Jump to content

R-T-B

Members
  • Posts

    2,019
  • Joined

  • Last visited

Everything posted by R-T-B

  1. That might work. Thanks for the idea! I'll try it soon here... would be nice to get another stab at multistar out but first I've got to do *sighs* real life work... lol. Thanks for getting back to me, I was stumped and didn't know what part was making the ECs. If it's just PostCaclulateTracking, I think we can work around this.
  2. I'd advise the same. And yeah, KopernicusExpansion only works in a limited sense in 1.9+, see this for the working bits; If after further testing it is indeed affecting in game clickable menus, I'll add a bug report, of course. It's probably stemming from the same issue either way, but I hope it's not that widespread.
  3. My branch fixes this. https://github.com/R-T-B/EnvironmentalVisualEnhancements I really didn't want it to come to this, but... This mod is quintessential to the KSP ecosystem, and is aparently abandoned other than @BIOZ's latest work. I assume your work is open domain, @BIOZ? If so, I will be integrating it into my tree and excercising a community takeover of eve, since my build has been the only one working right (right being mostly bugfree) for quite some time. I can open a new thread if desirable. There'll be a proper 1.10.x release today. @BIOZ, I can add you as a github contributor if you'd like, given you are obviously capable (and I'm overtaxed with Kopernicus as is). Anyways, until I have time to make a thread, get releases here: https://github.com/R-T-B/EnvironmentalVisualEnhancements/releases There'll be a true 1.10.1 ready one in a few minutes. (with @BIOZ's fixes, if he objects I will take down). I also integrated the cloud fixes mentioned above that i've had fixed for months now. PS: This project is a lower priority on my totem pole than Kopernicus, and as such, I'd really like someone to consider adopting it other than me. I'm just doing what must be done for now.
  4. I believe the last stable release on this branch may have fixed the issue with PQScities (monoliths, random discoverable things) "sticking" to the body from which they were templated. If anyone would care to confirm in a game test, that'd be great, though. That would move us basically to CKAN status, pending one more bugfix I think I can get done today. You are probably going to ask how to test. The easiest way I think would be to look for the "Mohole" on a body templated from Moho. It shouldn't be there. I'd do this myself, but ironically, despite being a bigtime Kopernicus dev now, I don't want to teleport to Moho or a body templated off of it as i haven't actually MADE it to Moho in my game yet. Yeah yeah, I know.
  5. Yeah, that's what I'm finding today. The code is in fact rerunning and removing the PQS that is the main menu, again. Which is what it should do, but it doesn't help us right here and now. We need to tell it to leave the main menu alone, lol. Whoever at squad thought the main menu should be a PQS is sorta... I mean if they were going to do that, why not from day 1? Why wait for 1.10.1? Sometimes their choices confound me. But I'm working on it.
  6. Heh, I didn't read it that way. If that's what you feel he's doing though, that is completely understandable. I've always modded for my own game first, others second. I imagine you are the same. I would ask you to remember linux-culture can just get a bit silly around windows sometimes (bite your tongue, steve_v, lol), and I think he was just trying to see if there was a bug to report. But beyond that I'll shut up.
  7. Proton is just steams wrapper for the tools i suggested (dxvk and wine, I think it adds esync too). At any rate, glad you do not carry bias. I found it hard to believe you would, considering the level of effort you put into backwards compatibility with the whole dual camera thing! I mean, scatterer is a graphics mod. Even on DX11 it hurts performance, that's just it's nature. It gets worse with dual camera though, and that's largely squads (or maybe even Unity's) fault because they have not implemented single camers in OpenGL codepath yet. I hear you. For the end user, this is the same ol "but it don't work good!" story. Sadly, not much we can do right now.
  8. I could do with better phrasing, but yeah, it might not be a bad idea to warn about the performance issues. It's tough for me because I sympathize with blackrack too, OpenGL is an utter mess right now on KSP. And it's not really the API's fault as much as squads and to some extent, Unity, but meh... End user hurts the same, you know? Ironically, navi's done better for me in linux than windows. In windows their driver is an unstable mess. lollerskates.
  9. It's more that AMD is woefully incompetent at linux and finally admitted it and just open sourced their driver. The linux community had it outperforming the windows binary drivers at OpenGL within a year. But that's OT, really. Sorry I can't be of more help. For your needs, you may try EVE atmospheres, for a basic (very basic) atmospheric effect. I think Spectra implements them. EVE works in 1.9 with my builds, I've heard. I'm no longer maintaining them since I joined Kopernicus, but you can still fetch them here: http://glacialsoftware.net/scattererBuilds/PATCHED_EVE_BUILDS/ Source is also on github, and in archive IIRC. Same license as original EVE. I don't blame you, but that 5FPS penalty every scene change sounds like a bug of some kind that maybe you could report and I'd at least take a brief look at it if not blackrack. Do you have any interesting log files from around then? If it's an easy fix I doubt blackrack would say no to a pull request. Granted, the CPU overhead may make DXVK a nogo anyways.
