Recherche avancée

Sur d’autres sites (7163)

• The complete guide on tracking your websites and web apps into multiple Piwiks and how to do it easily & efficiently

  23 février 2017, par InnoCraft — Community, Development

  Getting the tracking of your website and apps right is crucial to your success as you need to ensure the measured data is meaningful and correct. That’s why we, at InnoCraft, help our clients setting up their web tracking and digital measurement strategy. Some challenges include tracking your analytics data into multiple Piwik services as well as the tracking of single-page websites and web applications (covered in a previous article). In this blog post, we explain how to track your data into multiple Piwik websites correctly.

  Embedding the tracking code

  First of all you need to embed your JavaScript tracking code into your website or app as usual. If you haven’t done this yet : Log in to your Piwik, click on “Administration” in the top right and go to “Tracking Code”. There you have various options to adjust your tracking code to your needs.

  Tracking the same data into different websites

  Let’s assume you have set up the regular JavaScript tracking code and you want to track the same data into a second Piwik website. This second Piwik website can be either on the same Piwik installation or on a different Piwik. To do this, add the following line to your tracking code :

  _paq.push(['addTracker', 'https://$yourPiwikDomain/piwik.php', idSite]);

  It should look like this :

  var u = '//$yourPiwikDomain';
  
  _paq.push(['addTracker', u + '/piwik.php', var idSite = 2]); // adds an additional tracker
  
  _paq.push(['setSiteId', '1']); // configures your regular Piwik tracker
  
  _paq.push(['setTrackerUrl', u + 'piwik.php']);

  This will track the same data into website 1 and website 2 of your Piwik installation. You can also change the domain in addTracker to point it to a different Piwik installation :

  _paq.push(['addTracker', '//$differentPiwikDomain/piwik.php', var idSite = 2]);

  All Piwik tracker methods that you call afterwards will be applied to all trackers. Say you call _paq.push(['disableCookies']); _paq.push(['trackPageView']);, then both methods will be called on all tracker instances assuring they will behave the same and will track the same data into both Piwik websites.

  Tracking different data into different websites

  If you want to track only certain data into one website, and different data into an additional website, you need to configure the trackers differently. For example, you want to enable link tracking only for one tracker, but not for the other. The problem is that calling _paq.push(['enableLinkTracking']); enables link tracking on all of your trackers. To workaround this limitation, you can configure your trackers differently like this :

  window.piwikAsyncInit = function () {
  
      Piwik.on('TrackerSetup', function (tracker) {
  
        if (tracker.getSiteId() == 2 
  
           || tracker.getTrackerUrl() === '//$yourPiwikDomain/piwik.php') {
  
            tracker.enableLinkTracking();
  
          }
  
      });
  
  };

  Now it enables link tracking only for the tracker that is configured for a certain website ID or Piwik domain.

  Accessing a previously generated tracker instance

  When you configure a tracker via _paq.push, you create a so called “Async tracker” because Piwik will be loaded asynchronously and create the tracker instance as soon as it is loaded. If you need to get the instance of such a tracker, you can use the method Piwik.getAsyncTracker(trackerUrl, idSite). This can be useful if you have a single-page website and want to track different data into different websites :

  window.addEventListener('hashchange', function() {
  
      if ('undefined' === typeof Piwik) {
  
          // Piwik might not be loaded yet
  
          return;
  
      }
  
      var tracker1 = Piwik.getAsyncTracker('//$yourPiwikDomain/piwik.php', var idSite = 1);
  
      var tracker2 = Piwik.getAsyncTracker('//$yourPiwikDomain/piwik.php', var idSite = 2);
  
      tracker1.setCurrentUrl('/' + window.location.hash.substr(1));
  
      tracker2.setCurrentUrl('/mywebsite/' + window.location.hash.substr(1));
  
  });

  Tracking different data into multiple Piwik installations without using “_paq”

  Some users prefer to not use _paq.push and instead directly create tracker instances themselves using the method Piwik.getTracker(trackerUrl, idSite) like this :

  window.piwikAsyncInit = function () {
  
      var tracker1 = Piwik.getTracker('//$yourPiwikdomain/piwik.php', var idSite = 1);
  
      tracker1.disableCookies();
  
      var tracker2 = Piwik.getTracker('//$yourPiwikdomain/piwik.php', var idSite = 2);
  
      tracker2.enableLinkTracking();
  
  
  
      tracker1.trackPageView();
  
      tracker2.trackPageView();
  
  };

