In all of the arguments of late about Open Source and 3D printers and whatever Bambu Labs or Prusa might be doing with 3D printers and whatnot, I guess what I’d really like is for our 3D printers to not end up like 2D printers. Now, when I say 2D printer, I mean some things, none of them very positive.
Obviously, a 3D printer exists at the interface between the mechanical and virtual where you cross over from digital certainty to mechanical things that need to be lubricated and have backlash and, as such, it’s never going to really be a fully solid-state digital device.
Nonetheless, a Prusa-styled bedslinger is, if you ignore a lot of details, little more than the same sort of machinery as an ink jet printer but with one extra dimension. Obviously, this is a dangerous simplification of things because each side of the fence has some very unique requirements, etc.
So, let’s talk through what I mean by “Like 2D printers” and where we’ve gone so far with 3D printers and then I’ll talk through some concepts that maybe we need to think about, as a community.
What do I mean by “like 2D printers”?
I had to get a new 2D printer to print a few things after letting the broken printer sit there for as long as I dared. And the old printer died because it ran out of ink and somehow it decided that the new cartridge was empty and I looked at the cost of cartridges and decided that I didn’t want to risk the cost of another ink cartridge to see if I’d just gotten a bad legit-made-by-the-manufacturer cartridge or if the DRM had gone all fizzy or if there was some other weird failure mode that I’d have zero chance to debug or understand. Ergo, a bunch of potentially repairable plastic and electronic and mechanical bits became eWaste.
I haven’t been impressed by a 2D printer in… actually quite a while. In the early 80s, the best you could get was a dot matrix printer and everybody knew whatever you printed was off of a computer. In the mid-to-late 90s, you could easily pick up a decent laser or ink jet printer that could do something at least vaguely equivalent to a printing press in terms of text. Around 2000 there started to be 6-color CMmYyK photo ink jet printers that could actually produce a photo that looked… right. And then around 2005 or so, you had pigment-based inks such that you could print probably fully archival stuff if you treated everything right.
And, since then, I dono, there’s no quality improvements left? People don’t examine prints with a microscope to see detail that their eyeballs won’t catch in a print.
Conversely, the need for 2D printers has actually decreased. A coworker was very rationally complaining about having to buy a printer because he didn’t own one, didn’t need one, and he needed to print a legal document out and sign it and have someone else sign it and then mail it and he didn’t trust that if he took it to some public printing place that it wasn’t going to snork his document. This was right around the same time I’d read an article that suggested that, yeah, the privacy protections for those sorts of places really do suck so he’s being an entirely rational individual.
If you think about it, most folks have been able to get away without one these days because a regular old high-quality screen, especially the latest high DPI high dynamic range ones, is great. And we’ve had enough computer-using time to get comfortable with writing and editing and doing everything straight on the computer, although some of that is also figuring out what people really wanted out of their software. Plus, there are eSignature systems everywhere.
As such, the modern version of Late Stage vaguely-capitalism (where I’m casually lumping the Chinese command-economy communist system, the European socialist-capitalist system, and the American system all in there as “vaguely-capitalism”) encourages companies to be fully evil and to manufacture crappy printers that wear out quickly, require custom parts that you can’t just buy, and ensure that any competition is either dead or just as evil.
And one of the things that at least some chunk of the population is starting to wrap their head around is that this bizarre dream of recycling we’ve been sold was actually mostly lies. A lot of the energy used to create the printer is on things like the mechanical axes and motors and whatnot and unless you are scavenging the complete parts, the only thing that making sure it hits the eWaste bin is going to accomplish is that the landfill will be slightly less toxic.
I guess the important thing to highlight is that this is the thin end of a wedge that has been applied to folks in recent memory. There were durable printers, then there were cheaper-and-fancier printers where you paid out the nose for ink, and now you’ve got printers that are delivered broken.
It also feels, with the latest crop of motion-compensated super-fast CoreXY printers, we might be actually reaching a bit of a boring stagnation point for 3D printers that mirrors 2D printers when they got good at printing photos.
3D printers and promoting the progress of science and useful arts
3D printers came to popular usage not because there was an amazing advance in technology that happened to hit right then. 3D printers, as we know them today, exist because the patents on them expired.
