Recherche avancée

Sur d’autres sites (14260)

• Piwik is now using Github issues as our Issue Tracker !

  9 juillet 2014, par Matthieu Aubry — Community, Development

  This is an announcement regarding the Issue Tracker used for the Piwik project. We are excited to announce that Piwik has migrated from Trac to now using Github issues for managing our issues !

  More than 5,400 tickets and 20,000+ comments from 1,000+ users were migrated to Github. Read on for more information.

  Where do I find Piwik Issue Tracker ?

  • Piwik main project issue tracker (piwik/piwik)
  • Piwik Mobile issues (piwik/mobile2)
  • Device Detector issues (piwik/device-detector)
  • Piwik iOS SDK (piwik/piwik-sdk-ios)
  • See other repositories at github.com/piwik

  Benefits of using Github Issues for the Piwik project

  There are several advantages of moving to Github issues :

  • Faster and responsive user interface
  • Better cross-project referencing of issues
  • Ability to notify people with the @username functionality
  • No spam
  • Integration with Pull requests and our Git repository

  How do I get notifications for all Piwik tickets ?

  To receive notifications for new tickets or new comments in the Piwik project, go to github.com/piwik/piwik, then click the Watch button at the top of the page.

  

  In Github, watching a repository lets you follow new commits, pull requests, and issues that are created.

  How do I report a bug in Piwik ?
  See Submitting a bug report.

  How do I suggest a new feature ?
  See Submitting a feature request.

  Next steps

  At Piwik we care a lot about Data ownership. For this reason we need to have an up to date copy of all our tickets and comments out of github.com servers. Our next step will be to create and release as open source a tool to let anyone create a Mirror of their Github issues. See #5299.

  For more information about the Trac->migration, see #5273.

  We look forward to reading your issues on Github !

• Live streaming video on multiple platforms [closed]

  4 mars 2019, par Nikolay Nikolov

  I want to build an application similar to Twitch/YouTube, which mainly offers two things (and a couple of other, but they are not related to the question) - to record and send live streams and to watch other people’s live streams. Basically, if I wanted to build Twitch, where would I start in terms of protocols and back-end libraries for the processing and sending/receiving of video segments (more detailed questions follow) ? I am new to the video streaming software development and need a bit of guidance on where/how to start.

  Here are the details/requirements :

  • Video and audio
  • Scalability and low latency are more important than supreme quality
  • Adaptive bit-rate
  • No services like Wowza and such (I am willing to build the whole structure)
  • Has to work on iOS and Android (Desktop support is not as important)
  • The users should be able to watch every stream and every user should be able to stream through his camera
  • VOD is not as important
  • Going back in the stream is not as important

  If I have it right, this is how the whole process should work :

  • Android/iOS camera records video
  • Simultaneously the app saves every x seconds as a single segment and sends it to the server
  • Server processes the video in different bit rates and saves them
  • Another user requests stream based on the bandwidth of its internet connection
  • Server responds with a playlist of segments and sends each new chunk of video to the user

  Questions :

  • What protocols should I be using (HLS, MPEG-DASH, WebRTC, RTSP, etc.) and do these protocols have implementations on Android/iOS or do I have to implement them myself ?
  • What books/other resources would you recommend ?

  Thank you very much for reading my question and I look forward to reading your answers !

• Rolling screen capture with ffmpeg on windows

  11 décembre 2020, par gap210

  I have the following code to capture a video stream from my webcam. I use ffmpeg to write to named windows pipe, then read it with python and display with opencv. The thing is that the opencv stream is 'rolling' as shown here https://www.youtube.com/watch?v=H78TRo3DZIo

  


  If I capture the output to a video instead of a pipe, with the command :

  


  ffmpeg -f dshow -video_size 1920x1080 -framerate 60 -i video="USB Video" -c:v copy out.avi

  


  everything looks fine. What should I change to achieve the desired effect ? (non-rolling stream)

  


  My code below :

  


  import cv2
import time
import subprocess
import numpy as np

w, h = 800, 600

# Get frame generator
gen = ffmpegGrab()

# Get start time
start = time.time()

# Read video frames from ffmpeg in loop
nFrames = 0
cmd = 'C:/Users/......./Downloads/ffmpeg-4.3.1-2020-11-19-full_build/bin/ffmpeg.exe -f dshow -framerate 60 -video_size 800x600 -i video="USB Video" -pix_fmt bgr24 -vcodec rawvideo -f image2pipe -'

proc = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, shell=True, bufsize=10**9)
while True:
    # Read next frame from ffmpeg

    frame = proc.stdout.read(w * h * 3)
    frame = np.frombuffer(frame, dtype=np.uint8).reshape((h, w, 3))
    cv2.imshow('screenshot', frame)

    if cv2.waitKey(1) == ord("q"):
        break

    fps = nFrames / (time.time() - start)
    print(f'FPS: {fps}')

cv2.destroyAllWindows()