nick black
22461c7f5d
[drone] update gentoo to dankamongmen/gentoo:2025-01-02a
2025-01-02 13:46:06 -05:00
nick black
3a43881b01
only use _SC_SIGSTKSZ when it's defined
2025-01-02 11:49:24 -05:00
nick black
4adde8f63d
comment out gentoo drone until it's updated
2025-01-02 10:48:50 -05:00
nick black
287aa4dd58
[signals] don't free alternate stack until late
...
musl uses any registered signal stack for its thread
cancellation logic. since alternate stacks are a
per-thread deal, but signal handlers are per-process,
threads other than the main thread (having not disabled
their (shared) alternate signal stack) could get
whalloped by musl's "SIGCANCEL". ought close #2828 .
2025-01-02 08:47:08 -05:00
nick black
a8e797169a
blit.c: drop messy assert()
2025-01-02 06:38:17 -05:00
nick black
90bf8ba847
setup_signals: use sysconf(_SC_SIGSTKSZ) when available
2025-01-02 03:10:50 -05:00
nick black
ee50c6f006
ensure we're allocating sufficient memory for sigaltstack() #2828
2025-01-02 02:56:55 -05:00
nick black
c6b13e08a3
throw a diagnostic on sigaltstack() failure #2828
2025-01-02 02:10:50 -05:00
nick black
a924cede7f
ffmpeg: relax version check for AVFrame->duration #2828
2025-01-02 00:29:48 -05:00
nick black
544d4fccc3
cmake: use find_library for deflate, but pkg-style identifiers
2024-12-31 17:10:52 -05:00
nick black
f2545423a0
drone: use noble:2024-12-31b
2024-12-31 16:28:34 -05:00
nick black
69e75050fa
work around find_library
2024-12-31 15:34:02 -05:00
nick black
13db632d6a
drone: update alpine builder to 2024-12-31a
2024-12-31 15:26:42 -05:00
nick black
3a7b9d4149
apparently we *cannot* rely on pkgconf support from libdeflate ugh
2024-12-31 15:24:36 -05:00
nick black
304938e98f
drone: upgrade ubuntu builder to noble:2024-12-31a
2024-12-31 15:18:24 -05:00
nick black
4e42f9b5ed
ubuntu drone: update to 2024-12-31a from 2021-11-14b
2024-12-31 13:47:06 -05:00
CJ van den Berg
a7b64f6210
Workaround kitty F3 issue
2024-12-31 07:02:17 -05:00
Stephen Hurd
4526270689
Be more specific in SyncTERM DA response.
...
SyncTERM has constant values after CSI =, so may as well avoid false
matches.
2024-12-31 04:53:18 -05:00
Stephen Hurd
b8bdc445c6
Enable detecting SyncTERM DA1 response.
...
SyncTERM sends CSI = Pn c not CSI ? Pn c like VT terminals.
This results in notcurses hanging in init.
This required WezTerm da3 support to be removed, because the SyncTERM
and WezTerm responses could not both exist in the automaton. As it
turns out, a dig through the WezTerm git history shows that it never
actually sent CSI = ... ST in response to DA3, and notcurses took
no actions in response to it, so it should be safe to remove this
feature.
2024-12-31 04:53:18 -05:00
nick black
9631a79ebf
build without USE_DEFLATE on mac until they install pc file
2024-12-31 04:47:41 -05:00
nick black
3ec9983f02
add FIXME about ffmpeg linking issues with static notcurses
2024-12-31 04:47:41 -05:00
nick black
0fda32f982
cmake: add USE_STATIC_BINARIES option
...
based off @data-man's work in #2664 , add a
USE_STATIC_BINARIES option to CMake, dependent on
USE_STATIC being set. when this option is used,
link the notcurses binaries against the static
versions of notcurses. i've done ncman and nctetris
thus far. also, move to pkgconf for libdeflate,
requiring at least version 1.9 (that's when i added
the pkgconfig code upstream).
2024-12-31 04:47:41 -05:00
Dmitry Atamanov
11b45d4c27
Add USE_TFMAN_STATIC option
2024-12-31 04:47:41 -05:00
nick black
83e6a0a43e
bring up to speed with current master
2024-12-29 06:08:25 -05:00
nick black
39700bf64a
correct order of parameters to calloc() from #2768
2024-12-29 05:29:23 -05:00
Nat!
bf5e74c872
use better overflow check
2024-12-29 04:36:13 -05:00
nick black
b39785493e
clean up octitions
2024-12-28 14:58:59 -05:00
nick black
5e2dd99019
eliminate rgbas[] VLA, restore -Wvla
2024-12-28 14:58:59 -05:00
nick black
9f770c43fb
undo debugging
2024-12-28 14:58:59 -05:00
nick black
0f2e0d0a6b
hires: drop duplicate egc sets from sex/oct. compute partition counts
2024-12-28 14:58:59 -05:00
nick black
c6bd61495c
kill obsolete FIXME
2024-12-28 14:58:59 -05:00
nick black
23e1851561
complete octblitter
2024-12-28 14:58:59 -05:00
nick black
db83226614
more octitioins
2024-12-28 14:58:59 -05:00
nick black
12f9c38f7e
octblitter: fix transposition for upper one quarter
2024-12-28 14:58:59 -05:00
nick black
7c0f40614b
remaining octants
2024-12-28 14:58:59 -05:00
nick black
07c0828f5b
continue building up octitions
2024-12-28 14:58:59 -05:00
nick black
f435270222
build out oct[] and octitions[]
2024-12-28 14:58:59 -05:00
nick black
1d0976b0cf
degenericize blit_4x2() -> braille_blit()
2024-12-28 14:58:59 -05:00
nick black
edb8f31aca
comment up octtrans table some more, meditating
2024-12-28 14:58:59 -05:00
nick black
55e6e6790a
genericize hires_solver() by passing in egc/partition sets
2024-12-28 14:58:59 -05:00
nick black
67a122e857
genericize hires call to cell_set_bitquadrant()
2024-12-28 14:58:59 -05:00
nick black
2d9abd34eb
finish octtrans EGC table
2024-12-28 14:58:59 -05:00
nick black
b2846f73b7
continue building up octtrans literal array
2024-12-28 14:58:59 -05:00
nick black
0308af82a7
we can just reverse oct_egcs
2024-12-28 14:58:59 -05:00
nick black
74df4d77b5
genericize hired_trans_check() on EGC set, start defining octtrans[]
2024-12-28 14:58:59 -05:00
nick black
544586b8c9
genericize geometry for hires_solver()
2024-12-28 14:58:59 -05:00
nick black
b4dd7f1a7c
make hires_trans_check() generic across cellheights
2024-12-28 14:58:59 -05:00
nick black
22f99c650f
hires_blit: genericize rgb array prep
2024-12-28 14:58:59 -05:00
nick black
e4a777f940
start in on unified hires blitter (sex/oct)
2024-12-28 14:58:59 -05:00
nick black
27b49941a4
update sexblitter partitions to match fixed sexegcs
2024-12-28 11:18:08 -05:00
