 Liebe Teilnehmer am RC3, ich darf euch jetzt den Festi vorstellen. Er hat in seinem Projekt Boxes PY, was entwickelt, was ziemlich cool ist, um verschiedene Boxen und Kisten zu generieren, die ich danach einfach mit einem Laser Cutter ausdrucken kann und zwar so, dass ich die ausgeschnittenen Stücke dann einfach nur noch zusammensetzen muss. Aber für die Details übergebe ich jetzt einfach mal an Festi. Freue ich euch auf Boxes PY. Hallo, so, let's talk about, hello, do I have sound? Okay, let's assume I have sound. Let's talk about laser cutters and what to do with them, because let's start with the most important thing, laser cutters are amazing. From all the, from all the rapid prototyping stuff, they are in my mind the most worthwhile machines for hackerspaces or makerspaces. For me, the main selling point is their speed. They can do stuff quick. So if you have something that's like box size thing, you can do that within minutes, like 10 minutes or something, where 3D printing always takes hours or days depending on how big the thing is. The second thing is, laser cutters are very, very precise. We will see how precise they are, but they can easily produce precision within a couple of hundreds of millimeters, which is something 3D printers can do and well, billing machines can do, but they have a lot of other difficulties to achieve that. And they're inherently 2D, which means they're easier to understand and easier to use. Basically, all you need is a 2D drawing and a laser cutter can use that. While many true 3D machines need complicated toolpaths and you need a toolchain that actually is able to deal with that. And so in hourly hackerspaces, while the introduction to laser cutter is about one hour long, but after that you can just use the machine easily. So, how do they work? Basically, they work great. If you have a bit of a closer look, it's all smoke and mirrors. It's so much smoke and mirrors, that you need some fume extraction. And that's a bit of a problem, but it's doable. If you look even closer, you see a lot of scary things that are in this huge box. And there's a nice mix of fragile glassware, water and high wall ditches. So, when you take a look at clothes, it's even more scary than just lasers. And of course, those relatively cheap and affordable laser cutters are using CO2, lasers which are invisible. So, even the danger that is from the laser is invisible. And the laser is strong enough to make stuff disappear. So, laser cutters look very, very scary. But in reality, they're not as bad as they look or sound like. Because the whole laser stuff is all enclosed in this box, which is made of steel. And if you get a good laser cutter, which is class one, it's basically the same glass as a CD-ROM drive, which basically says there's a laser, but it's inside and can't get out. This thing we need to be careful is if you do stuff you shouldn't or do repairs. So, then be careful. Otherwise, it's actually a pretty nice machine, all the difficult and dangerous parts are on the inside. And if you don't do anything stupid, there's very little chance of you getting hurt. So, I'm much more worried about other tools in our Hacker space, like circular saws or hammers or just normal electricity out of the outlet that people plug into some weird apparatuses. The only thing that actually requires care is that the laser cutter also always tries to burn your house down. The laser basically evaporates the material, it cuts and basically creates a flame, a small fire instantly all the time. And just a small air pump that basically blows out this fire all the time. So, the only thing you really, really, really need to take care of is watching your laser cutter all the time. You really can't leave it unattended. But if you do that, you should be all fine, it's all good. So, what can laser cutter do? It can cut stuff, we will talk about what exactly in a minute. And as an alternative, if you die down the power and beat it up, you can also score the surface of the material. So, you can scribble on that or put markings on that. It's also very useful. On the Internet, I've seen people are using this to bend acrylic with heat, but I've not done this myself. The trick here is that the laser is focused in one point. And if you move your material out of this point, the focal gets bigger again and so you can spread the heat on a bigger surface. And it can be used to heat stuff, but it's kind of hacky. So, only hackers might want to use that. Then there's a whole world of stuff about engraving. It basically scans lines over the surface and you can then score or fill whole areas for adding text or something like this. You can even do dittering, so there's some software chain involved to prepare the images so this looks good. But it can be done. There are also a couple of cool specialty stuff that can be used. One of them is dual-layer acrylic, which is basically acrylic with a very thin layer and a second color on top, which can be etched off. So, you have a two-color sign, for example, as a result. And we also have some special laser rubber that allows to do stamps, which is always a nice thing to have. When it comes to materials, the main thing we are using is plywood, typically birch or poplar, which is relatively light, which helps cutting, because the laser cutter needs to evaporate and burn the material so the more dense and more thick the material is the slower the cutting goes. For typical two or three or four millimeter plywood, you can cut with like 30 millimeters per second, so it's quite quick. You have to be careful, there's one kind of dark glue that's used for birch plywood, sometimes that can't be cut. It's just a bit of a material that won't evaporate and won't burn. So, no matter how powerful your laser is, it won't go through there. I had to learn