dnrops
/
libcimbar
mirror of https://github.com/sz3/libcimbar
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
627 Commits 38 Branches 15 Tags 12 MiB
Commit Graph

627 Commits

Author SHA1 Message Date
sz 7bb1bc8896 -DBUILD_PORTABLE_LINUX for static linking against opencv and libc++ 
This gets us back to glibc 2.27 (ubuntu 18.04) -- debian stretch would
be nicer, but the default build tools (c++1x, cmake) gets slightly more
interesting.

The output binaries seem like they could be smaller...
 2022-02-19 19:52:04 -06:00
sz3 729eb7ebcb
Merge pull request #59 from sz3/bugfix-misc 
Misc bugfixes + dependency upgrades
 2022-02-19 19:48:46 -06:00
sz 9e0dae0be8 Compile time check is fine, I guess 2022-02-18 20:11:44 -06:00
sz c05e3b5f1f hash difference aside, the binary differences are minor 
So I'm ok with cv3 and cv4 returning slightly different answers.
 2022-02-18 19:55:52 -06:00
sz e698786669 Update concurrentqueue (1.0.3) 2022-02-17 21:00:04 -06:00
sz fec665f245 Update fmt (8.1.1) 2022-02-17 20:57:30 -06:00
sz 0f2665aa60 Upgrade intx 2022-02-17 20:47:41 -06:00
sz cc6abee11a Update picosha2 and stb_image 2022-02-17 20:30:40 -06:00
sz 39b37d9e80 Update wirehair 
Including my custom CMakeLists.
 2022-02-17 19:10:51 -06:00
sz 82ed9e9643 oof 
unsigned being 32bit was doing some heavy lifting here... but let's be
explicit.
 2022-02-17 18:55:03 -06:00
sz 28559cb521 I think we need calib3d now? 
It's only needed for undistort. It be worth making optional...
 2022-02-17 18:28:41 -06:00
sz 178e02adeb Make CMakeLists smart enough to find opencv4 
since the include path is different...
 2022-02-17 18:10:54 -06:00
sz b0091d1919 new catch2 to make compiler happy 2022-02-17 18:03:44 -06:00
sz 52547aee2d Include to make g++11 happy 2022-02-17 18:03:37 -06:00
sz 7c545bb701 cerr 2022-02-17 17:55:37 -06:00
sz c5f76768c5
Bugfix (maybe): properly clean up old decodes 
This requires some more thoughts -- the shape of the problem is that the
encoder and decoder are running out of band with each other (very
intentionally -- it's the whole point). We want to "keep up" with what
the encoder throws at us, but we also need to clean up our list of
in-progress (and finished!) decodes at some point -- we don't want it to
be unbounded.

