PowerBuilder History, Powersoft History (2004)

Powerbuilder divided biz software developers into two camps: the smart people who hated because it was too opinionated and confined, and everyone else who understood that business results were what mattered.

Powerbuilder offered ROI like nothing else in that era if you needed custom software in your business.

It's interesting that the currently-hot idea of Entity Framework was the core advantage of Powerbuilder with the Datawindow. It still feels like the web is catching up to Powerbuilder in some ways.

PowerBuilder was so far ahead of its time and did so many things right. Microsoft pretty much ripped the datawindow concept early on in .NET's life with datasets, but didn't get it nearly as right as PB did (in the beginning anyway).

Pity they don't open source the thing and maybe some life could be breathed into it.

I spent 6 months moving 13 Microsoft Access apps to Powerbuilder equivalents. We went from daily problems to I could actually enjoy my morning cup of tea.

Oddly enough as someone who graduated with a Bachelors in 2013, I never would have thought I would have had hands on experience with PowerBuilder. But my first job out of college was a company that sold a product built in PowerBuilder and they hired me to do some work to gradually phase in newer web technologies. Said employer actually paid to have a PowerBuilder consultant/trainer come in and give me (and some others) on site training. Of course by this time SAP already owned PowerBuilder.

To be honest, I don't hate it. It wouldn't be my first choice as a tool, but it really wasn't bad. It certainly made trade offs...but I can't remember any time I thought a design decision was "terrible".

We still have some legacy PowerBuilder apps running today, and I am responsible for one. I still have to change some things in the app from time to time.

PowerBuilder is (was) an extremely powerful rad tool. The datawindow still has no rivals for crud operations on data and for reports.

The web has taken over, but the productivity this tool gives you is nowhere to be found in the modern dev stacks.

I spent 4 or 5 years doing Powerbuilder apps, both for an Enterprise employer and as a consultant. I grew to know it inside and out, which made for a valuable skill for a while. But, as the article said, the web came along, and I hardly noticed when I closed up my Powerbuilder IDE for the last time.

I worked with Powerbuilder for a year around 2012 as a junior. I have pretty fond memories of a completely broken workflow UI/UX. The project I was working on was an in-development replacement for a COTS COBOL application for use within a nieche market. I was just happy to land my first development job, but there were some alarm bells going off given the application hadn't had an initial release, some 10 years after development officially began.

The IDE view would allow you to select a function to view / modify via a dropdown list - it wouldn't jump to the section in code but be roughly similar to the usefulness of a one-function-per-file methodology might seem in Visual Studio, but without the Ctrl+Tab.

That in itself wouldn't have been completely horrible if you could actually navigate while the previous buffer had invalid code, or just had an intellisense equivalent that actually worked. Alas, if you couldn't remember some function or global variable name you'd need to comment out enough code to get the file into a buildable state in order for you to use your mouse to select the appropriate item from the function dropdown list in order to view the relevant section of code elsewhere within the same class file.

There was also inconsistent build behavior, I gather most compiled languages contain reference errors or similar when performing incremental builds but this fucker would actually fail mid build for no apparent reason - load it back up and rebuild the same stuff and it'd all work fine. It wasn't my computer, all devs had similar issues across multiple operating systems and minor revisions.

The application itself also had a crazy inheritance (think WYSIWYG objects with 8+ ancestors) and event hierarchy (pre-event, event, post-event. pre-save, save, post-save) which I think was sortof an immature OO and GUI ideals, implemented poorly. In powerbuilder terms this meant that not only did I have 400 events in my dropdown list of stuff that I can't access unless I create massive code blocks, but there was a second dropdown containing an equally crazy inheritance hierarchy with 400 functions at each level.

Oh, also don't forget that functions do not implicitly call their ancestors, but events do (or was it the other way around, I've tried my best to block this out of memory).

Needless to say, I bailed for greener pastures, I've since lost contact with everybody there but occasionally I check their website - they added some screenshots of the new application sometime in 2014 but I still can't find any indication it's actually sold to clients. I'm sure there are people out there that say Powerbuilder wasn't that bad, but personally I'd rather code in VB6.

"It was an offer like no other offers. The groom asking for Powersoft's hand in marriage was Sybase and the billion dollar dowry offer was very seductive. So a wedding/merger was arranged on February 13, 1995. I hope they took pictures during the wedding ceremony and honeymoon because the "paper valuation" (the deal was done with Sybase stock—worth $904m) didn't last long. The bad news arrived in the form of fabricated (Sybase) sales results. Sybase stock took a tumble, along with the fortune of many Powersoft executives..."

Whoa, flashbacks!

One of my first jobs as a grad was to build a telnet interface into a PB app. Fun times.

Too bad it's basically dead at this point.

I hate powerbuilder with a passion.