TuesdayI arrived at Nuremberg after a very long travel and nearly no sleep, so I didn’t go to the “warming up” party in the bar, but instead met up with C. and we went to a Balkan restaurant (Balkan Spezialitäten) for dinner. We got Cevapcici sausages with some sour cream (called ajvar) and some kind of bell pepper sauce (kaymak, which confused me greatly since in Poland it’s a name for a sort of caramel) on the side, with some very nice bread (börek). The food was fine, but not amazing. But the restaurant was pretty quiet, probably because it seemed to be more focused on locals than tourists and it was pretty small, so we could talk.WednesdayState of Libre Graphics by MABarbeThe first actual day of LGM started a bit late, at 1 pm, but I used all that time in the morning just to sleep. First item on the agenda was the State of Libre Graphics, which is always exciting since you can learn about many small projects you don’t know existed.Some of the projects that are quite interesting:Blockbench – a simpler 3D modeling tool, made for low poly models, with painting tools and even animation. They have a partnership with Mojang, which uses Blockbench for models for Minecraft, and with Hytale so it becomes the main modeling tool for the modding community.Coollab, Processing, p5.js and OpenRndr – they all seem to do roughly the same thing, which is procedurally generated 2D graphics. Coollab seems to be a software with a node editor, while others seem to be libraries you can use to code the graphics. I haven’t explored them enough to know for sure what would be differences in how they work; I have only seen how OpenRndr works, and it seems like it’s a coding library in Kotlin which spawns a window with the graphics you coded. It also has some capabilities to make a GUI in that window.Friction – a motion grahics software, I believe that’s an alternative to Natron and Adobe AfterEffects.Glaxinmate – a 2D vector animation tool. It’s still in version 0.6.0, and it’s part of KDE.Graphite – a software for vector graphics and animation. It provides morphing between shapes, it has both layers-based compositing and node-based generation of graphics. And it’s supposed to release 1.0.0 version really soon. (Seems already more advanced than Glaxinmate).Pixiv – another node-based pixel art and animation, both frame-by-frame and generatedOurPaint – a digital art software, it has some elaborate pigment implementation.Hyper 8 – if I understand correctly, it’s a static page generator (like Hugo, which is what Krita’s website is built on), except focused on videos. It has browsing, searching, playlists, localisation, collaborative editing, and interactive searchable transcripts, and all that while promising the setup to be really easy, not requiring any database, programming nor maintenance, and compatible with all webhosts.Ladron de Flores – that’s actually a movie that uses lots of FOSS software like Inkscape, Krita, Opentoonz, Kdenlive, Audacity and Aegisub (a subtitling tool).Ossia Score – a software for live art installationStellarium – desktop planetarium, it is used now in Vera C. Rubin ObservatoryTiXL – realtime/live motion graphics (which I guess makes it similar to Ossia Score?)Trying out BlockbenchAfter the LGM I went through different projects to check them out; I already made a voxel lamp in Blockbench and imported it into Blender. It took very little time to achieve this result. With polygonal editing, I had a bit more trouble, especially since Blockbench has alt-clicking for some shortcuts, and that of course causes trouble on my Linux. I really hope I’ll get a hang of the polygonal editing too, since voxel editing is still a bit tedious (especially since I don’t really see a way to merge the cuboids together), but if not, it’s still great.Trying out PixiEditorI also checked out PixiEditor. Not gonna lie, the website and the software itself looks really fancy and modern. You can clearly see the modern web influences in the design. The Steam page looks pretty good too. You can see that the people working on it know what they’re doing. There are also some fun quirks of the interface; there is for example a mushroom guy who’s giving you tips like the old Clippy, but more adorable, and some text in GUI that clearly shows that it was written by a Polish person that is also not the overly formal language you see in Krita (for example “Ale tu pusto!”, which means roughly “wow, how empty is this space!” on the empty list of projects). It also has quizes on the documentation page to check your understanding of the topic. Makes learning the software more fun and makes the software itself look more casual as opposed to formal and stiff.For painting it’s still a bit limited because it doesn’t have many ready-made brushes; I think there is only six of them. I knew there is some node-based engine within so I duplicated a brush, edited it, and finally found a way to the node editor, and I have a bit of mixed feelings about it. Obviously it’s amazing to