Jump to content
3DCoat Forums

Sorn

Contributor
  • Posts

    224
  • Joined

  • Last visited

Recent Profile Visitors

2,935 profile views

Sorn's Achievements

Novice

Novice (3/11)

  • Great Content Rare
  • Great Support Rare
  • Helpful Rare
  • Helpful Rare
  • Helpful Rare

Recent Badges

129

Reputation

  1. In my machines, running different Arch variations, I couldn't use 3DCoat because any interaction with file browser crashed it instantly. I tried this method suggested above, installed distrobox and Ubuntu 22.04 (with nvidia to have 3D acceleration) and oddly enough 3dcoat-Ubuntu22.04 still crashed but 3dcoat-Ubuntu20.04 worked! Apparently I can use 3DCoat again on my machines! Yay! Thank you @veris for this very practical workaround! Edit: on a different (newer install of the host system) machine, 3dcoat-Ubuntu22.04 worked. The file manager looked different too. Anyway, thanks again.
  2. @AbnRanger, let me politely decline to answer your reasonable inquiry about an example(s) of other DCC software with similar interfaces as the one I'm suggesting. Let me tell you why, though; fair is fair. I do know of some, several, for DCC and other fields. But presenting them as examples will easily turn the discussion into a direct comparison about features and workflow paradigms between them and 3DCoat. Thus, rendering the discussion about UI pointless and orphan, mixed up in a horrific cacophony of all sorts of strong opinions about software preferences. Maybe you and I would manage to restrain ourselves, barely, but this is a public forum. It doesn't end well. Second, and most importantly, I really don't care at all about what other DCC software do or not do if I'm thinking about this one: I'm suggesting what I would do if I could, to make the best possible UI for 3DCoat. That, in turn, would possibly be the envy of everyone else, but that is absolutely not my concern. I really don't care if no one has done it before, or if everybody has already implemented it. I'm just stating what my preference is and I try to articulate it in case it makes sense to others and the developers, who knows, maybe pick some ideas from here or refine some of their own, or steer an internal development already underway in this direction. Or nudge them into committing to it if on the fence. They'd possibly already have thought about it, it's not that these fine developers are completely new to this or are totally clueless. They seem a pretty capable bunch. I won't be angry nor disappointed if none of this gets done. But I really don't care if others do it or not. I want 3DCoat to do it, and that's what my focus is. In the past I've been in talks about similar issues with developers of other software, commercial or internal, and I tend to explain the same thing. Sometimes the feedback turns into "here, the limitations are those for us, and a complete overhaul is not possible at this stage. How would we make it as close as possible? Would that work?... and so on". At other times it's absolutely out of the question for some reason and it's fine. And sometimes they like what's already in place and the users too, so no harm done. Still, I'm right and you are not! Nah, all is fine.
  3. Exactly my thoughts, @digman. I'd let the UI as is, if it depended on me. And allocate some time to develop what would be needed to make it as plastic as possible, as stated before. When ready, switch the interface with the new, using the current one's layouts as the default presets of the new. Fix current needs if urgent, really needed, as @AbnRanger states, in the interim. But keep those fixes at the most basic and minimum: the truly necessary for right now. I do understand, nonetheless, that in most DCC programs, some tools ARE the interface, they are coded for the code that manages the interface we see. So sometimes it's far from an easy switch. Many (if not possibly all) tools would need to be touched, and some testing afterwards would be needed to see if everything works fine. And that, assuming the code of the "new" engine for the interface is flawless (it seldom is, by no one, on the first iterations). So here we are. What do I propose, then? Well, leave the UI alone, focus on a mid-term overhaul of the interface code to make it plastic (all dockable, un-dockable, resizable, respositionable, "live" colour coding of the interface with a panel, font choosing and scaling free and "live" (as in no reboot), the whole lot). In the mid term, it's an overhead of work, interfering with current plans of development. With the right plan, allocation of hours/resources weekly, as time goes by, while being tested in stages, it will come to fruition. Once that is done, you have a magnificent platform, with ABSOLUTE freedom for users and no more wondering about what tool goes where, and all things related. Also, users will give another level of feedback too, and new workflows will naturally appear. Many puppies will lick our faces happily and the sunshine, through the window facing the garden, will be warm and cosy forever... that kind of thing. I think of this as an investment in the mid-term/future rather than a "solution" for the present, a solution for the present that still takes resources now and will inevitably will come to bite us again, and again, as time passes, new tools emerge, new workflows are needed. That'd be all from me on the matter. I hope it inspires good things and if nothing I said is feasible or deemed necessary, let's all move along and do some 3D stuff anyway.