that the hard way. Another thing that's not that common, but really nice is leather. Leather is very easy to cut. And you can cut in holes to sew them into complex shapes. Make sure it's actual leather, or if it's some sort of artificial material, you have to check that same caveats as the four other plastics apply here. The acrylic is the plastic to cut with a laser cutter, because it cuts really, really nicely. It can also be cut deeper than all other materials, because when cutting, the surfaces are reflective, so they refocus the beam into the cutting area. So you can cut even deeper than your beam would normally be able to reach. There are a couple of other plastics that can be cut. You need to be really careful so there are no hyalogens, especially chlorine in there. So PVC is something you really shouldn't cut, because it creates an acid that will damage the lungs, and even worse, it damages the machine, so you really don't want that. So there's a flame test you can do to transfer chlorine. And if the flame gets green, you really shouldn't use the materials. There are a couple of materials that can be cut, but most of them don't cut as nicely as acrylic. So they have the tendency to melt, and then the molten material is blown over the edge, and you have small droplets in the bottom and stuff like this. The thinner the material is, the less this matters. If you have thick material, there's a lot of molten stuff coming out of the cut, and so your mileage might vary. It's possible to engrave glass. The glass went hit with the laser cutter, cracks on the surface, so it gets mated. So you can even make a small impression with enough power. So you can put, for example, letters or some image on there. The same is true for metals, but it's very inconsistent. It depends a lot on the chemistry of the metal itself. We had a couple of really good results with stainless steels, but some other steels just won't take a color at all. There's also some spray-on color that you can burn onto metal, but that's pretty expensive. So you have to try what works or what doesn't. One can try to put oil on the surface and burn this into the metal to get some blackening effect similar to normal blackening of metal, but it depends a lot on the type of metal you're actually dealing with, so it's worth trying. So this is a talk on its own, but I'm just adding it here quickly. So if you don't have a laser cutter and think they're too expensive, machines like ours, if they're upgraded to European standards, so they're really true Class 1 lasers and are safe. They're pretty expensive. They can be around about 10,000 euros. So if you surprisingly don't have that money and don't want to put some CCC funding into that, what we did is we took a credit from our members. We charged per minute. We have lowered the price since now, but at the beginning we charged 50 cents per minute. The machine actually ran. Whenever someone needs someone else to laser for them because they can't operate the machine or they're an external customer, we paid the operator another minute. So if they work for one minute, they have one for their own. And of course, that needs to be paid by the person that wants to have their stuff cut. And we charged another 50 cents just for good measures for external people who don't contribute to the rent of the room and the electricity bill and anything else. So that's the way to refinance your laser cutter. The question is, how do you motivate your people to actually put that much money on the line? And what we came up with is just paying interest in laser minutes. This is a great benefit that it doesn't cost the hackerspace any money and laser minutes are easy to produce and still with 50 cents per minute it's easy to come up with an interest rate that's very, very competitive compared to banks. And we allowed paying back the loans within 10 years on the terms of the hackerspace which worked very, very great. After three years we were able to pay everything back and so we were able to lower our rates after that. So, if you need help with that, contact me after that. So, back to actual laser cutting. The difficult thing with laser cutting is that there are 2D and the question is, why the Congress is in 2D, most of us live in 3D and so you want something that's not flat. And there are basically two ways of doing that. Let's say two and a half. The easiest one is just cutting multiple layers and stacking them on top of each other. That's something, I'm not doing a lot, but it's something that's very interesting if you want to do something with circuit boards, for example, you often can just have one plate on top, one plate on the bottom and just have spacers in between that holds your circuit board. And it's good enough for many applications. The trick here to use is to make holes for alignment pins. You can just use some nails in there or you can laser cut a piece of wood that you slot in or you can have some kind of sea clamps that sandwich the stack of layers. And the trick is that these help while gluing the stuff together. Otherwise, it's very difficult to align stuff because if you don't have enough glue, it's not holding very well. If you have too much, things start sliding around. So alignment pins is a trick and it's easy to add. So many things here, you can just use Inkscape, click to take together a basic shape, copy them a couple of times and then cut away what space you need inside. The other obvious solution is to just use finger joints and some box-like structures. That's what we are talking about in this talk mostly. So what's the problem? The problem is, I had a laser cutter, we were buying one in my old Hacker space but I was looking for a box generator. And it was 2013, so a lot of software projects that look better now, like FreeCut, weren't still