get nonlinearity, but I have this impression every time I actually try to do something with nodes that it’s a very inefficient system of writing that nonlinearity. In PixiEditor it’s of course still a new system, so you can’t expect it to be already perfectly adjusted for optimal use and I’ve never actually used any node system extensively, so I’m not a good judge of it. It looked promising, though, so I want to try to make some of my favourite brushes there to test the painting side of it properly. Layers system is also node-based (more specifically, a proper directional graph, instead of a tree like in Krita with Clone layers introducing a bit of nonlinearity). I haven’t really explored the full potential of the software yet.Freeze+Press – talk by Christoph HaagFreeze+Press seems to be a small printing workshop of sorts that is focused on experimental printing and using FOSS and CC. Currently they have a website where you can buy some stickers (ImageMagick’s wizard, git’s logo, broken image logo, the Canonical tiger), posters and books. A portion of the talk in the beginning was about Christoph’s own story in FOSS and graphics design and how Freeze+Press was created.The elephant in the room: who owns the rights to an image? – talk by Eylul DogruelAfter the coffee break, there was a talk by Eylul with a title “The elephant in the room: who owns the image?” which was, how you can guess, about AI generated images. I highly recommend you to watch the recording on the Chaos Computer Club website: https://media.ccc.de/v/lgm-2026-110662-the-elephant-in-the-room-who-owns-the-image.At first, she changed the question into “Who owns the monkey-generated photo?”, and I understand that it might be similar from a legal standpoint, but for me, that’s not a good analogy, because monkeys are sentient. It would be more comparable to a baby taking a photo, or maybe a intellectually disabled person (since a baby would grow up to become an adult with all the rights and capabilities), taking a photo. From that standpoint, imho the owner of the image is that person or that monkey, and since they can’t represent themselves, then the guardian should represent them and control the copyrights. That’s of course all completely not applicable to an AI… Monkeys aren’t trained on stolen images, monkeys also aren’t blindly generating average photos out of all the photos they’ve been trained on. In case of AI, the AI is the combination of all that stolen data. A monkey might have an owner or a conservation organisation that should handle their copyrights; but AI was created by the company that owns it, from all the pieces of stolen data. There is, unsurprisingly, a lot of differences between a monkey and an AI.It’s a shame the time for questions was so short; I could imagine quite a long discussion about it. I also noticed that, in general, there is a bit of disconnect between the way artists and programmers see AI; and since LGM is sort of a mix of both (when some days layer someone asked who is an artist and who is a programmer, nearly all people raised hands to both questions), you could see voices from both sides. Artists are usually (not always) very against AI on ideological basis, while programmers often just see it as more of a tool.ctx vector desktop — rasterizing as late as possible – talk by Øyvind KolåsThe next talk was about a ctx vector desktop. I’m still not exactly sure what it is, but it might be because it seems to be many things at once. The main part is a software rasterizer for vector graphics, and everything else seems to be something on top. There is a terminal that is capable of ctx vector graphics and also audio (the terminal itself also written to use ctx), there is a protocol that you use in the terminal to draw the vector shapes, and of course the vector rasterizer itself that later combines it into raster so the downstream (a display server?) can render it as it usually renders all windows. I might be wrong though. If it seems interesting, go watch the talk, Øyvind surely explains it better and without the confusion I have.Beginner-friendly shader programming in p5.js v2 – talk by Kit KuksenokP5.js is a Javascript library that is used for making procedural art, but also focused on teaching users how to code. For the 2.0 version they want to simplify the API for shaders even more so that the user can start coding and making art with as little previous knowledge as possible.The part of p5 for shaders is, I believe, called p5.strands. The user doesn’t have to define their shader completely, doesn’t have to know how the lights work or anything, instead they can add modifications over a default shader. It’s also written in a way that is familiar to users who used p5.js sketches (which art programming still on CPU). The point of it is that users get familiar with shader programming concepts while still using more familiar syntex. P5.js also supports proper GLSL shaders, so they are also available for more ambitious students.Phosphor for