98610982c9
export the final dbrush_t right in ExportBrushList_r instead of doing it ahead of time
2022-08-03 05:09:39 -04:00
58d1e5230e
remove hullbrush_t; just directly work via the bspbrush_t since we have the bevels pre-calculated in mapbrush_t
...
track the mapbrush we were created from in bspbrush_t; we use this later for outputting the collision brush for q2bsp. bevels are busted though..
2022-08-03 04:47:50 -04:00
24168c8e40
clang pass
2022-08-02 21:57:33 -04:00
5f35575c02
remove map.faces and map.brushes; now, brushes are stored inline in mapentity_t, and faces are stored inline in mapbrush_t
...
calculate brush bounds (and winding) and add bevels after all of the rest of the map has loaded, so that it has proper world extents
calculate world extents right after the map is loaded
2022-08-02 20:59:08 -04:00
9ca28dff61
bevel
2022-08-02 19:48:17 -04:00
e14c2772a6
add ability for winding storages to reserve memory (for clip mainly, so it doesn't need to allocate on each addition; we know the winding will have at least the input number of points)
2022-08-02 18:10:20 -04:00
4c6d952841
plane hashing
2022-08-02 17:44:14 -04:00
a75f4239e3
node_t planenum
2022-08-02 16:04:48 -04:00
46a643c3f0
face_t using planenum
2022-08-02 15:45:06 -04:00
e57aa36e2e
use stats rather than logging every time these specific instances occur
...
restore the "how close the match is" block
2022-08-02 15:18:53 -04:00
1d729ba573
side_t using planenums
2022-08-02 14:55:38 -04:00
1ad0eb530d
going back to using planenums, but going for the qbsp3-esque system
2022-08-02 14:17:59 -04:00
93655913c0
Merge branch 'brushbsp' of https://github.com/ericwa/ericw-tools into brushbsp
...
# Conflicts:
# include/qbsp/qbsp.hh
2022-08-02 06:24:46 -04:00
dd287f30d4
Merge branch 'brushbsp-midsplit' into brushbsp
2022-08-02 06:09:28 -04:00
53adb9fed5
split winding_base_t into two types, so that we can switch out the backing storage
...
default winding_t for the main polylib to use heap-based, exact-memory-usage-only, winding
2022-08-02 06:09:20 -04:00
8476e2861c
this all matches release now
2022-08-02 04:29:36 -04:00
e05a2bdf75
qbsp: "-add additional.map" feature
2022-08-01 19:13:10 -06:00
cc9dd986c6
adopt parent node bounds if a leaf is unbounded now too
...
move midsplitbrushes into same group as blocksize
always return a split rather than erroring
list out which method is used of the three
2022-08-01 15:51:57 -04:00
52dff47a86
fix linux build
2022-08-01 12:47:20 -06:00
1751733ddc
qbsp: add experimental MidSplit re-implementation, enabled by default
2022-08-01 12:30:04 -06:00
da4c45d62f
qbsp: fix crash on DM6
2022-08-01 12:13:21 -06:00
755a9e65ca
use std::set instead of vector for original_brushes since we need them implicitly sorted and no dupes
...
parallelize PruneNodes_R and CalcTreeBounds_r
2022-08-01 13:19:50 -04:00
26a18d5cc3
add logging::header/funcheader to easily and consistently print function or "active routine" headers
...
add OOP percent_clock wrapper to make printing percents easier, not used yet
2022-08-01 09:35:21 -04:00
aac9513417
qbsp: parallelize MakeTreePortals
2022-07-31 21:39:47 -06:00
f665cd5cb6
testqbsp: rename our copy of base1.map to base1-test.map
...
for easier loading in game in Q2
2022-07-31 21:29:25 -06:00
d1c3b56e1d
don't jitter sun entities
2022-07-31 09:15:51 -04:00
6a3e971a11
Merge branch 'brushbsp' of https://github.com/ericwa/ericw-tools into brushbsp
2022-07-30 06:39:20 -04:00
e60babdb9c
add q3map2-style -blocksize option, using the same algorithm from it. it's optional and disabled by default.
...
pass through the "use mid split" boolean again
remove node_t::side; appeared to be unused in our current code, and needs to be removed anyways to support the other plane splitters
re-introduce ChooseMidPlaneFromList, but comment it out as it currently fails on a lot of BSPs.
2022-07-30 06:39:14 -04:00
e1b6aba014
portals.cc: log number of tree portals
2022-07-30 00:29:54 -06:00
360daea172
fix test failure from map parsing
2022-07-29 22:41:57 -04:00
d8fd13d481
use <= gating for brightness for surface and bounce lighting, since these can often be very small values (provides a moderate speed increase as well, counter-balancing the next change)
...
bounce lighting generates a grid of bounces rather than a single light, improving lighting in conditions with larger faces or where the midpoint is blocked by other geometry
2022-07-29 22:26:07 -04:00
91124d296f
fix mismatch in surface lighting cull calculation
2022-07-29 21:29:04 -04:00
521daad432
always use world entity for looking for _wad key
2022-07-29 21:20:10 -04:00
77e5990016
set default subdivision to Q2 non-software maximum when Q2 is used (twice as large as Q1)
2022-07-29 13:42:26 -04:00
ea5e7f5874
fix reported bounce light count
2022-07-29 13:27:31 -04:00
85f25bbc78
Q2-specific fix; -1 lightmaps are always fullbright, which can cause actual full-dark areas to be full bright by mistake
2022-07-29 10:40:02 -04:00
14ac139cce
fix race condition; map operator[] modifies if it doesn't exist
2022-07-29 09:42:11 -04:00
b1b4cb0e09
try relative to current working dir if can't find archive (need to search relative to .map too)
2022-07-29 04:49:54 -04:00
3ee28fc416
fix crash in light from bounce lighting storage
...
move bounce lights to a forwardly linked list
fix minlight affecting style 0 for switchables
2022-07-29 00:41:24 -04:00
7810875860
strip trailing spaces from keys
2022-07-28 20:41:43 -04:00
dd39bf9940
better print for areaportal warnings
2022-07-28 20:41:33 -04:00
6399c743a7
fix bogus brush splits from old Q2 world extents
2022-07-28 20:00:43 -04:00
d33acbe12c
remove snapvertices cmd
2022-07-28 09:25:19 -04:00
faa4c38985
fix logging being disabled wrongly at the last steps
...
remove vertex snapping done as a post-processing step; it causes hall of mirrors and doesn't work the same way as qbsp3. need to find a better option.
2022-07-28 09:21:42 -04:00
1d8a0c1df3
clang-format pass
2022-07-28 06:14:54 -04:00
e03d99d807
fix spacing
2022-07-28 05:59:06 -04:00
d9a9046e46
fix some foot-guns with qbsp_plane_t - type is auto-set now
2022-07-28 05:51:55 -04:00
64cd39c229
add optional vertex snapping to match qbsp3
2022-07-28 03:56:12 -04:00
3fc6225fee
although it's rare, if there's a case where a zero-angle triangle is *required* to close up a face (MWT determines it's part of the optimal case), we'll just let it through
2022-07-28 01:35:12 -04:00
c9935aaa38
hook up merge and subdivide stats
2022-07-27 23:02:09 -04:00
Jonathan
Eric Wasylishen