The bug this fixes should manifest as files refusing to
progress/download -- the decoder (app) has already reserved the "stream
slot", finished a download, but failed to clean up the mess to let it be
reclaimed. The workaround for CFC is to close and re-open the app, which
will throw out all in progress downloads and start from a clean slate.
 2022-02-03 21:59:09 -06:00
sz3 f752be47d6
Merge pull request #58 from sz3/service-worker 
Add service worker script for cimbar.org
 2021-06-30 23:48:39 -05:00
sz 19c7904af8
No PWA for now 
The most useful implementation (android) has some very strange
performance issues, so offline cimbar.org will be the checkpoint.
 2021-06-24 00:28:08 -05:00
sz3 ef6df875eb
Merge pull request #57 from sz3/css-fix 
Misc css changes/fixes for cimbar.js
 2021-06-20 19:48:50 -05:00
sz 8103bf0ac2
Possibly functional? 
Getting some absymal performance on the android PWA, not sure what
that's about...
 2021-06-20 19:26:29 -05:00
sz 74ceae31c9
add favicon... 2021-06-12 22:17:03 -05:00
sz 8fa1509a61
Cache more stuff? clear old caches? 2021-06-12 21:59:13 -05:00
sz e20a3cd75f
WIP? cache please? 2021-06-12 21:21:09 -05:00
sz 99eb349ac6
Merge branch 'css-fix' into sw 2021-06-12 18:47:11 -05:00
sz 8d4cbc9a1c
Try using a button element? 2021-06-12 18:46:58 -05:00
sz a7f663aee7
Set zoom on the canvas element, not the whole page? 
What we're really after is to make the canvas fill the screen on
high-dpi (phone) screens. Scaling the whole page achieves the effect,
but with some collateral damage of weird sizing issues.
 2021-06-12 03:14:13 -05:00
sz d0dffa2ade
Slightly more responsive to window size changes 2021-06-12 00:28:04 -05:00
sz 9e8e5528e0
Fix the background css 2021-06-06 23:36:32 -05:00
sz 989796bed2
service worker attempt (doesn't work) 2021-06-06 22:30:24 -05:00
sz3 b39469ce6f
Merge pull request #56 from sz3/cimbar-send-refactor2 
Remove opencv highgui code path
 2021-04-29 20:14:55 -05:00
sz 3e1e4350aa
Update dependency list to single out GLES 2021-04-29 20:04:59 -05:00
sz 2da86feb5b
Find clang 2021-04-27 20:26:05 -05:00
sz 257973d8b2
Slightly more readable, maybe? 2021-04-27 19:41:09 -05:00
sz 0a6bda1b7b
killing the opencv_highgui code path after all 
imshow() + waitKey() doesn't fit how I want to do frame pacing.
 2021-03-30 19:25:39 -05:00
sz 50d65a7eea
Fix cimbar_send encode_id. 2021-03-30 02:16:44 -05:00
sz 8216c3b811
Bugfix -- reset _frameCount when we reset the stream 2021-03-30 02:07:54 -05:00
sz 0ca7dd1e12
Rename the fountain_encoder_stream's `reset()` to `restart()` 
Since we tend to use it in a shared_ptr, it's pretty confusing to use
`reset()` as the function name.
 2021-03-30 01:50:57 -05:00
sz 22b58ff414
Switch cimbar_send to use cimbar_js 
For the moment, we'll ditch the shakycam (on) and rotate (off) options.
Might get rid of rotate altogether?
 2021-03-30 01:43:41 -05:00
sz 6b7022f546
Add header for cimbar_js, and add extra params to configure() 2021-03-30 00:54:44 -05:00
sz 71ab9eae76
Moving cimbar_js into src/lib/ 
I'd like to make cimbar_send use cimbar_js. We'll see how that
goes...

(there's an argument for removing the opencv_highgui code path here, and
I did that on an alternate branch. But I think I'll keep it for now)
 2021-03-30 00:14:51 -05:00
sz3 4415602d77
Merge pull request #55 from sz3/decode-more 
Minor tweak to the decode's preprocessing "sharpen"
 2021-03-29 17:08:48 -05:00
sz 918ffb20f7
Update build instructions to match reality 
(emcmake does things to the ENV, it's probably not what you want for
make)
 2021-03-29 16:47:08 -05:00
sz 6376e123f1 indentation 2021-03-08 01:47:56 -06:00
sz fc5ef1c1e9 Merge remote-tracking branch 'origin/master' into decode-more 2021-03-05 21:23:46 -06:00
sz 79a91b28e5 Smaller image 2021-03-05 21:17:31 -06:00
sz3 9ffb94e5cd
Merge pull request #54 from sz3/glshake 
Encoder improvements. 118 KB/s
 2021-03-05 21:12:01 -06:00
sz 61fff8fd01 I don't think I mentioned compression anywhere. 2021-03-05 07:38:29 -06:00
sz fc04ea6038 more docs? 2021-03-05 07:33:10 -06:00
sz 8e5fa906a4 Misc docs updates, including new performance numbers! 
100 KB/s is defeated
 2021-03-05 07:18:59 -06:00
sz 206c7868ee Add a few tests for the larger size 2021-03-04 21:17:06 -06:00