was SO late to the game.

But Unicode and database agnosticism should have been top priorities. It tooks years for Unicode to get implemented somewhat correctly, and then several versions more before it finally became the default just a few years ago. In the mean time, if you wanted to do globally-distributable software, you pretty much had to either struggle with the limited support in Delphi or switch to .NET.

And Borland/Inprise/Embarcadero were so married to the low-end, under-featured, under-performing Interbase that the closest you could get to support for other databases was either the anemic and irritating BDE or their ADO components. True cross-database support was non-existent until they purchased OpenDAC and renamed it FireDAC. And FireDAC is cool, but it took them until 2014 to bundle it with Delphi? Dude, I made my own database agnostic library and toolset fifteen years before that, and it worked with every major SQL database out there. It's just not that hard a nut to crack.

Delphi has been the most under-appreciated, under-developed, and poorly guarded flagship product I've seen for a major player in the development industry. At one point it took CodeGear to dig the product out of Embarcadero's trash bin and breath new life into it before Embarcadero would admit that the IDE still had a huge following and could be profitable. It's frustrating to have bet my career on it. Not that I haven't made good money, but it's a struggle nowadays to find Delphi jobs, and it's all the fault of Borland back in the day, who set the tone for the miserable stewardship practices of Inprise and Embarcadero.


Todd Grigsby