6 ms·

Which is ironic given one of the original motivations behind Richard Stallman getting involved in open source and part of the foundational history of the whole

Which is ironic given one of the original motivations behind Richard Stallman getting involved in open source and part of the foundational history of the whole GNU FOSS organisations and ecosystem… is that he got frustrated with a printer and wanted to write his own software to run it (basically wanting to write his own firmware) for it. Given we have open source 3D printers… why are there no good open source regular printers? Does no one really care that much? Is it too much of a niche desire given the ubiquitous nature of shitty printers round the world being sold for less than they cost in a parasitical effort to trap owners into a sunk cost fallacy of paying for overpriced ink?

unwind 4y ago
The argument I've been hearing when this comes up is that the mechanical side of ordinary 2d printers is too high-precision for hobbyist work. This sounds plausible, the actual nozzle holes in an inkjet printer are around .003 inches (76 µm for the rest of us) according to Wikipedia [1]. That sounds like some pretty high-order manufacturing is needed, I'm not drilling that out in the shed. Perhaps the best bet would be to find a commercially available printer (laser or inkjet) and reverse engineer it, perhaps replacing the control board with a Raspberry Pi or something more suitable (ESP32?) to drive the existing electro-mechanical components with an open firmare. [1]: https://en.wikipedia.org/wiki/Inkjet_printing#Continuous_inkjet https://en.wikipedia.org/wiki/Inkjet_printing#Continuous_ink...
- techdragon 4y ago
  That’s the argument I usually hear as well. But a lot of the fidelity of such a high precision system is basically “wasted” reproducing the kinds of analog effects that exist at the edges of the ink meets paper part of writing/illustrating with a physical implement. Plotters with pens are still a commercial niche because the need for precise clean vector graphics to be mass produced in small batches of specific content (think custom topography maps with survey and drilling markers, or mission maps for a specific regional campaign, or other commercial line art work) It’s sort of surprising there isn’t a well established reprap of plotting given it’s simplicity. The biggest missing piece is arguably the paper feed. And as for an inkjet… why does a reliable printer need to be that good, I’d be pretty happy with a “higher dpi” pseudo dot matrix printer using a lift retract style mechanism and a hacked up refillable ink pen. Tapping out the art one dot at a time. It’s not a bad argument for a lot of consumer printing use cases, but the “they need too much precision” never felt compelling to me given the breadth of printer technology both past and present.
  - EdwardDiego 4y ago
    You describe a rather niche printer.
  - caf 4y ago
    Even 20 years ago most of that plotting work had been taken over by large-format inkjets - I used to be involved with a few print shops around that time.
  - PeterisP 4y ago
    The "high precision" is required for the very basic use case of printing a page of black on white text that does not look horrible, and being able to reliably print a 30 page document without jamming. Making the paper feed and printing head positioning hardware to do that is 90% of the way of the expense of building a decent printer.
- RobotToaster 4y ago
  The actual print head maybe, but there's no reason a commercial one couldn't be used, the same way people don't make their own CPUs.
- alpaca128 4y ago
  The most realistic approach is probably to build an open-source printer which can drive proprietary cartridges (which would be refillable because the printer won't complain). People could even design different cartridge holders for a few different models. It wouldn't be easy, but 3d printers also made immense progress despite technical challenges. I first heard about them in the reprap days and now we can buy well-engineered 3d printers with fancy features like resonance compensation ("input shaping").
  - capableweb 4y ago
    People who are hacking in the free time often go towards hacking on things that are fun and interesting. When 3D printing appeared, it was (still is) immensely fun and interesting; "printing your own physical objects?!". On the other hand you have 2D printing on paper, which is nowhere near as fun and interesting, so you have less people gravitating towards hacking on 2D printing software and hardware. Granted, a few FOSS giants are obviously maintaining the stack, which is amazing. Same with programming languages for example. You give a developer unlimited free time to learn whatever language they want, they'll probably either go for a language that was newly released and hyped (like Rust) or something obscenely old (like COBOL), probably won't spend time learning PHP or Perl which are too old to be new but too new to be old.
- onewheeltom 4y ago
  Sounds like work. This is just printing. I can always print at the FedEx down the street.
Moosdijk 4y ago
I think the tolerances are much finer in regular printers, leading to the fact that not anyone can create and tune their homebrew printer. Building and tuning a 3D printer requires much less precision, making it more suitable for open source (for everyone) usage.
michaelt 4y ago
> Given we have open source 3D printers… why are there no good open source regular printers? Most engineers can get an adequate printer - such as a small office laser printer - for 1-2 days wages. And can print things at work, their boss doesn't mind a few pages here and there. And can get photos printed at the supermarket for 10 seconds wages. So there's not much motivation to scratch an itch. Existing printing technology can also produce photo-quality prints, and is heavily cost-optimised (albeit with very expensive ink). So there's not much motivation to improve on the state of the art in regards to print quality, or upfront purchase price.
rtkwe 4y ago
Beyond the basic accuracy needs being vastly higher as others have said there's more going on in a paper printer than a 3D printer. Principally the actual paper feeding and movement is pretty complex and well beyond anything needed for a 3D printer.

