T-Junctions
- Thread starter MegapiemanPHD
- Start date
They're brushes that have been converted to models, haven't you been reading the thread?
Why would you want my wooden scaffolding model, when you could make your own version that has exactly the right shape you want? No one wants to build around my weirdly specific props. The only ones you might want are the three sizes of pillar since they're not actually fitted to a specific area. But then the same argument applies, you can make an exact copy out of brushes, so why bother?
Why would you want my wooden scaffolding model, when you could make your own version that has exactly the right shape you want? No one wants to build around my weirdly specific props. The only ones you might want are the three sizes of pillar since they're not actually fitted to a specific area. But then the same argument applies, you can make an exact copy out of brushes, so why bother?
Because Hydro is a launch map with virtually no detailing. 
At least compared to the amount of bling we're putting in maps today.
At least compared to the amount of bling we're putting in maps today.
-notjunc
works as advertised, it skips tjuncs. irregularities may happen, and considering you probably have a large map, they will happen.
if I may give an advice, before trying to fix tjunc related problems, always ask yourself "are my tjunk errors a side effect of a convoluted layout?". the anwser is yes in many cases. most of the time you can achieve the exact same results with a more straight forward layout.
Last edited:
I am curious what you mean by convoluted layout and the correlation with t-junctions. It sounds like you're suggesting that simpler layouts can produce less t-junctions with the same amount of actual detailing. Which intrigues me.
And speaking of alternate ways to improve t-junctions, is it possible that tricks like using 45 degree joins on corners and such, particularly for func_detail buildings, might reduce them?
And speaking of alternate ways to improve t-junctions, is it possible that tricks like using 45 degree joins on corners and such, particularly for func_detail buildings, might reduce them?
well, if you have height variations forcing you to have stairs, you can question whether or not a height variation achieves anything. in the case where the variation does absolutely nothing, remove it if possible. (height variations are not strictly good, who would have tought!)
you can have a couple separated rooms and figure out that you could instead have a single room that achieves the same results. then, a single beam will contribute to the detailing of what was once 3 different rooms and you'll find that you probably need less beams etc...
on top of that, visleaf divisions increase your waterindice. so yes, a straightforward layout should in theory help against t-juncs
you can have a couple separated rooms and figure out that you could instead have a single room that achieves the same results. then, a single beam will contribute to the detailing of what was once 3 different rooms and you'll find that you probably need less beams etc...
on top of that, visleaf divisions increase your waterindice. so yes, a straightforward layout should in theory help against t-juncs
Last edited:
Ah. Well that explains Hydro then; it's mostly huge open areas. If only there were a way to disable automatic vis cuts that occur at certain intervals, you could probably get away with even more still!
Look at the -block and -blocks options for vbsp.exe, I think that might influence the visleaf creation.