  4. Let me try this again, about the UI: What I'd like is a drag'n'drop of everything, from panels, to bars, to windows, palettes... to be able to place them, to detach them from the main UI to make floating panels/windows and be able to attach them back again to the main UI and, if at all possible, to be able to move windows to another monitor and keep focus on them whilst working (so not having to click on the window to interact within it with the mouse/pen). The point is not to be able to set up different layouts to work with, but to do this on the fly, as it were, while working on something. In that maner, at any given time, the user can adjust, adapt 3DCoat to the user's needs. In conclusion: user should be able, at any given time, to place anything anywhere, add or remove, dock and un-dock, while working. That to me should be the aim of the engineers of any 3D application. Give your users the freedom to use your tools however they please, let them see what works best. Provide sensible defaults (ask feedback for defaults so it's easier for newcomers to get a grasp of what's what) and let the users evolve their UI while they work and play. Hopefully this is not controversial. Otherwise I don't know what to say, really. Edit: I forgot one thing, a mode with a button on the interface and keyboard shortcut (so doable with keyboard and mouse/pen at user's discretion) to toggle everything off and back in place as it was to make the window clean of everything and anything but the geometry or working asset at the moment of the toggle activation should be implemented as well. Again, to me, all efforts of the engineers should ideally be concentrated in making all of the above possible, even in stages if necessary until that is a reality. Spending time with UI setups is going to be an endless effort, as new needs and worflows arise, particular cases appear, niche opportunities present themselves, user idiosyncrasies... let them free, they'll invent new workflows with your tools, new combinations: use your users to show you and other what your tools can do instead of anticipating what they might do and how they may do it. There, that'd be all. Edit (2): Nope! I also forgot the super-promise-it's-for-sure-the-last-thing: text should be scalable at all places, and colour of the interface, panels, all of it should be easily changed too. Even while working, without having to restart the application. Now it's final. Final final. Last definitely final.
  5. I don't know exactly what to answer to that... question? Nor do I see anything to do with what I said, frankly. My concern would be: the UI should be totally configurable by the users. All panels, tools, bars, windows... the lot. On any 3D software that should be the aim. And engineers should focus on making that possible, not on the possible arrangement combinations of tools and helpers on the screen.
  6. And some users just want different UI setup for different sessions, or for fun even. The UI should be set up by the user, not the engineers. When developers ask for feedback on UI I get cold sweats. Like when a client wants to "help" with colours. We'd all be better off if they stopped concerning themselves about "the UI" in any shape or form. Please give us a UI we can tweak at will, and stable (no crashes), thanks very much. That's it.
  7. Or maybe there's a safe call to the file browser that's not problematic if on different version of Gnome. I don't know. I'm open to test whatever needs to be tested, gather information as requested too. Other users willing to help and I will need some lead, here; that's all.
  8. There's absolutely NO other program with this issue that I know of. On Gnome standard, either X11 or Wayland, ALL work and can interact with the file browser. It has something to do with 3DCoat's implementation. The fact that works in Ubuntu (I assume it does) doesn't mean it can't work in anything else. It's quite baffling. Anyway, I've been in contact with Sergyii before, about this issue. If they don't want to fix it, we can't help. If they want to, they could tell us what to try, steps to follow to help nail down where the problem is. Because, there is actually a problem there. The last time I was in contact with him he was open to fixing it but said something along the lines of he didn't know how or what to fix to make it work. Hence, if he really doesn't want to schedule proper time to fix it himself, I offer to help out. The users have an invested interest too in fixing this. So we could follow guidelines, track, gather data for him. But if the only answer is only Ubuntu (and an outdated one at that) then we're at a loss, here. At this point I feel it's pointless if we are told this ridiculous crash has anything to do with not being Ubuntu. Again, absolutely everyone else does it perfectly fine. A shame, really.
  9. I upgraded my license to support the development of 3DCoat, but unfortunately I have no use for it in its current state. I don't use Ubuntu and it is impossible to work with 3DCoat. Anything that touches file manager crashes 3DCoat. It's the only program that displays this behaviour of the many professional tools I use (almost none of them officially support my distribution either, but that basic interaction doesn't seem to be a problem for absolutely anyone but 3DCoat). Is there anything we could do to fix this issue once and for all? I'd gladly test whatever is needed if developers care to provide a set of steps to try and figure out what's happening. Maybe some other users could join following steps and reporting so we can fix this ridiculous situation. It's a pity this is a problem. I know of others that already have stopped using 3DCoat for the same reason and it's a waste of a really good tool. 3DCoat can do many things really well, and provides solutions for common tasks on the 3D content creation process with interesting workflows. Let's fix the issue! Come on, let's do it.
  10. Guides updated for 3DCoat 2023, to avoid confusion and mentioning differences in the process since 2023 (two executables, none sharing the same name with prior versions of 3DCoat). Happy 3DCoating!
  11. So Mac and Linux won't receive any video/documentation, some procedures to test? I may have missed it, admittedly. But the omission on the official YouTube channel is peculiar. Also, what if Blender is installed somewhere else? Thanks!
  12. I had a suspicion that it could have to do with XDG_DATA_DIRS but couldn't spend time trying different things to see if it worked. Because possibly different environment variables are set up in different systems, that would explain the apparent randomness of the crash and also its consistency if it happens. And possibly why it doesn't if started as super user instead of plain user. I don't know when will I be able to find the time to test this, though. I just put it here in case anyone else want to give it a go, and see if we can find a way to check and correct the issue once and for all. Good luck! Edit: possibly glib version differences and or relations to XDG data variables could be at play, here. I may be wrong, though. My knowledge is very limited here.
