Shadow Hog Posted March 8, 2014 A little disclaimer: I'm touching up a gigantic SRB2 map somebody else made. Reason being, there are huge graphical glitches in some spots, and from what I'm guessing, it's because the outermost linedefs of the rooms in question, for whatever reason, don't have Impassible checked. I'm not sure how that happened - personally I make damn well sure all of my one-sided linedefs have the flag, be they for Doom or SRB2 maps - but being that I didn't make the map, I can only imagine the reason. Now, collision-wise, it's fine; an old quirk of SRB2 that made Impassible lines not-so-impassible meant that the outer wall of every stage has to be a zero-height sector instead (the community calls it a "thok barrier", but whatever, zero-height sector makes more sense in a Doom context) which players definitively can't clip through. However, I thought the renderer used Impassible one-sided linedefs to determine when it should stop rendering; is that the case, or am I mistaken and it's actually just one-sided linedefs in general, in which case I'm wasting my time? Only reason I really even suspect this is coming from when I was working on my 32in24-13 map last year; I believe a sector I was making forgot to either have Impassible checked or Double-sided unchecked, which wound up with me falling into the ground in a wide, flat area that wasn't even directly by the bugged linedef (which I obviously fixed before release, since MAP29 has no such issues in the final build). Obviously, these flags are no joke; I just need to better understand some of the more subtle quirks behind what they're doing. 0 Share this post Link to post
Kappes Buur Posted March 8, 2014 Shadow Hog said:..... I'm not sure how that happened - personally I make damn well sure all of my one-sided linedefs have the flag, be they for Doom or SRB2 maps - but being that I didn't make the map, I can only imagine the reason...... The original map was probably made with Wadauthor, where linedefs had be set explicitly to be whatever and that could easily be overlooked. 0 Share this post Link to post
Shadow Hog Posted March 8, 2014 Y'know, that's entirely possible. I know Mystic was big on WadAuthor for the longest time, while I hopped over to Doom Builder the instant it became a thing (and Doom Builder 2 thereafter). Anyway, putting the flag in all the relevant places didn't fix the issues; turns out it's ZenNode being kind of dumb. ZDBSP seems to fix it... for the split second I can see the map before the game crashes. No idea why it loads it correctly long enough to render a single screen before crashing. I think the crash is related to SEGS being too large ("SEGS is too big for vanilla Doom and some ports. (38319 segs)") and SRB2 not yet supporting ZDoom's extended NODES format, but you'd think it'd crash before it even has a chance to render, surely. Given that we want to push out the new version of SRB2 next week, it'll just go out with the graphical glitches for now. Doesn't affect collision any, so ah, well. 0 Share this post Link to post
fraggle Posted March 8, 2014 I can't think of a circumstance where you'd want a one-sided linedef that wasn't impassible. The impassible flag shouldn't affect rendering though - as far as I know it's just used by the collision code. 0 Share this post Link to post
Enjay Posted March 8, 2014 I was under the impression that there was something in the engine that forces itself to treat 1S walls as impassable regardless of the absence of the 1S flag. If I am correct, then that would be why the collision works - the engine is explicitly set up to ensure that it does. 0 Share this post Link to post
Kappes Buur Posted March 9, 2014 Enjay said:I was under the impression that ..... Well, that must be the case, because when looking at Rex's THE PHOBOS DIRECTIVE I remember that all 1s lines were without the impassable attribute and the map worked just fine. And he used Wadauthor at the time. 0 Share this post Link to post
Enjay Posted March 9, 2014 Yes, it was actually one of Rex's WA maps that led me to finding this out. I ran one through the DeePsea error checker and it flags up 1S lines that are not marked as impassable. I asked about it *somewhere* (here, ZDoom forum, not sure) and I'm pretty sure that I was told about the way the engine treats such lines as an answer. Given that the engine doesn't really care whether 1S lines are flagged impassable or not you could argue that the DeePsea error checker is being a little over zealous. Personally, I like having the lines flagged appropriately, but just for my own satisfaction. I think most editors do it by default anyway and not flagging them is more of a WA quirk than anything. 0 Share this post Link to post
Memfis Posted March 9, 2014 Hm, I've heard that when the Plutonia 2 team noticed that all maps made by Gusta (another Wadauthor user) had non-impassable one-sided linedefs, he had to fix that. So I thought there must be some difference. Or maybe they did that just to be safe. 0 Share this post Link to post
boris Posted March 9, 2014 If it really was a problem there'd be a shitload of problematic maps from the late 1990's and early 2000's. 0 Share this post Link to post