unwind 4y ago

The argument I've been hearing when this comes up is that the mechanical side of ordinary 2d printers is too high-precision for hobbyist work. This sounds plausible, the actual nozzle holes in an inkjet printer are around .003 inches (76 µm for the rest of us) according to Wikipedia [1]. That sounds like some pretty high-order manufacturing is needed, I'm not drilling that out in the shed. Perhaps the best bet would be to find a commercially available printer (laser or inkjet) and reverse engineer it, perhaps replacing the control board with a Raspberry Pi or something more suitable (ESP32?) to drive the existing electro-mechanical components with an open firmare. [1]: https://en.wikipedia.org/wiki/Inkjet_printing#Continuous_inkjet https://en.wikipedia.org/wiki/Inkjet_printing#Continuous_ink...

techdragon 4y ago
That’s the argument I usually hear as well. But a lot of the fidelity of such a high precision system is basically “wasted” reproducing the kinds of analog effects that exist at the edges of the ink meets paper part of writing/illustrating with a physical implement. Plotters with pens are still a commercial niche because the need for precise clean vector graphics to be mass produced in small batches of specific content (think custom topography maps with survey and drilling markers, or mission maps for a specific regional campaign, or other commercial line art work) It’s sort of surprising there isn’t a well established reprap of plotting given it’s simplicity. The biggest missing piece is arguably the paper feed. And as for an inkjet… why does a reliable printer need to be that good, I’d be pretty happy with a “higher dpi” pseudo dot matrix printer using a lift retract style mechanism and a hacked up refillable ink pen. Tapping out the art one dot at a time. It’s not a bad argument for a lot of consumer printing use cases, but the “they need too much precision” never felt compelling to me given the breadth of printer technology both past and present.
- EdwardDiego 4y ago
  You describe a rather niche printer.
- caf 4y ago
  Even 20 years ago most of that plotting work had been taken over by large-format inkjets - I used to be involved with a few print shops around that time.
- PeterisP 4y ago
  The "high precision" is required for the very basic use case of printing a page of black on white text that does not look horrible, and being able to reliably print a 30 page document without jamming. Making the paper feed and printing head positioning hardware to do that is 90% of the way of the expense of building a decent printer.
RobotToaster 4y ago
The actual print head maybe, but there's no reason a commercial one couldn't be used, the same way people don't make their own CPUs.
alpaca128 4y ago
The most realistic approach is probably to build an open-source printer which can drive proprietary cartridges (which would be refillable because the printer won't complain). People could even design different cartridge holders for a few different models. It wouldn't be easy, but 3d printers also made immense progress despite technical challenges. I first heard about them in the reprap days and now we can buy well-engineered 3d printers with fancy features like resonance compensation ("input shaping").
- capableweb 4y ago
  People who are hacking in the free time often go towards hacking on things that are fun and interesting. When 3D printing appeared, it was (still is) immensely fun and interesting; "printing your own physical objects?!". On the other hand you have 2D printing on paper, which is nowhere near as fun and interesting, so you have less people gravitating towards hacking on 2D printing software and hardware. Granted, a few FOSS giants are obviously maintaining the stack, which is amazing. Same with programming languages for example. You give a developer unlimited free time to learn whatever language they want, they'll probably either go for a language that was newly released and hyped (like Rust) or something obscenely old (like COBOL), probably won't spend time learning PHP or Perl which are too old to be new but too new to be old.
onewheeltom 4y ago
Sounds like work. This is just printing. I can always print at the FedEx down the street.

Moosdijk 4y ago

I think the tolerances are much finer in regular printers, leading to the fact that not anyone can create and tune their homebrew printer. Building and tuning a 3D printer requires much less precision, making it more suitable for open source (for everyone) usage.

michaelt 4y ago

> Given we have open source 3D printers… why are there no good open source regular printers? Most engineers can get an adequate printer - such as a small office laser printer - for 1-2 days wages. And can print things at work, their boss doesn't mind a few pages here and there. And can get photos printed at the supermarket for 10 seconds wages. So there's not much motivation to scratch an itch. Existing printing technology can also produce photo-quality prints, and is heavily cost-optimised (albeit with very expensive ink). So there's not much motivation to improve on the state of the art in regards to print quality, or upfront purchase price.

rtkwe 4y ago

Beyond the basic accuracy needs being vastly higher as others have said there's more going on in a paper printer than a 3D printer. Principally the actual paper feeding and movement is pretty complex and well beyond anything needed for a 3D printer.