that great. I was looking for an open-source solution, something that would be fun to code, something that would be able to deal with different material thicknesses, that would do curve corrections or would account for the amount of material that's removed by the laser. So I was looking for an open-source solution by the laser. I urgently wanted to do flex cuts just to see how they looked like. They were basically making wood bendy by cutting it a lot. Of course, I wanted finger joints. And it turned out, there wasn't a great solution there. So I started Box.py. I implanted in Python. It basically is a huge turtle graphics project. So it's basically just you start in one corner, go along the edge, turn 90 degrees, do the next edge, and repeat until you're back where you were before. And repeat that for all parts. So it's conceptually really, really simple. And that's on something, strength of the project on the other hand, it limits what can be done recently. So there's no huge graphics engine where you can cut one shape out of another or something like this. But it's basically just go along, turn, turn, turn, turn, be back where you were. So for everything that's made of rectangles or simple shapes, that's really, really fine. If you do something complicated, it's probably not the right tool. And there are a couple of generators that come against that edge already. Back then, I was basing it on Cairo, LibCairo, that's the graphic libraries from GNOME, but it's has been replaced. There's a command line interface, that's how it started. There's no web interface, which we'll look at in a minute. And it's now even available as Inkscape plugin, so you can generate a box right away in Inkscape. There are now about over 100 different generators. We'll look at a few of them now. That's the web interface. It's not great. I'm not a graphics designer. I'm also not a web designer. I'm happy that it does something. There are a lot of generators in each of these points that can be opened. And this is the interface for generating a box, basically. They all have the same layout. On the very top, with the error going to the right, there are settings for different edge types. In this type, only finger joints are used here. So there's only one, you can open that. We'll look at the details later on. Then there's the section with the per generator parameters. So there's the different sizes. Many of the boxes have an outside checkbox, where you can decide whether the measurements are for the inside of the box. So if you want to fit something in there or if they're on the outside of the box, if you want to fit the box inside something. A lot of boxes do have the ability to change the edge type on top or bottom. We'll look at those edges types in the examples. And they may have a couple of other parameters. In this case, we can decide how many walls are there on the side, how many angles. And we can decide if we want a lid basically or not. And on the bottom, there's the default settings, which we look at a bit bigger. So some of them are really important and you need to take care of them every time. And some are not that important. The really important is the thickness. This is about the thickness of your material. I strongly advise to always measure your material before putting it in a laser cutter. Because many materials don't always come in the size that they are supposed to be. So plywood, swells and shrinks. Although that's not that critical because it can be forced into its nominal dimensions. But especially acrylic is very brittle and if it doesn't fit, it won't fit. And you really don't want to use a file to enlarge all the holes that you just cut. So do measure the thickness and check twice every time. So you can select different formats. Default is SVG, which depending on the program you're using to use your laser cutter may be a good choice or not. It also can do DXF, which is the second proper format. The second really, really important value is on the bottom, which is burn or RK curve. This is how much the laser is outset to account for the thickness of the cut. This needs to be pretty precise for things to fit. And I typically change this in 200 of a millimeter. So it can be 0.1 or 0.08. Or if you want to lose of it, 0.06. The problem is this all depends on your machine and on the material used. And might even change with if you're using different speeds and powers a little bit. So it's something that needs to be tried out or you need to have experience with that. There's a tool that we're just looking at for checking this. So there are a few more stuff here that's not that important. There's the tubs, which allows you to leave small gaps in the cuts so pieces don't fall out. That might be useful if you want to keep all your pieces within your sheet of material. That's supported by most parts. There might be some hand-programmed stuff that's not supporting them, but most places to support them. So pieces stay in. This is also very critical if you want the pieces to pop out easily. Otherwise, you need to cut them. So 0.1, 0.2 is a good value there. Even if you have 1.1 mm too much, it's really hard to get pieces out, especially smaller ones. But you can cut it with a knife if you need to. Debug allows you to put boxes around specific features to check alignment. It's only something you need if you're actually programming new generators. And there's the reference, which renders 100 mm Reference to make sure your drawing has the right scale. Because a lot of tools mess up the scale of your drawing. Because both SVN, SVG and DXF don't really come with a built-in measurement system, but they have just numbers and different tools have different ideas of what they exactly mean. So if you're new, you may want to keep this. If you have your toolchain established and