I am building Open Source machnery too; could really use that new license Prusa talks about, thoughts on how to proceed in this business
Hey everyone, I am developing an open source, 3d printable Energy Recovery Ventilator. This is a machine which provides clean air taken from outdoors to the interior of a building without the need to heat or cool the air (much). It does this by capturing heat(/cold) from the outgoing air stream and transferring it. The website is www.openerv.ca.
I am looking at Prusa as a company as inspiration, although paradoxically, I should admit up front I have never actually looked at their source code and am not actually that familiar with the whole thing (I will before I buy). Also I don't have a Prusa and am not sure I will buy one (hopefully). If I do it will be a genuine one. Same as I do with Arduino, and this is not for philosphical or ethical reasons, but because I have discovered that actually the arduinos from aliexpress are not nearly as good. This IMO is an illustration that the source code is not the whole picture for manufacturing a device, not by a long shot.
I think this discussion started by Prusa's blog post is very important. Technology is important and open source is very important, to society. It is a powerful force partly because it can add up over long time periods and harness collaboration of large numbers of people across the globe. IMO Open Source is sensibly regarded primarily as a method of collaboration to accomplish things, especially engineering and scientific things. Gravity rules the cosmos, even though it is extremely weak (about 10^-42 times the strength of the electrical force). This is largely because it always, always adds up. Some lesson.
I was reading the OSHWA website earlier today, and it has become clear that their definition is not applicable to many types of hardware. It is too narrow, by far, and this is very disappointing to me (I've known it for a long time, but just reviewing it). It makes sense for e.g. an arduino, which is not that hard to design. However as Prusa points out, a 3d printer for instance, or an ERV is not actually that easy to design. Especially not from the ground up. I see no real value that accrues from 1:1 clones produced by people working in poor conditions, with the profits all accruing to rich people. All it does is make rich people richer. And we know where these types of people tend to steer the development of the product ecology - made for the garbage walmart trash.
A lot of people are saying things are changing in the 3d printing industry, that no one in their right mind is going to buy a prusa at it's high price when you can get a fairly operational printer for way less. However there is nothing stopping Prusa from getting their stuff made for cheap in areas where labor is cheap, and there is no reason to expect the cost to performance ratio is any better with a closed source system. It's only the quantity of capital poured into the close source systems that aids the way they achieve surprisingly good *apparent*, superficial, cost to performance ratios, however the ratio after a thorough accounting is damaged greatly in many ways on purpose by them as they implement planned obsolescence or simply leverage the short sighted nature of most people to get away with cutting corners etc.
Part of the reason aliexpress arduinos are so bad is they don't expect it to matter if the thing doesn't work. They don't really have a reputation and you can't get a return or whatever. Good chance you will not remember they were good and buy more even if they are. As long as they don't get kicked off the platform that's all they care about. Drive buy, nail people once each. Make your getaway. As long as Prusa keeps up the good stuff I don't see them disappearing.
My ERV is similarly going to be open source primarily to benefit the customer directly, not some wealthy and unethical industrialist. It's extremely rare anyone would print their own ventilator, and it's also highly unlikely anyone will contribute to the design significantly. Therefore the type of licenses used for software or electronics have little applicability.
It's also important to note that it must be combined with commodity, and ideally open source, parts all the way down. I had a problem recently where I could not obtain the motor driver boards I was planning on using, because although they were open source they were quite complex and the components were not widely available. Lesson learned.
The maintenance cost of my ERV is greatly reduced because people can get replacement parts anywhere, and the lifespan is greatly increased therefore (also through design like replaceable bearings). Custom parts are all printable, the electronics are all standard type parts/can be substituted with others (power supply etc) or are open source themselves. I used a raspberry pi pico and MicroPython as well as some Arduino. The motors are nice motors and unfortunately there is limited standard interface, however I made the parts that connect to the motor and the motor driver such that you can adapt them with minimal effort should you be unable to obtain the exact same model of motor, for instance. More importantly, I got one that has replaceable bearings and put thermal overload protection features in so there is very minimal need to every replace it (only upon physical damage really). There are things you can do, I believe.
I am sure there are similar problems with e.g. the load sensor in the mk4 for bed levelling. I am not a fan of requiring licensing for replacement parts, I think as system for rubber stamping approved parts is a good idea but if customers wish to gamble with knock off parts...
I need to put an acknowledgement page up soon because I also use Cura, Marlin, SimpleFOC, and a bunch of other open source stuff. Again, this is very much not for philosophical reasons. To be honest, I use Creality Enders as my printers, yes they are kind of crummy but I can always get parts for them and they are pretty repairable. I used to be a cnc machine technician so I know my way around, but it still takes significant work and I can see why they have a bad reputation. However many of the processes are actually process challenges that are not affected by the machine design very much - temperatures, extrusion and feed rates have to be in the right range, as I'm sure you know, and while default profiles help, they cannot always win the day.
Also aftermarket parts are quite important in my experience. Better bowden tubes, stuff like that. All machines have design flaws. I would never buy a closed source printer because I know there will be problems I cannot fix and parts I would need to make with it but could not. The fan blades in my ventilator, for instance, are 25 mm long walls 0.4 mm wide (one road) and at an overhang of 65 degrees. I absolutely needed that geometry to make good fan blades. I cannot depend on any proprietary printer to be able to do that. It might, it might not. With open source, I know I can always get under the hood and get things done, at least within technological limits.
I'm also aware that buying proprietary stuff is like buying lead paint - it works great in some ways, but you are setting yourself and everyone else up for a fall in the longer term as the companies will lure us in and then up their prices for parts and materials etc. This is no philosophical matter. There were a number of companies, if you recall, that tried to put encrypted chips in their proprietary filament rolls, and make it so you could not use any other filament in their machines. That seems to have gone out of style now, but I kid you not.