  10. I would actually advise giving dxvk and wine a try. It performs... pretty damn well with it's DX11 layer. Provided NVIDIAs vulkan driver is ok (it works by translating DX11 calls to vulkan ones). PM me if you need help with that. The results are surprisingly good. While I don't agree with @blackrack's linux/OpenGL general dislike, I do agree that there's not much he can do to fix it. The OpenGL implementation in Kerbal at the moment has been badly neglected in the stock game. There isn't much modders can do about that. Ie, it runs like crap on Windows, too. The only driver configuration that seems to run KSP OpenGL mode acceptably is AMD's full open source stack, and even it is questionable and riding on the advancements of open source rather than the cards themselves. For now, use dxvk, extract 90% of native performance, and blame squad.
  11. Thanks. Tomorrow that's one of the bugs I intend to take a hard look at. All the evidence helps. Most likely. Multistar is a nonfunctional mess right now, sadly.
  12. So in other words, testing/confirming functional status will be a major pain. Gotcha. Take your time. We all want that product but nothing happens overnight that's any good.
  13. This is what we advise, actually. The bleeding edge is more for users who just don't care if something breaks here and there, and modders who want a snapshot of what is to come. Don't make main releases off of it really.
  14. Indeed, I explained the crux of the problem above. If you need more details, let me know:
  15. Thanks for the report. Getting over some minor sickness (not COVID-19, thank god), but I'm betting @Vabien is right that it's a formatting issue. I'll give you more advice shortly. Just starting to function again and doing some chores.
  16. Yeah. We version lock for a reason, it was kind of bad we ever did not. Really bad things can happen longterm if built against the wrong files. Mind you, for Kittopiatech and not a real save game, you may not care. But just sayin'. If there's enough demand, I can look into the Kittopiatech bugs as soon as we hit the major blockers for 1.9.1 stable release (CKAN). It shouldn't be super hard to make it work (hope this isn't famous last words). Yeah, I'm amazed no one reported this in 1.8/1.9 era of builds honestly. It's getting better, not all templated PQS's "stick" now, and it should be fixed entirely soon.
  17. Ooh, interesting! Thanks, that's more evidence towards what I think is going on, maybe it'll help with a fix soon... not just for the core main menu bug, but the issue with templated bodies. It seems at some point, it just stops trying to detemplate certaib things, hence the main menu starts working again as a side effect (as strange as it is, the main menu in 1.10.1 seems to be a PQS of Kerbin). Need to determine why. Thank you for pointing this out. I think it's a critical piece of the puzzle.
  18. Time to bump stable branch from oblivion, we have some new stable code to push out in a release, at long last! https://github.com/prestja/Kopernicus/releases
  19. Personally, if I just don't have to recode Kopernicus for it, I'll take it as a win.
  20. I think I know what's going on there actually, it's an old debug logger running out of control. I apologize for leaving that in there. Unfortunately, there isn't much you can do right now but wait for a bug fix, but since you've reported it, I am now aware of it and will start work on fixing it. It should be fixed sometime tomorrow when I make my next release. You won't have to wait long. The log entries shouldn't be hurting performance or doing anything bad, but they are more than a little spammy and do need fixing, I agree. Next release will have that fixed. Scratch that. The debug logger isn't the source of these, thus that won't be an easy fix as I thought. I have no idea where there are coming from, honestly, as that string isn't even used in Kopernicus (I searched for literally just "doesn't have" ). Maybe, it's just from the stock game when you load 3rd party assets for all I know? Does this occur with any planet pack or just JNSQ? I know there used to be a way to turn off the main log file but I don't know if that still works. I think it used the old game launcher found in the games folder. Yeah, as a former HDD user, I can feel your pain here. I'll keep looking into this.
  21. Good to hear. So that leaves determining if templates are working right on a fresh load. I'll let some other users test that though... I suspect they mostly are. Thanks for the help!
  22. That's fine. If BE6 continues to work fine for you do report back, I'll be able to run a compare and see what changed/broke between then and there. Probably some residual multistar gunk, if I had to guess. We messed with that a lot. I hate to say it, but I kind of feel like I may have actually fixed this bug but your save game is just got a inflight cache of a lot of errenous data. If that's the case, reverting is really the only true fix, or simply ignoring it and traveling to a new body (this would flush the cache, theoretically). Thank you for playing guinea pig anyways. Maybe make a save game backup next time if you care about the game though? I know I would if you put that much effort into a probe.
  23. Yeah, it's also hard to determine how much is save-game caching going on and how much is truly there and generated fresh. I need a deterministic test I can do to go to a templated body and look for an artifact with a simple teleport, but I've yet to come up with one. If anyone could name some teleportation coordinates in a pack, I'd be most grateful. Right now, given I am running JNSQ, my best bet is probably the Mohole on Aden.
  24. Hmmm... I'll need to take a closer look at this tomorrow. Darn. So much for simple fixes, lol. At least it's better than it was, if not fixed.
  25. Lol that'd be awesome if it wasn't a bug. Looks like a cool fog effect.
×
×
  • Create New...