It took that long because nobody was working on it, because it wasn't obvious that a low-cost 3-D printer was feasible.
The 3-D printers you're seeing today are basically the series of RepRap designs, named after famous scientists who studied self-reproduction: Darwin, Mendel, and Huxley. The RepRap project, which started in 02005, is the reason this happened. For the first several years, it was about half a dozen people: Rhys Jones, Patrick Haufe, Ed Sells, Pejman Iravani, Vik Olliver, Chris Palmer (aka NopHead) and Adrian Bowyer. The last three of these did most of the early work. Once they got it basically working, after many years of work, a lot of other people got involved.
There were a series of developments that had to happen together to get a working low-cost printer. They had to use PLA, because the plastics conventionally used (mostly ABS) had such a high thermal coefficient of expansion that they needed a heated build chamber. They had to design their own electronics, because Arduino didn't exist. They had to figure out how to build a hotend that wouldn't break itself after a few hours. They had to write a slicer. They had to write a G-code interpreter. They weren't industrial engineers, so they didn't know about Kapton. They wasted a lot of time trying to make it work without even a heated bed, to keep costs down. They improvised leadscrews out of threaded rod and garden hose. They made rotational couplings from aquarium tubing. Lots and lots of inventions were needed to get the cost down from US$60k to US$0.3k, and lots and lots of time was wasted on figuring out how to get the resulting janky machines to be reliable enough to be usable at all.
Starting in the mid-90s, Don Lancaster was excited about 3-D printers, which he called "Santa Claus machines" https://www.tinaja.com/santa01.shtml, when he could see that they were possible. He wrote lots of technical articles about building what he called "flutterwumpers": "low cost machines that spit or chomp". https://www.tinaja.com/flut01.shtml. For example, in https://www.tinaja.com/glib/muse140.pdf in 01999, he describes Gordon Robineau's low-cost PCB drill driven by MS-DOS software over a serial port, with a schematic. (The fishing-line cable drive sounds imprecise, since this was years before Spectra braided fishing line.) But nobody listened. I don't know if he ever built so much as a sign cutting machine himself.
Journalists like to talk about the patents, maybe because they're legible to nontechnical people in a way that difficulties with your retraction settings aren't, but when I was obsessively reading the RepRap blogs in the period 02005–02010, I can't recall that they ever mentioned the patents. They were just constantly hacking on their software, fixing their machines, having then break again after a few more hours of printing, and trying new stuff all the time. I don't think the patents even existed in their countries, and they were researchers, anyway, and generally patents don't prevent research. Maybe there's a vast dark-matter bulk of for-profit hackers who would have gotten involved and started up profitable consumer 3-D printing companies before 02005 if it hadn't been for the patents, but who never got interested because of the patents.
But what I saw was that businesspeople started commercializing RepRaps once the open-source RepRap hackers got them to work somewhat reliably. Before that, they mostly weren't thinking about it. After that, most of them spent a lot of years shipping very slightly tweaked RepRap designs. Josef Prusa got involved in the RepRap project and redesigned Ed Sells's Mendel, and everybody copied him, and he famously started selling it himself, very successfully. https://reprap.org/wiki/The_incomplete_RepRap_Prusa_Mendel_b... And more recently Bambu Labs has apparently gotten the machines to be much easier to use.
> Journalists like to talk about the patents, maybe because they're legible to nontechnical people
Journalists usually repeat what said experts, unfortunately, not always include enough context to understand.
Patents context is global West, where patents are extremely powerful, so nobody could violate patent and make business on this. And as I know, in many cases people scared even to talk about something patented, and to do something scared like from hell.
On global East (example ex-USSR), patents are mostly harmless, but also don't exist all other economy infrastructure, need to make big things (I mean free market, open borders, powerful VC funds, powerful industry).
Generally, as in this case, a lot of the R&D comes from academia, and from the open source community. For-profit companies aren't very good at stuff that's far out, so they generally free-ride on open-source efforts like Linux, Python, the Web, and RepRap.
The 3-D printers you're seeing today are basically the series of RepRap designs, named after famous scientists who studied self-reproduction: Darwin, Mendel, and Huxley. The RepRap project, which started in 02005, is the reason this happened. For the first several years, it was about half a dozen people: Rhys Jones, Patrick Haufe, Ed Sells, Pejman Iravani, Vik Olliver, Chris Palmer (aka NopHead) and Adrian Bowyer. The last three of these did most of the early work. Once they got it basically working, after many years of work, a lot of other people got involved.
There were a series of developments that had to happen together to get a working low-cost printer. They had to use PLA, because the plastics conventionally used (mostly ABS) had such a high thermal coefficient of expansion that they needed a heated build chamber. They had to design their own electronics, because Arduino didn't exist. They had to figure out how to build a hotend that wouldn't break itself after a few hours. They had to write a slicer. They had to write a G-code interpreter. They weren't industrial engineers, so they didn't know about Kapton. They wasted a lot of time trying to make it work without even a heated bed, to keep costs down. They improvised leadscrews out of threaded rod and garden hose. They made rotational couplings from aquarium tubing. Lots and lots of inventions were needed to get the cost down from US$60k to US$0.3k, and lots and lots of time was wasted on figuring out how to get the resulting janky machines to be reliable enough to be usable at all.
Starting in the mid-90s, Don Lancaster was excited about 3-D printers, which he called "Santa Claus machines" https://www.tinaja.com/santa01.shtml, when he could see that they were possible. He wrote lots of technical articles about building what he called "flutterwumpers": "low cost machines that spit or chomp". https://www.tinaja.com/flut01.shtml. For example, in https://www.tinaja.com/glib/muse140.pdf in 01999, he describes Gordon Robineau's low-cost PCB drill driven by MS-DOS software over a serial port, with a schematic. (The fishing-line cable drive sounds imprecise, since this was years before Spectra braided fishing line.) But nobody listened. I don't know if he ever built so much as a sign cutting machine himself.
Journalists like to talk about the patents, maybe because they're legible to nontechnical people in a way that difficulties with your retraction settings aren't, but when I was obsessively reading the RepRap blogs in the period 02005–02010, I can't recall that they ever mentioned the patents. They were just constantly hacking on their software, fixing their machines, having then break again after a few more hours of printing, and trying new stuff all the time. I don't think the patents even existed in their countries, and they were researchers, anyway, and generally patents don't prevent research. Maybe there's a vast dark-matter bulk of for-profit hackers who would have gotten involved and started up profitable consumer 3-D printing companies before 02005 if it hadn't been for the patents, but who never got interested because of the patents.
But what I saw was that businesspeople started commercializing RepRaps once the open-source RepRap hackers got them to work somewhat reliably. Before that, they mostly weren't thinking about it. After that, most of them spent a lot of years shipping very slightly tweaked RepRap designs. Josef Prusa got involved in the RepRap project and redesigned Ed Sells's Mendel, and everybody copied him, and he famously started selling it himself, very successfully. https://reprap.org/wiki/The_incomplete_RepRap_Prusa_Mendel_b... And more recently Bambu Labs has apparently gotten the machines to be much easier to use.