Tectonic Typefaces – talk by Émile GreisThe last talk was about Tectonic Typefaces in Phosphor. If you are, like me, unfamiliar with font-specific lingo, a tectonic typeface is a kind of typeface that’s composed out of usually same-shaped small elements, usually dots. It gives the typeface this blocky appearance. I think all pixel-art typefaces could be defined as tectonic typefaces. The main point is that it can be defined as a table with some cells being filled and some not. Then the rendering can be more creative, you can add slopes and rounded edges and different sizes for different cells. Looks pretty interesting, though I can imagine that it’s pretty niche.Dinner timeAfter the talk there was a photowalk planned, but I was at the (very brutal) photowalk the previous year, and I knew going to the castle meant quite a lot of workout, because it’s on a steep hill, and that would probably just be a part of the attraction, so I decided to skip it.For dinner we went to Tucher Mautkelleron the Mauthalle street, hoping for better food and maybe some space to talk as well, but it was unfortunately very loud (and quite expensive). I ordered the salmon trout on spinach with rosemary potatoes, and a side salad. The side salad was definitely not a side salad, and not very great either, but the fish was great.ThursdayRe:WIRE workshopOn Thursday there was a workshop focused on the past and future of the LGM. I made some notes about the past. The future was still discussed later on BoF and the workshop on Friday.(This was a place where I wanted to include a photo of a page of the first issue of the LGM magazine, because it contained some funny information about Krita and Gimp I think; unfortunately, I couldn’t take the photo at the moment I saw it, and then I forgot. I also took some notes about the history of LGM, but maybe I’ll write about it next time).Cherry tree blossoms outside of Zollhof buildingLunchFor lunch we went to grab something easy to eat from the main train station. I found an amazing chicken wrap in one of the stores in the food court there (place called Dean & David, the wrap was called “Chicken & pesto”, 4.95 euros, I bought two more for my travel back and it was so worth it).Soldering workshop with Simon BudigLater we went to the soldering workshop. I signed myself as intermediate, even though I’ve never held a soldering iron in my hand before (but how hard can it be, right?). The theory is relatively easy, but we started from the resistors, which are incredibly tiny and finnicky to solder onto the board. First you had to heat up the tiny pad on one side, and then put solder on it so a tiny bit of melted solder stays there; then, using tweezers, put the tiny resistor in between the two tiny pads, and finally heat up the other pad and add some solder to it. The next step was adding LED bulbs; they had a much longer “legs”, so you in theory didn’t need to be as precise anymore, but they had a tendency of bouncing off of the board when you weren’t looking, and keeping both legs flat on the board wasn’t that trivial. Unlike the resistors, they also had a correct and an incorrect orientation: you had to make sure to solder the shorter one to the pad marked with L. Thankfully I didn’t make that mistake. For the “spider” with a switch, the only difficulty was that you can only heat up one pad and one drop of solder at once, so preparing all three pads on one side is a mistake. But it was still managable. The last part was the battery holder, which wasn’t really complicated, though for that you needed an inside knowledge that you can heat the holder’s pad up from inside of the holder after putting the solder on the pad, which binds them together. I think I did it correctly the first time, actually, but I wasn’t sure, so I redid that part (and getting a battery was tricky, so I only checked with battery after everything was done). I also had to redid one of the LED bulbs since one of the legs jumped out of the solder. But all the lights worked correctly the first time I put the battery in. Hurray! Full success. (Just don’t look at the finished board).The soldering workshop was located in a reception area of a local theatre, because the venue didn’t have appropriate fire safety certification or something. We did have to remove some carpets from the floor though. After the workshop, I tried to teach M. and C. how to play Star Realms with Polish cards; we were sitting among all the chaos, so it was of course a disaster. The main rules were explained though. No one won, because we had to finish abruptly when the workshop ended and there was time for cleaning up the space.There was a group photo made during the coffee break.Gimp: A community, free software – a talk by JehanLast year I could see for myself how Inkscape works and organizes themselves, so I hoped to learn the organisational structure of Gimp this year as well; the talk was good, but I hoped it would go a bit more into details. Then, of course, I spent a whole day with