This is a very US-centric view, but in the US, the constitution states “The Congress shall have power … To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries” which then got turned by congress into the modern US patent and copyright regime.
I reject wholeheartedly the idea that a small group of slave-owning privileged white male people some centuries ago had a monopoly on clear thinking, but there’s a few important things to consider with this. Primarily the task of promoting the progress of science and arts.
An invention that is never disclosed is protected, however it’s easy to lose. Consider the Antikythera Mechanism. It dates way back in history and someone figured out how to make it… but the knowledge with how to make it was lost. Then many centuries later, other people figured out how to do make similar mechanical computers. Ergo, you want people to disclose inventions.
However, and maybe our present-day view of Late Stage vaguely-capitalism weigh into this, there’s some idea that you shouldn’t invent a single thing and never see any “benefits” from it.
Hence, the patent system exists as a compromise between the two interests. You should publish your invention but also be incentivized to invent.
Obviously this system doesn’t really work out very well in practice in a lot of ways. I know some people who have a series of interesting inventions that were patented and everything… and they never saw a dime from them, because we’ve built a whole complex structure of contracts and case law and unfortunately you can end up signing the wrong contract and trusting the wrong people.
Likewise, there are a lot of patents where I look at the claims and feel like what the patent is about is obvious… except I’m not sure if it’s obvious obvious or only obvious in retrospect. How novel is squirting plastic filament or a purge tower? How novel was it at the time?
Late Stage vaguely-capitalism is a sort of thing that we’ve created as a structure. It is as arbitrary as how we somehow decided that assigned-male-at-birth-babies got wrapped in blue blankets and assigned-female-at-birth-babies got wrapped in pink blankets and, as it turns out, that color choice was a recent decision and very little about it maps to reality. Or gender, for that matter. Patents are not based on deep brain structures provided by evolution, therefore we should feel free to change them to better service our society as we need.
Regardless, 3D printers could have hit the market a decade earlier, except for patents. You could control a 3D printer with an Apple II connected up to some standard 74xx logic. Nothing in a printer is fundamentally novel in terms of fabrication technologies.
The actual modern 3D printer as an object that you could acquire and use for things came about because of the RepRap folks who had extremely world-changing altruistic motives. When Sanjay Mortimer said ”We build space shuttles with gardening tools so anyone can have a space shuttle of their own” the community was literally meaning space shuttle sorts of objects based on the ”Clanking replicator” concept of space exploration. Hence, it was important for this to be an open source thing and, not only that, but the idea was that you could build yourself a starter printer from easily fabbed material (A RepStrap) to print your printer or you would have someone else print you a set of parts to print your printer.
Over time, companies emerged. This guy, Josef Prusa, took the second RepRap design (The Mendel) and started to simplify it, calling it the Prusa Mendel, which eventually became the Prusa i3 open source printer. This other guy, the Sanjay Mortimer from the quote, figured out how to make a hot end that could actually deliver decent print quality, creating the V6 hot end. Josef Prusa formed Prusa Research and Sanjay Mortimer formed E3D.
The thing is, for a while, there were a set of actual open-source and open-architecture printers. Making a RepStrap was a bit of an ordeal and not actually very fun for most folks. But if you got a decent variety of printer models, you could actually get an open source printer where you could download the files, print replacement parts, make your own modifications, and so on. Furthermore, there was an even wider market for open-architecture printers such that if someone came out with an amazing new hot-end or nozzle or controller, you could swap it out.
However, if you look at the recent movements in the printer market, it’s not looking so good.
The Prusa i3 Mk 3 was an open-source printer. The Prusa i3 Mk 4 is, at present, not an open source printer. While Prusa marketed it as an Open Source printer, the facts of the matter is that at the publishing of this blog article, you cannot download the design files like you could with the Mk 3. The CORE One doesn’t even claim to be open source anymore.
Likewise, neither the Prusa i3 Mk 4 nor the CORE One are an open architecture. The Nextruder nozzles are proprietary.
Bambu Labs makes no pretense about being open-source, although the slic3r license is such that they had no choice but to keep their fork open. And, in the same bad way that the new Nextruder requires special nozzles, so does the Bambu Labs. Furthermore, there’s a whole giant cloud offering for remote control where everything flows through the Bambu stack… and they just decided that they were going to force you to always use the Bambu stack to control your printer.