  13. @SERGYI it may have escaped you, but it's not the first time I specifically explained the procedure to reproduce the crash in this very thread. Also I'll quote part of an e-mail I sent to you on the 27th of June, one that you kindly replied to then, so you must have read it. Let me know if you need further information. As stated, though, this won't help me anymore, but it might be of use to many others. Thanks for considering it. Quote from the e-mail: "(...) Bug reproducible steps: - Open Smart Material Editor - Try to load an image - Click on a folder in the file browser (nautilus/Gnome Files). - Instant crash: 3DCoat disappears." Quote form a forum post in this thread: Hopefully this will help you to sort it out. Have a good one and see if you can find and smash that bug!
  14. HI, good to know there's a new build for Linux. I've tested it again in both trial versions (3dcoat-Ubuntu20.04 and 3dcoat-Ubuntu22.04) and, again, if I launch 3DCoat with sudo it doesn't crash. If not lauched with sudo, when editing a material in Smart material editor and clicking to add or change an image texture, navigating the file manager crashes 3DCoat. I don't think it's related to my system, as many other applications (all of them, actually) don't have this problem. Even if this doesn't happen in Ubuntu (I don't know if it crashes in Ubuntu or not). Still a pity something like this prevent non-Ubuntu systems to work. Again, I haven't seen this anywhere else: all applications I use, including 3D/2D DDC software, both commercial and free applications, work on my machine, even if they don't officially support my Linux distribution. On my second workstation, the same is true. By the way, in case it could help: if opening the file manager to load a saved project (File>Open in 3DCoat) it works just fine and it doesn't crash. My license has expired so I can only test it with the trial (my last build working with my just expired license is April's 2023.17), but that shouldn't be an issue either, I assume this part of the program should work the same regardless of licensing status. I don't know what the problem may be here, and my limited programming skills and the time dedicated to find a possible root cause for this have not allowed me to find a viable solution, besides having to run 3DCoat as root, or super-user, which is far from ideal (albeit it then works). But then I don't know if this opens any more issues when trying to use the link to Blender, with file permissions, configuration files and folders... I just didn't test that. Others with newer current licenses may be able to check that, surely. I did my best to help here, but I'm at a loss. I won't change my system just for 3DCoat for now, but others may find the solution to the issue, if developers' constrains prevent them to allocate resources to fix this particular problem. I wish you all the best! Anyway, thank you for your efforts and openness to feedback, @SERGYI
  15. Thanks @Carlosan, the issue is known for a while, though. I already e-mailed him and got an actual direct answer from him. In all my interactions he's been always supportive and keen to explain what was needed. I'm not complaining, really. And we've moved far from the early days regarding building and publishing Linux versions of 3DCoat, yes. For the better, of course! Their situation must be really tough, and their country may need them too. So they have to split their time between family and friends, country and this product. I'm in awe that they manage to produce anything in this conditions, and admire them for what they are accomplishing. Not complaining. Months ago I even offered to send them hardware to code and help with logistics if needed. It was not deemed necessary at that point in time. So not complaining and I try my best to be constructive. The thing is, though, that my professional needs are not met here. All in all, I do understand that Linux must be at the bottom of their list (smaller user base, alone, would make that choice unavoidable, besides hiring more coders). And as soon as they can we do get builds. And that's fantastic. When something that requires lots of changes or more structural code, they have to adapt it to Mac, so it understandably takes time. Later, the same for Linux since it's the same guy doing both, and other tasks too. So from time to time, we get behind. But meanwhile that happens there are no more bug fixes for weeks, for months. And yes, once the code is there, he updates builds regularly, so when they can, they really do produce the builds within days, even hours. So I know they care, I know they listen, react, adapt. But I also know that during the year, there are gaps with weeks and months with no new tools (I can live with that) but, crucially, no bug fixes. And that is the problem. I can't rely on a software I like done by people I admire because if I get stuck because of a bug or report one, I don't know if I'll get a fix for it in a reasonable time. So I do understand they have limited resources, they do care and they are doing even innovative tools with their country under attack. But I need this tool for work. And I need it for Linux. I'm just a tad disappointed because I really like 3DCoat and where is heading. So don't take me wrong, but I can't tell my clients to understand certain things. Or spending many work hours looking for workarounds. Or worrying if I'll find a bug on the next project and then get stopped again and moving to another software to finish the job. I hope things will turn around, and I wish them all the best. But I have not had bug fixes since April. And since my license is expired now I won't get them even if fixed. I got a little carried away, here? Well, sorry about that. I wish I could help the situation but I don't see how. And it's a pity. Thanks for listening, and now I leave you with your regular programs. Have a good one!
×
×
  • Create New...