View Single Post
Old 7th March 2019
Originally Posted by SluttyMcSlut View Post
I think that infamous "used to work at Cakewalk" thread on Reddit told the sad tale of Cakewalk at that time. The anonymous OP suggested that there was a dichotomy in direction and leadership. I got the feeling that it is probably the usual stuffed shirt management style where they have no idea what it is like to work on the tools. They just want sales and wind up setting new feature deadlines that were so tight that many other things had to fall by the wayside.

Originally I too thought the same thing as you about Noel still being involved with the program. So far though he seems to be captaining it along nicely. I remember after the Gibson collapse he stated that although he was in the chief seat, he was over ridden on some decisions.

I think he was being a mixture of both nice (professional?) and protecting his ego there. As I think it is likely from seeing how Cake was run back then, that the supposed over riding on "some decisions," were actually "most decisions." Either that, or he was given leeway. But that leeway was an incredibly small space in which to attempt to manoeuvre his plans, wedged up against the crushing wall of producing and prioritising new features.

If that wasn't the case, it certainly is weird that him plus a really small team can all of a sudden bang out some serious stability fixes and address long awaited design flaws. I don't want to curse it by speaking to soon, but it certainly appears that they have righted the ship.
With software, it can be much easier to manage a small team than a large one.

Back in the late 70s, the first 'real' desktop database was written as a personal project by a guy (Wayne Ratliff) who had worked at JPL (the legendary Jet Propulsion Laboratories in Pasadena, CA, USA). He wrote it for PTDOS for Intel 8080 processors to keep track of football pools and called it Vulcan. He found a small market for it but then took on a partner with business and marketing experience; they renamed it dBase and released it for CPM and then PC-DOS. Their first release dBase II was quite basic. They added several more programmers bringing the core dev team up to, as I recall it, 5 members who created what became dBase III. I think they added a bit more programming staff for the fix-it release dBase III+

And that product sold like hotcakes. Next to Lotus 1,2,3, it was probably one of the biggest hits on the desktop in the mid 1980s.

And THEN they added a bunch more staff for a massive rewrite for dBase IV.

The programming staff kept expanding -- at the end, I think, they had hundreds of coders. And deadlines kept getting missed. Meanwhile, dBase III+ had created its own ecosystem. It had opened up the world of relational databases to desktop developers -- and brought custom turnkey application development to small businesses and consultancies -- and created third party add-on markets dependent on it.

When dBase IV finally came out, well past a succession of projected release dates, it was woefully, incredibly buggy. It had, by one report, over 1000 potentially data-destructive bugs in the first release.

The company issued fix versions but never recovered -- though the basic dBase III data format still survives as a niche tool.