Inkscape (which was delightful) and Gimp only had 20 minutes to describe themselves. One of the things I noticed is that they have a lot of success with their GSOC students, with most students becoming regular contributors after the project ends. I think we can still improve on that aspect in Krita.Dinner timeLater me and some others went to a burger place near the main road in the Old Town. I got a classic one, it was pretty nice. We talked about so many things — definitely a bit about AI, about an illustrator, who lost nearly all of their commission work (presumably to AI), about more ethical AI projects, and then of course NFTs, and then finally stumbled upon pigments mode for digital art software (and color management in general). And that lasted the next two hours. There are three separate topics in it: one for proper color management and what is really even is, one for HSX “models” which are just coordinate systems for the RGB color space (so they change depending on the RGB profile you’re using, but there are better and worse ones for different purposes), and the last one, for the pigment mode.FridayCounterpunch Font Editor: Faster, Safer Complex-Script Design – talk YanoneCounterpunch is a font editor with a great name. And it was made by a person who designed font that Krita uses for its logo. It’s browser-based and I can’t really tell you much about it, because I haven’t made a font in my life, but it’s in alpha and apparently it fixes some issues that other font editors often have.ReWiring the Video Editor – Timeline as a Node, a talk by MABarbeMABarbe did an amazing talk about a new concept for a video editor, node-based, but also with timeline (or more precisely, timelines, since it is a node and you would be able to have multiple of them at once). It would have some auto-adjusting capabilities so if you want to just reorder clips, even if they are not the same length, you can just switch them around and the rest of the timeline will be adjusted to the new state automatically. MABarbe also said that the common workflows should all have their own nodes with GUI instead of requiring users to fight the node system for something it wasn’t meant for. I was very impressed with this idea; I have tried to use Kdenlive, and even when I knew exactly what I wanted to do, it was not that trivial to use it with the rigid timeline system. Of course I’m not really a video editor and I don’t even know Kdenlive that well, but the system M. was talking about just seemed more suited for the way I’m thinking about video editing (my internal paradigm).It’s all still in the idea phase, there isn’t even any proof of concept, but I (a layman) was convinced. I wonder if it would be possible to implement it into Friction or Natron instead of a brand new software; but that of course depends on many things.LunchOn lunch we went to a falafel place. I got a classic falafel sandwich, while C. got the Arabic one, which turned out to be cut into 6 or 8 pieces and served in a takeaway salad bowl. It was fine, though I felt like it would benefit from more tomatoes. At the end of the break, I tried to rehearse my lightning talk, but I didn’t manage to say it in full. It was very nice that people wanted to listen to it, though, since they were bound to hear it again very soon.Arabic falafel on a plateLive Coding for Artists and Designers – workshop by Abe Pazos SolatieThe workshop about OpenRndr was very interesting. Such a shame I didn’t bring my laptop. M. was sitting next to me though, so we were sort of coding together. To get OpenRndr to work, you need to install IntelliJ Idea (to get Kotlin), and then just downloading the template project from the git repository of OpenRndr. The full guide is here: https://guide.openrndr.org/setUpYourFirstProgram.html. The language looked quite alien to me at the first glance, but I had to be brave and act as an expert to M. (who doesn’t know much programming at all) and explain what things mean when I still wasn’t exactly 100% sure yet. But, ultimately, it’s just a bit different syntax, that’s all. You just need to know what to put where, and having an example source file, you could guess quite a lot. (Turned out that most of the confusing syntax was just lambdas over lambdas with brackets being an unnecessary part of the syntax in those cases, therefore omitted, therefore nothing looked like normal functions).What’s interesting about OpenRndr is that apparently unlike other live coding projects, when you create shapes, you can take them in vector form and apply all sorts of transformations and calculate more things from them. For example when you create a contour from a set of points, you don’t just get to draw that contour, you can get the vector shape characterised by that contour and for example add arrows representing normal vectors to it. The program is also interactive, so you can make it listen to mouse clicks and for example change colors. It also knows the passage of time, so you can animate the shapes. There