nick black
cd5c913994
[yield] yank demo_nanosleep #1613
2021-10-22 22:00:41 -04:00
nick black
edc10fc054
[yield] pull demo_nanosleep() #1613
2021-10-22 22:00:41 -04:00
nick black
c2ed7d19fe
[yield] two threads #1613
2021-10-22 22:00:41 -04:00
nick black
d3d418bb07
[yield] split main loop out suitable as pthread #1613
2021-10-22 22:00:41 -04:00
nick black
ef2c15abf2
[xray] decode in both threads concurrently, 40% FPS improvement #1613
2021-10-22 00:06:32 -04:00
nick black
1bdeb0d258
[demo] FPS not FPSS (thought i already made this change)?
2021-10-21 21:36:41 -04:00
nick black
85b6d55577
[utils] implement notcurses_osversion #2293
2021-10-21 19:02:09 -04:00
nick black
6e2ab83a08
[utils] declare notcurses_osversion #2293
2021-10-21 18:36:52 -04:00
nick black
20fa1db5dd
define clockwait save in presence of _USE_GNU #2291
2021-10-21 17:07:57 -04:00
nick black
9b30487e4a
[windows] define pthread_cond_clockwait() #2292
2021-10-21 17:07:57 -04:00
nick black
a3575ebfd8
[input] switch to pthread_cond_clockwait() #2291
2021-10-21 17:07:57 -04:00
nick black
864e2710a3
[demo] use pthread_cond_clockwait() with CLOCK_MONOTONIC #2291
2021-10-21 17:07:57 -04:00
nick black
3e0ce64b76
[input] try to use CLOCK_MONOTONIC for condvar
...
By default, Pthread condition variables use the wretched
CLOCK_REALTIME for pthread_cond_timedwait(). We reaaaaaaly
want it to use CLOCK_MONOTONIC instead. Where supported,
use pthread_condvar_setclock() to initialize the input
condvar with CLOCK_MONOTONIC, since every other operation
we perform is clocked against it. Closes #2291 .
2021-10-21 17:07:57 -04:00
nick black
a087960ab7
[notcurses-demo] write input readiness to correct pipe side #2290
2021-10-21 14:02:21 -04:00
nick black
6da04e04db
[cmake] drop -fcx-limited-ops
2021-10-21 05:11:18 -04:00
nick black
b0e4bae2fe
[cmake] can't use -ffinite-math-only with libav
2021-10-21 04:48:31 -04:00
nick black
398266d10c
[make_sprixel_plane] properly set outx following scale #2278
2021-10-21 03:54:54 -04:00
nick black
d6e574f131
[input] on construction failure, return NULL initdata #2289
2021-10-21 03:16:53 -04:00
nick black
7071cb42fa
[cmake] add USE_ASAN option to build with addresssanitizer
2021-10-20 12:03:08 -04:00
nick black
06de7ac582
[init_banner] no extra space when we don't know a version
2021-10-20 09:53:06 -04:00
nick black
86b81dfbd1
[sixel_refresh] bounds-check when damaging render state #2287
2021-10-20 09:41:24 -04:00
nick black
2b86835faa
[ncplane_erase] call destroy_tam() #2286
2021-10-20 09:31:44 -04:00
nick black
7be722edb3
[ncplane_reparent_family] resolve use-after-free on error path #1348
2021-10-20 07:06:51 -04:00
nick black
854003208e
README: unicode 14 full emoji list
2021-10-20 05:19:30 -04:00
nick black
1e4ce97750
[banner] display OS in init banner
2021-10-20 04:49:08 -04:00
nick black
6411b628a4
[utf8_egc_len] don't disqualify output based off wcwidth() #2285
2021-10-20 03:20:55 -05:00
nick black
041cb1de80
some iTerm2 notes
2021-10-20 02:59:19 -05:00
nick black
f002cd3e4c
mention mintty-direct and iterm2 terminfo keys
2021-10-20 03:49:09 -04:00
nick black
9483278ba1
dump sprixel_auxiliary_vector, save 50% of kitty/fbcon auxvec space
2021-10-20 02:09:53 -04:00
nick black
d857fedf16
[tester] use NCLOGLEVEL_INFO for initial setup
2021-10-20 01:29:07 -04:00
nick black
2f2c42d311
[CMake] throw -ffast-math into the mix
2021-10-20 00:53:08 -04:00
nick black
089bea99b6
[CMake] remove duplicated cmake_minimum_required()
2021-10-20 00:36:11 -04:00
nick black
2c89553437
[bitmaps] kill memory leak on resize failure path
2021-10-19 20:23:22 -04:00
nick black
5ce9f57540
[plane_blit_sixel]: call destroy_tam()
2021-10-19 19:54:49 -04:00
nick black
18425cd426
[ncplane_reparent_family] don't use invalid pile reference #1348
2021-10-19 19:41:35 -04:00
nick black
a4896720e7
[keller] if we expect success, but fail, error out #2278
2021-10-19 18:05:46 -04:00
nick black
3dafea8e91
add notcurses_canpixel()
2021-10-19 18:04:58 -04:00
nick black
bfdd2ed347
[linux] %hu -> %u for new ioctl
2021-10-19 14:58:39 -04:00
nick black
61b149b357
[boxdemo PoC] properly recover from error #1348
2021-10-19 13:49:45 -04:00
nick black
a12d7a8d15
[base64_rgba3] make pixel assignments a bit clearer
2021-10-19 13:42:04 -04:00
nick black
fa3c872ae6
[grid PoC] free rgba after loading visual #1348
2021-10-18 19:31:44 -04:00
nick black
51043fc959
[putenv_term] don't dereference NULL when lacking TERM variable #1348
2021-10-18 19:29:31 -04:00
nick black
5010911a08
[tabbed PoC] kill dead stores #1348
2021-10-18 19:28:14 -04:00
nick black
b544364ab4
[sprixel] void* for auxvector
2021-10-18 19:11:51 -04:00
nick black
c2fca2392b
[compare_versions] work on empty version string
2021-10-18 19:11:35 -04:00
nick black
f71a0617f7
[fbcon_rebuild] kill dead store
2021-10-18 14:52:57 -04:00
nick black
6d31d6a89b
[automaton] refresh pointers after realloc #2279
...
Wherever we realloc() the escnode array --
create_esctrie_node() mercifully being the sole
place this goes down -- it invalidates all existing
pointers into the array. Properly rematerialize
invalidated pointers by preserving indices around a
creation call. Closes #2279 . Thanks, scan-build!
2021-10-18 14:48:52 -04:00
nick black
f4ca9abc9d
[reel PoC] ignore EvType::Release events
2021-10-17 18:42:12 -04:00
nick black
9097568ac0
document NOTCURSES_LOGLEVEL #2277
2021-10-17 17:28:35 -04:00
nick black
27917d12ec
add set_loglevel_from_env() #2277
2021-10-17 17:12:52 -04:00