Even E3D, one of the first companies to make a hot-end and nozzle combo that didn’t suck, is moving people to their patented Revo nozzle and system.
The homesteading open source 3D printering might have been a feverish dream, I dono.
If you notice, I’m differentiating between open-architecture (meaning, you can swap out bits and repair it yourself, more or less) and open-source (a term centering around freedoms of use).
There’s a whole “homestead” thing in the US that was apparently always a real-estate scam. The idea was that you’d get a block of land that was “yours” and you could move out and start a new life with your family, away from everything. Naturally it wasn’t actually “yours”, it was taken from the indigenous people who lived there before. And everything about what builds modern culture is built around sharing and collaborating, so living on your own is putting yourself at a severe disadvantage so if you look at the simpler times that occurred more naturally in Europe, you see a much more sustainable social structure that is not building a little house on the prairie but instead communities edging up against farmland. And I forget most of the details, but the people who actually benefited from the whole “homestead” thing were other wealthy people, not the suckers who signed up for backbreaking labor being “pioneers”.
RepRap printers as a prototype for the sort of “Clanking replicator” one could land on the moon or mars to autonomously manufacture a colony is an interesting idea… but notice that the word “Colony” that I deliberately picked to make a point here isn’t a very nice word in a bunch of circles either.
Thus, while we’re off subject, I want to highlight that there’s all of these big fancy dreams of open source 3D printers and micro-factories and colonizing places… and when reality hits, for the people involved it feels like they have a choice about staying a dreamer or chasing reality with less of the dream. And then there’s, of course, sketchy people who are out to make a buck on people’s dreams.
I really like a lot of things about open source but I also have seen the unintended side effects involved and I’m completely prepared for us to all realize that it was always a mirage as well, just a different mirage.
What’s important is that, even if the fever dream of clanking replicators building a space colony for pioneers to inhabit isn’t actually going to happen, giving us a useful new tool for stuff that’s neat but maybe a little more mundane is a success, whereas all of us getting screwed over like homesteaders is failure.
Likewise, it’s not one thing that is going to keep 3D printers useful user-friendly devices. Open architecture and open source are two nice things, none of them solves all problems.
Case in point: The Sovol open source SV08, based on the Voron 2.4
The Sovol SV08 printer is an interesting project, because Sovol took the open source Voron 2.4 design, made a bunch of changes, and are selling it as a new open source printer.
Looking through Angus from Maker’s Muse’s video, you can see why they had to make some of these changes — he’s able to slide it all together quickly and otherwise get it going without nearly the same degree of drama required to make a genuine open source 2.4… as well as being able to manufacture the printer efficiently. And in this case, because it’s very obviously based on the 2.4, they released the design files.
It’s an interesting move because there’s arguably more value to be derived from a designed-for-mass-production CAD model of something like the SV08 than there is for a designed-for-home-assembly Voron 2.4 and while there’s a lot of generally sinophobic bellyaching about Bambu Labs allegedly starting with a Voron printer and then replacing all of the bits… I have to say that the SV08 is almost as different from a Voron as a Bambu Labs printer. To me, it feels like this is more of a branding exercise because if Sovol had released the SV08 as a new printer it wouldn’t be nearly as noteworthy than releasing it as the SV08, based on a Voron 2.4.
However, in the last few minutes of the video, Angus suggests that you look at all of the consumables, such as the integrated nozzle-heatbreak that’s very much specific to the SV08, and maybe think about buying a few of them. Because, guess what? The SV08 uses a nozzle-heatbreak that is specific to the SV08, so it’s not in common with any other printers, not even other Sovol printers. Thus, in spite of being open source and based on the Voron 2.4, this printer is only usable so long as the nozzle-heatbreak is available.
Furthermore, you cannot expect any of the mods designed for a 2.4 to fit and your SV08 won’t get a Voron serial number.
Likewise, I’m focusing on the nozzle-heatbreak that Angus mentioned in his video, but I do wonder if there are other specific-to-the-SV08 parts in the design.