is a lot of different packages it can use, most notably you can also make a GUI for your graphics to, for example, control the thickness of the lines while the program is running. Overall it looked a bit like a game engine, just with a way more 2D vector graphics options. I was considering using it for prototyping some things like a gradient editor for Krita I’ve been thinking of (with defined curved lines across the HSY space).Which also reminded me of Seexpr library in Krita. Such a shame it’s so user-hostile and so limited. I hope that I could someday get into it and implement two important features: being able to output opacity, and using Seexpr as a filter, using the projection of the layer below as a base. If those two things are done, then we’d have to make the GUI less user-hostile. Maybe someday we could have an easy-to-use shader layers in Krita.Real ultramarine rocks (located outside of workshop room)Lightning talksI won’t be describing each lightning talk, but there was one that caught my attention. It was an ad-hoc talk about mental health. Started very well, encouraging the atendees to take care of their mental health. I totally agree, that’s a very important topic, something that everyone needs to pay attention to. But then it went downhill, towards, I’m sad to say, pseudoscience; the speaker mentioned that we have no evidence that the consciousness is formed in the brain (which is untrue, we definitely do have evidence of that) and discouraging people from taking psychiatric medication (which is straight up dangerous).Especially the last one bothered me, because there are several different disorders that improve greatly from being treated with medication, and many kinds of those medication is proven to work really well. We have good anti-psychotic medication and schizophrenic people really need it. We have good mood stabilizers and people with bipolar really need it. We have good ADHD meds and people with ADHD, maybe for different reasons than people from groups mentioned above, really need it too. And sure, there is a lot of people with depression and anxiety and our antidepressants and anti-anxiety medication isn’t as effective as some other medication, but it is still proven to work in a decent percentage of people, and there are multiple kinds if one of them doesn’t work on you; and encouraging those people to not take medication is also harmful: especially if someone is easily influenced and stops taking it abruptly. Depression can be deadly and stopping taking medication like that increases the risk even more.And sure, psychiatry and psychology is still in infancy when compared to standard medicine, because it only started being treated seriously way later. But it’s still the best we have, and it contains only and all the things that we know about human psychology. We can’t just say “oh but it’s not perfectly advanced yet” and discard *the only thing we know for sure* for some feel-good notes and “common sense” theories or “spiritual healing”. Discouraging people from seeking evidence-based help can, in my opinion, lead to all kinds of adverse effects. Spiritual healing can always be applied on top of actual medicine.Of course, that was just a lightning talk, improvised just that day. Still, I think it’s an important topic that requires quite a bit of care and foresight applied when talked about.Liblast 3D asset pipeline – talk by Tobiasz ‘unfa’ KarońThe talk about Liblast was mostly showcasing the pipeline for creating assets for the Liblast game (which is a 3D shooter) that allows them to create decent-looking 3D assets pretty fast. I wrote more about the pipeline further down, when talking about the workshop.Tixl Skill Quest – talk by Thomas MannTiXL has an interesting idea for onboarding new users: they created a whole tutorial series in the software itself. TiXL is a software for real-time motions graphics, I believe it’s purpose is the same as ossia score. The tutorial seems to be sort of kid wheels for the software, you are guided to do tasks that use more and more advanced features. It reminded me of the Godot’s GDQuest’s tutorial series.Design Students Experimenting with Free Software – talk by Lila PagolaThis talk was really interesting. Lila is a teacher in some kind of design school and she tries to open the minds of her students to FOSS software in a pretty creative way. She gives them an assignment to choose one FOSS software that seems like it could be useful for one of their typical design or graphical projects (since they are on very different paths, their typical projects differ significantly, so each student just chooses the one that fits their work), do that task in that software, and then evaluate the software, and I think even the feedback was supposed to be divided into two different categories: actual problems with the software, and issues caused by their lack of familiarity with the software. She said that in the past she just wanted them to evaluate the software on its own, without