you're confident that everything works fine, you can also set it to 0 and it will disappear. And you won't need to delete it before actually cutting. So in the bottom, here is no image, but if there's an image, it will be off the finished product. Well, an example image of the product will be shown here, and there might be building instructions. But to be honest, there are very, very few generators that come with them. So if you want to write some open a ticket, I will add them here. At the bottom, you can choose a language. It currently is translated to... It is in English, it is translated to French and Chinese right now by Cotabrotas. So here's the Brunt Test, which allows you to check different fits. I need to speed up a bit. So there are a couple of different boxes. I have just put a couple of images here, for examples. On the top left is the one we've seen before. That's the angled box. The important thing is on the bottom right, which is the universal box, which is a normal box, but you can change the top and bottom edge and allow to add a lot of different features to the box. One thing that's worth knowing is, per edge type, there are settings that can be adjusted. It's not really intended to adjust everything, but there are a couple of things that are useful. So the angle is something you should leave alone, because it's typically set programmatically to something useful. There are different styles. There's a style that has small springs to make it easier to assemble. And this assemble the finger joints. The stuff that's really interesting to change is the fingers and the space, which is basically the width of the fingers and the space in between. This is a bit weird at first, because these are in multiples of the material thickness. That means if the fingers are of size 1 and the space of size 1, all fingers are basically little squares. And the default is to have them rectangular, one thickness high, two thicknesses wide. Another thing that's worth adjusting is the surrounding spaces, which gives how much space needs to be left under left and right, outside. The two is actually pretty nice. It means basically one space on each side and one complete space needs to be on the left and the right. But if you have a very tiny box or a tiny part, that means that you don't have any fingers at all. So it might be worth reducing this to like one or one-half to get this one finger that keeps your parts together. So these three are actually the ones you should have in mind. If you don't like your box, then you can adjust the width of the fingers in the spaces and the surrounding spaces. And that's basically all you need here, really. Many of the other settings can be ignored unless you have good reasons to play with them. Same for most other edge types. So, here an example of flexboxes, which was one of the first things I wanted to do. In retrospect, they're kind of a gimmick. So, it might work better. The current default settings are very, very flexible. They might be a bit better working if they were a bit wider and sturdier. But it's a nice gimmick, basically. It takes quite a long time to cut because there are a lot of lines to be cutting. I'm relatively proud of the one in the bottom right, which actually works. So you can slide on. The lid disappears underneath. Unfortunately, the photo doesn't show it. There's a second bottom underneath, so it actually really disappears and you can put something in there. So, let's come to the really useful stuff. This is trace and draws. You can either, on the bottom left, there's something you can put into a drawer. You might want to adjust it by hand with Inkscape or something. If you have a drawer, that's not perfectly square. There's not really something to do that within a generator. On the outside right, there is the type tray generator, which allows you to give all the distances between those walls. And on the left, there's an extended version of this, where you can not only give the distances you want for those walls, but you can also basically delete single walls. The interface is a bit, let's say, okay, it looks like this, scaled up. So it's basically a text area, where you can add those numbers, and then you can delete the single lines that represent those walls, and those walls will just not be rendered. So you can do more complicated setups like this. Then there are a couple of shelves, which are also useful. The Rhine rack is actually just around the corner. The middle shelf is the one in our hackerspace kitchens, to have all kinds of spices. Two on the left are actually contributions from the parties that needs and submitted their generators. Those shelves are actually in use at a lot of places. This is our snack bar in our hackerspace. It's a bit difficult to see in the pictures, but those shelves are slanted forward. So stuff will slide to the forefront. Of course, we had to do all kinds of corona stuff with Boxes Pi 2. That's the sad reality we had this year. There's a whole section for tools and sled walls. There's no good photos yet, because I'm not that happy how this works out yet. The actually finished items are okay. They're actually cool, but the problem is because the sled wall is actually meant to be ... So the sled wall holders are meant to be stamped from steel. So they are wide and strong and steel. I have only basically those small hooks. They might have thin plywood. And while they are able to hold the stuff up, they are easy to break, especially when you install it or uninstall it. And it's easy for them to get snagged and then broken off. So I might change that to actually use aluminum profiles to hook in there. But they're there. So there's another section with parts and samples. There's all kinds of stuff and we don't have pictures yet, but you can just slide through there. Let's