So, even worse than the fundamental economic model behind open source being somewhat questionable in a Late Stage vaguely-capitalist system, here’s a very nicely packaged open source printer that still fails to move the needle in the direction I think we need to move it, where you are empowered to customize and repair your technology in such a way that it doesn’t turn into eWaste.
One important point that people forget, when talking about open source, is that open source is almost meaningless in a vacuum. The real value is often times the community. If the SV08 sold a hundred printers, is support ends, it’s effectively useless for the the community of the hundred people who own that printer because actually designing replacement parts for discontinued bits or adapting open source community projects … perhaps mounting a Xol or StealthBurner or DragonBurner toolhead where the stock toolhead was … requires a lot of time and effort and skill and energy.
However, it does seem like a bunch of people did purchase a SV08 and thus there are quite a few printers going around, which means that the printer is more likely to attract useful mods because it has a community.
But just as Josef Prusa opines about how the current crop of open source licenses are too much of a burden for a company like Prusa, I think that open source licensing is not enough of a means to ensure that I am empowered to maintain and modify my 3D printer for the indefinite future even if the companies who made the parts go away.
There’s another thin wedge aimed at us and we should be fighting it
3D printers are a pretty darn handy tool for a bunch of things, and it’s important that we preserve our own access and understanding of these tools.
So far, printers with DRM’d spools have failed in the marketplace. A friend had a 3D printer that was supposed to give the simplest possible experience there and it didn’t work out well and the company ended up folding and so she had a broken printer that she never got around to hacking so it can work as just a plain old printer, so it became eWaste instead.
I guess you could say it’s a victory for now, one only akin to the victory that printers don’t require special paper. Then again, Bambu has a RFID tag on the filament and a reader on the printer. They spend an unusual amount of effort making sure that the printer is required to connect to their upstream services. And they also prevent you from installing your own firmware. At any point, they could decide to do all sorts of things that are profitable to them and unfriendly to you like force you to only use Bambu spools of filament.
Thus, we return to my statement. Can we have 3D printers that are not like 2D printers? There seems to be a scale.
The “Can I prevent my printer from becoming eWaste” score
An open source printer using parts that are themselves open-architecture gets an A. If it’s got a lively community that’s likely to work around shortages, discontinuations, patent screwage, et al, it gets a 100% A+ score. A lot of Vorons are using the vaguely proprietary Bambu hotends or the very much patent-restricted Revo hotends, but it’s designed that you can use a different hot-end.
A commercial Stratasys printer that has DRM’d spools, uses parts very specific to Stratasys that nobody else knows how to make, that requires so much effort to get working if Stratasys decides to stop supporting it, gets a 0% F score.
A closed-source Prusa i3 Mk 4 or CORE One gets points deducted for lacking design files, the nextruder, the specialized controllers, but still comes gets some points, maybe B or C range because the design is otherwise well understood. If Prusa disappears the printer won’t be eWaste.
The SV08 might be a little bit higher, on account of being actually open source?
A printer released with a full suite of design docs, spares, etc. using open-architecture parts that is not open source could get an A.
A Bambu labs printer gets a lower score because, while it’s just as restricted as the CORE One, it contains features that are there to actively work against you, so it’s in the C to D range. For example, adding the RFID tag to automate filament settings has a few small advantages for the end user but mostly enables spool-locking. Combining that with locked firmware means that you can’t undo the locking if they decided to deploy it.
Most, but not all of my vibes-based scoring, could probably get turned into some sort of scored checklist, which I’m going to leave as an exercise for the reader. For example, it’s entirely reasonable and something I’ve seen in practice to mark down any company that has ever removed functionality remotely.
You try convincing a Bambu buyer to get a Voron instead, I dare you
The most popular license for open source was created because of 2D printers. This guy was annoyed because his 2D printer had driver issues and that led him to start a foundation and do a bunch of stuff. On the other hand, he’s also a deeply unpleasant person at a time where a variety of alternatives to fully-locked commercial software were being experimented with and he happened to be one of the loudest voices and therefore we’ve got this whole community built around this idea. The problem is that, because he’s a deeply unpleasant person who enabled other unpleasant people, the collective group of people unable to empathize with the needs of normal human beings and how they would approach computers has arguably held us back.