having a project they have to do in it, and she said it didn’t bring much success: students actually using the software was crucial. She also said that after her classes, some students do start to use FOSS software. They are often surprised that it is possible to do things in free software, while up to that point the only thing they knew was one specific “industry standard” program they were required to use for some other classes.EveningBefore dinner me, C. and M. played Star Realms again, this time successfully, in an empty community area. I was also showing the good quality black pens you can buy in Poland that are perfect for drawing.Later for dinner me and M. went to the burger place again, taking C. with us. This time I ordered some weird burger with potato chips. It was way too big to eat it the normal way, so I had to split it horizontally in half and eat it with a fork and a knife. I wanted to try it to understand the reasoning for potato chips on a burger, unfortunately, this experience didn’t enlightened me. Maybe if I added them myself, they’d add crunch, but that way it didn’t really do anything, and the flavour was somewhat lost among all others. So I think it’s just a gimmick.My view while eating the burgerSaturdayImposition methods for bookbinding – talk by Tom LechnerTom Lechner has this very interesting software called Laidout, which is a desktop publishing tool with some vector graphics capabilities. The talk was mostly about various methods of imposition, meaning arranging pages onto physical paper sheets. There are several different usecases, for example when you want to print a sewn together book, you need to print text pages in specific order so when you fold and sew together all the pages, what the reader sees is actually all pages in order. You might want to print a big poster or maybe a sewing pattern on a home printer which only prints up to A4, so you’d want to arrange pages first so you only print the parts you need. Or you might want to print a folded pamphlet, or maybe a sphere-like polyhedron with a map of Earth. It can also be any other 3D shape (from what I can tell), as long as it consists of polygons.Learnings from Our first 3D Game in Godot – talk motion ensemble (an animation studio) and Alexander LehmannI’m writing this part more than two months later, so I don’t remember the details, but I couldn’t just omit this talk. Or rather, this game. The studio was tasked to make a game for an organisation or institution whose whole job is asking the government for information and bothering them until they give that information. Just paperwork after paperwork. So, they made an incremental game (think cookie clicker, but better) in 3D in Godot (see the screenshot below) where first you gotta write the letter, then print it out, then send it, stuff like that, and each stage is shown with this low poly 3D in a very organic way – you can see every letter, everything. Later of course you can advance and get more funding and buy stuff that makes you faster, and what’s best, at the end you can hire more employees, and each and every employee you can hire is a direct representation, with a face, of an employee of that organisation. I really loved this idea. And the game looked fun (at least for the first hour, hard to tell how complex it can really get. The screenshot below is just a small part of of it, of course, but it shows the process of writing, and the stamping).LunchDuring lunch me and C. stayed in the community area in the venue. We intended to play the game again, but somehow talked through the whole time. For sure we were talking about the Welcome Page widget, and C. was trying to prove to me that converting a 8bit/int image to 16bit/int image in Krita creates incorrect results. Turned out the conversion works fine, but the dialog for the conversion is a bit user-hostile — I believe it’s the same widget that’s used for creating a new image, and when you switch from 8bit/int to 16bit/int, the default colorspace switches from sRGB-trc to a linear sRGB. That caused difference in rendering. Switching it manually back to sRGB-trc in the dialog and only then converting solved the problem.Liblast 3D asset pipeline – workshop by Tobiasz ‘unfa’ KarońLater I went to the Liblast 3D asset pipeline workshop. Liblast is a libre multiplayer 3D game. There is an older legacy version and a newer one, with rewritten codebase, still not yet feature-complete comparing to the old one, but much faster. The workshop is about the tools they use to create relatively good using assets in a short amount of time.The basic idea is relatively simple: use pre-made normal map, automatically applied onto the mesh, for as much of the content as possible. When unwrapping, the script in Blender compares the faces with areas on the normal map and assigns the UVs in a way to add bevels to the edges of the faces. The normal map includes a few different sizes, both square and rectangular, and there are both vertical and horizontal