look at the time. We need to hurry up a bit. Then there's a MISC-Session, which has all kinds of other stuff. There's an acrylic-colour insert that just got donated by a contributor like a couple of one or two months ago. On the top left, there is my Makita Akku-based power supply that makes use of a cheap Chinese Bench power supply. And there are all kinds of other things that are in there. Then there are a couple of bigger things that uses Box.py for as a basis. One of them is Flipper, my Cocktail-Borobot. This was done basically to see with how little vitamins one can build a Cocktail-Bot. So it's able to use 15 Bottles. And what it does is the thing on top turns around and then pulls these levers to tilt the bottles and then pour stuff into ice and go through the ice, it gets outcooled and mixed. Everything is done with only two stepper motors, whose heat sinks you can see or the heat sinks of the drivers you can see on top. Another thing is the Autobot. We did a couple of years ago, which is basically just an integrated Hingebox with Lex. We wanted to do this as a project for kids. Origin was 3D printed and 3D printing such a thing takes one evening per robot and now it takes like 12 or 15 minutes. Then there's our rotary attachment, which probably still needs a bit of work and love. So there's now a section. A lot of, there are a couple of people who do need finger joints and stuff, but don't want to go through the hassle of actually programming a new generator. And so what they do is they just take some boxes and cut them into pieces with Inkscape and then make something new out of that. We'll just want to show a couple of examples so to get a feeling of what can be done. Don't ask me about the details, because that's not what I do. So, there was the charging station for the Conquerors, I think two years ago. There's an useless machine, which should actually be a generator, but you can also just combine two different boxes and have one. And here's a Maslow-type milling machine that we are currently building that's also made out of simple pieces of boxes pie, heavily modified, to be honest. So, I will just be a bit of late, where we'll skip this part. Generally, if you design something, it's important that you make good use of the material and a way to get stiffness and things is basically to enclose volume. That gives the stuff strength and to keep all your forces within the walls. Basically, the more boxy it is, the stronger and stiffer and better it is. I want to show here some detail of the Maslow-Machine, where you can see really nicely how all the parts are basically trying to wrap around the bolt they're holding and basically supporting the part from all directions. So, whenever something needs to move, it needs to twist a whole wall within itself, which is, of course, not that easy. The thing is, wood is relatively flexible, so you need to get a stiffness and strength from the form itself. And that's a good example of how this can be done, instead of just having one wall that attaches. So, if you're interested, you can also do your own generators. If you don't want to use to go down the inkscape routes, there's some documentation that's linked on the generator on the bottom. It might be worth asking for help first, because I'm a lazy guy, so a lot of things can be done easily, if you know how to do them. And if you're not, there's a long route that you probably want to avoid. Architectures, although it's just total graphics, there are a lot of abstraction layers to make this feasible. So, there's the top layer with the UIs. There are the generators. There's code for handling the settings. And there are setting glasses for all those edge types. Then there's a handful of parts that are basic shapes. They do have callbacks, so whenever you need holes or features within one of those walls, you have a callback to use or you have a callback per edge, you can use to put those in place. In there, the main work of the library is done by the edges. And so adding a new edge type allows you to add features to the outside of some parts. So a lot of things can be done easily by just trading a new edge type and put it at the right place. And at the bottom are, of course, drawing permatives, if you really want to draw stuff on your own or use them to implement the edges. That's basically the very, very rough overview. There's some documentation that's worth looking at, or you can just ask me or Penetrate it. Happy Lasering! Thank you for a nice talk. I hope all the people could get something out of it and want to play around with that. Just one question at the end. I have an answer from the Signal Angel. There's Douglas. He has a 3D printer for which a laser attachment is available for about 30€. It has a violet laser with 500mW. Is this sufficient to cut wood? Let's say this is a dangerous toy. So this is just enough to damage your eyes. I mean, you can probably cut paper with it. And if you're running it very, very long, it will cut through wood at some point, in no point. So our laser cutter has 80W. And while it might not focus as well as a diode, it's probably not a good idea. And the thing is, you really should have a closed box around your laser. And the box should not let any of the laser type you're using through. So the CO2 laser has the benefit that it doesn't go through. So a lot of materials are not transparent for the CO2 laser. But ultraviolet laser is another story. So it will probably go through acrylic and stuff like this. And you really don't want UV lasers in your eye. And you won't be able to have fun with boxes like this. That won't work. There's a reason why our machine costs like 3 zeros more than that. Okay, thank you for the information. That's it from the Q&A. For everyone here at remote congress, I wish you a nice event.