There’s a sort of person who builds a Voron and tunes it and upgrades it and whatnot to make their Voron the best possible 3D printer ever, much in the way that one would tune a sports car. I am not that person. I, like a lot of people, mostly want to 3D print neat things for my projects, where a good fast printer is better than a slow cruddy one.
Part of how a Bambu Labs printer just works is because they have a bunch of carefully specified parts built for mass production such that it arrives without requiring you to spend a bunch of time fiddling with things. Replacing extrusions with a more solid metal frame and designing parts that cannot be 3D printed but are instead milled from metal helps as well.
Some things happened that caused me to be extra-suspicious of devices that exist within the vague form factor of a Bambu Labs device, where it’s presumably running some Linux build, phoning-home, etc. that aren’t even 3D printer specific. Therefore I decided that a Voron was a great way for me, with my skill set, to get a faster printer. Which is not for everybody. The solution to this is not that everybody builds a Voron.
The most recent Bambu Labs thing..
This has been sitting around as a draft for a while. Every time Prusa or Bambu did something silly, I’d update it a bit more. The most recent Bambu labs change, where they suddenly decided to remove public APIs so you were forced to only use Bambu’s tooling, is a decent moment to pause and consider where we’re going. A lot of people got very excited and angry very quickly over this, for good reason.
Bambu removing public APIs from the device and denying you functionality you had at the point where you purchased the device is the thin end of the wedge. We’ve had enough history with the 2D printers to know where this goes.
It’s important to understand that Bambu is spending a chunk of money each month to host, support, secure, and upgrade their Bambu cloud services. This model is built into the design of their printers and was a deliberate choice on their part. As you are not paying a monthly fee for the use of these services, they are faced with any number of options for extracting the money to keep those services running out of you. Depending on their own internal cost models that you have no visibility to, there’s a fairly high probability that it will make logical sense for Bambu to add a long-term monthly subscription fee, brick your printer, remove key functionality, lock your printer to Bambu spools, or other such measures regardless of what they might say now.
It’s about community standards
People supporting the lage-stage vaguely-capitalist system defend advertising as being there to “inform” people, suggesting this idealized world where a person sees a video on youtube where a creator uses the printer that Bambu sent them, is enthralled, and suddenly realizes that buying a 3D printer, maybe the Bambu printer, would be life-changing.
I’m going to be extremely rude here and parallel the Bambu situation to the recent revelation that Paypal’s Honey screws both content creators and consumers out of their money. I always figured the Honey app sounded fishy and I’m not alone. However, a wide variety of YouTubers were recommending Honey to their audience for a long time, in a completely legal and aboveboard fashion as a paid spot.
Once it became fishy, a lot of people started to take a step back. On the other hand, it looked pretty darn fishy from the start… so a lot of youtubers made a deliberate choice to have not gotten involved in the first place, although I can’t actually critique somebody who is just trying to get by keeping their channel going in a world where every year YouTube cuts their payouts a little bit more.
Bambu sent a lot of people a lot of printers, in a completely aboveboard fashion. The printer, in version one form, is actually pretty great, so between just having enough money to send out a ton of reviewer printers and letting people say whatever they wanted, was good enough to give them a lot of press. Giving people who don’t review printers a free Bambu is the same thing.
Meanwhile, the Anet A8 very justifiably earned a reputation as a firestarter. When Bambu realized that one of their printer had the same sort of Anet A8-esque firestarter properties, they spent a lot of energy making sure to smooth things over with people.
Presumably part of the trick to sending out printers to “inform” people is to get them to people who are the least likely to ask deep questions that would make Bambu look bad.
Thus, the most important thing to come out of this is not for people to get lectured on how they should own a Voron, it’s for printer reviews to talk about the long-term maintainability of a printer, how easy the consumables (which includes nozzles and hotends and moving parts) are to find and replace, and point to the mere presence of specific features by Bambu and others as being there to eventually support consumer un-friendly features.
Maybe we need to make some scoring for how consumer-unfriendly a 3D printer is. Maybe “Oh, yeah, Bambu printers are nice out of the box but it really looks like they are going to end up like your ink jet printer,” if repeated enough, will make both manufacturers and buyers act a little more carefully.