rectangular shapes, so if you later want to for example add dripping paint to the material, the unwrapped UVs are consistent with the orientation of the face in the world (the bottom is always in the same direction). There are two types of bevels, square and rounded, and there are multiple other additions and some space if you want to add something more. You can change the scale between the normal map and the mesh so it fits your mesh best. Sometimes you need to adjust the UVs, either because it’s a more difficult framgent (not perfectly rectangular), or for other reasons, and of course you also need to mark the seams. Besides normal maps, there are also some dirt maps and common markings textures (warning markings, metal plaque, this sort of things, because the game itself has an artstyle with a industrial, brutalist vibe).Tech-wise, the key element of the pipeline is a Blender addon called DreamUV by Braam Eulaers. It requires an atlas. The atlas in liblast case is called “hardsurface_atlas_0“. A TexTools addon is also recommended. There are some more maps as well, not just normals.To test it yourself, first you should watch the talk, then go to Liblast repository: https://codeberg.org/liblast, and download a runnable project here at branch devel-unfra: https://codeberg.org/Liblast/liblast-framework/src/branch/devel-unfa (or maybe specifically at hash https://codeberg.org/Liblast/liblast-framework/commit/a2781c6726255ac026cc8cc5c67ff04c84d5f262). You gotta use it with Godot 4.6.2. For the atlas, go to liblast-assets at branch feature-hardsurface: https://codeberg.org/Liblast/liblast-assets/src/branch/feature-hardsurface or even more precisely, https://codeberg.org/Liblast/liblast-assets/commit/4ca78bd5501f49811c61125f4b5f42632d318aad (they are at textures/hardsurface_atlas_01/export).Normal map for the workshopUpStage – talk Helen Varley JamiesonUpstage is a sort of a video call application where you can do a lot of things to the way the main screen is presented. You can add images, text and other elements, you can scale things down, add a background, add a video stream with green screen applied. It can be used for artistic and theatrical installations.EveningAfter the ending we went to the hotel to switch some things up (I wanted to leave the card game there, and get all the warm clothes I could get, and I also made myself a hot tea as well in a thermos cup), and then we went to look for something to eat. We walked past the talafel place, saying that the sandwiches were pretty small, but as we went into the Old Town, I think we, or at least me, started regretting it, not seeing anything suitable to grab to eat on the go.Finally found a street with many different options. Some people opted for a kebab, but me and K. went for an unsuccessful search for some different kind of wrap. After a few failures, we decided to step into a fast food place with fried chicken. I got a long sandwich and a round sandwich, knowing one fast food sandwich would be too little. There was an option for Polish in the self-order kiosks, but it was clear from the first screen that you should only use it for amusement and not for actual ordering, since the first question, takeout or for eating in the restaurant, had takeout translated to “Podnieść” (which sure, means “picking up” literally, but not only it’s a wrong grammar, but also the word is only used for literally lifting things up in the air; the appropriate translation would be “Na wynos”. There were several other funny mistakes later, as K. noticed).The meetup place was a small plaza next to the castle, with a café and many restaurants around. The only trouble was that the floor was paved with a big, uneven, slippery rocks, and the filling in between them didn’t reach the top of the rocks, so you couldn’t really sit comfortably on it. But it had nice a view, there was a lightened up tree over us, so we spent all time until half past eleven talking. After that we returned to the hotel and, after saying goodbye to M., the rest of us decided to play the game in the hotel lobby. C. really wanted to explain the game, and then he was of course winning again, but with a few carefully directed “a selected enemy discards a card from their hand” cards I took the lead and it started to seem like I might win, but then we had to stop playing since we were all a bit too tired to finish it properly.Nuremberg at nightSundayOn Sunday we ate breakfast together and got to the station. I bought my favourite wraps at the station food court (amazing decision). Travel went pretty smooth, despite the fact that I had to use a replacement bus for part of the train journey; thankfully it was sufficiently marked, I got there in time, and I also got to the later train station way before the departure as well. After two trains and one bus I also hopped onto two trams and finally, very late in the evening, got home.Note: I wanted to write about and describe every single talk, but it was just too much. It’s already pretty long.