  We usually don’t recommend creating trackers manually as it is more complicated and you need to make sure to configure trackers in the right order. For example to prevent the setting of any cookies, it is recommended to call disableCookies before calling any other methods. If you want to create your trackers manually and you use any of the following methods, make sure to call them in this order :

  disableCookies(), setAPIUrl(), enableCrossDomainLinking(), setCookiePath(), setCookieDomain(), setDomains(), setUserId(), enableLinkTracking()

  Roll-Up Reporting – the easy and efficient way

  Often users track data into multiple websites because they need aggregated data over all their websites. They want to see all statistics for a single website, but also which pages were viewed across all their websites, or how much traffic they got from a specific website or search engine across all websites. This means they add a second tracker to all their websites and track data not only into the regular Piwik website, but also into one additional website that gives them statistics over all websites. This has several disadvantages :

  • Complexity in getting the tracking code right and the time needed to integrate and maintain it
  • Slower website performance because everything needs to be tracked into several websites. This can decrease your conversions and sales
  • Slower Piwik performance because it has to handle twice as much traffic. This means tracking becomes slower, generating the report becomes slower, and the database gets twice as big

  Luckily, there is a better solution called Roll-Up Reporting. With Roll-Up Reporting, you can get aggregated data over all websites and / or for a group of websites without any of these disadvantages. It lets you create as many Roll-Ups as you wish and you can choose which websites’ data should be aggregated together into a new website.

  We had customers who were able to remove one Piwik tracker because of this feature which resulted in less server costs, a faster website, and a faster Piwik. On top of all these advantages, it also lets you view the Visitor Log, Real-time Map, and other widgets and reports across several websites.

  Questions ?

  If you got any questions, please let us know and get in touch with us. You can find more information about the Piwik JavaScript tracker on the Piwik Developer Zone. There is a section dedicated to Multiple Piwik Trackers.

• Install ffmpeg on Centos6 x64 using FFmpegInstaller 8.0 Issue [on hold]

  21 mars 2017, par woshka

  I have the following issue while installing ffmpeginstaller8.0 on centos 6 x64
  on the mplayer installtion it fails as followes on cpu.c
  what is the issue and how to solve it ?

  Thanks

  EDIT : I have found the best solution to install ffmpeg and all of it’s libraries with this link enter link description here

  YASM    libavcodec/x86/vp9lpf_16bpp.o
  
  YASM    libavcodec/x86/vp9itxfm_16bpp.o
  
  YASM    libavcodec/x86/vp9lpf.o
  
  YASM    libavcodec/x86/hevc_mc.o
  
  YASM    libavcodec/x86/vp9itxfm.o
  
  AR      libavcodec/libavcodec.a
  
  make[1]: Leaving directory `/usr/src/ffmpegscript/mplayer/ffmpeg'
  
  make -C ffmpeg libavutil/libavutil.a
  
  make[1]: Entering directory `/usr/src/ffmpegscript/mplayer/ffmpeg'
  
  CC      libavutil/cpu.o
  
  libavutil/cpu.c:20:23: warning: stdatomic.h: No such file or directory
  
  libavutil/cpu.c:28:5: warning: "HAVE_SCHED_GETAFFINITY" is not defined
  
  libavutil/cpu.c:34:5: warning: "HAVE_GETPROCESSAFFINITYMASK" is not defined
  
  libavutil/cpu.c:34:36: warning: "HAVE_WINRT" is not defined
  
  libavutil/cpu.c:37:5: warning: "HAVE_SYSCTL" is not defined
  
  libavutil/cpu.c:48: error: expected '=', ',', ';', 'asm' or '__attribute__' before 'cpu_flags'
  
  libavutil/cpu.c: In function 'av_force_cpu_flags':
  
  libavutil/cpu.c:86: error: implicit declaration of function 'atomic_store_explicit'
  
  libavutil/cpu.c:86: error: 'cpu_flags' undeclared (first use in this function)
  
  libavutil/cpu.c:86: error: (Each undeclared identifier is reported only once
  
  libavutil/cpu.c:86: error: for each function it appears in.)
  
  libavutil/cpu.c:86: error: 'memory_order_relaxed' undeclared (first use in this function)
  
  libavutil/cpu.c: In function 'av_get_cpu_flags':
  
  libavutil/cpu.c:91: error: implicit declaration of function 'atomic_load_explicit'
  
  libavutil/cpu.c:91: error: 'cpu_flags' undeclared (first use in this function)
  
  libavutil/cpu.c:91: error: 'memory_order_relaxed' undeclared (first use in this function)
  
  libavutil/cpu.c: In function 'av_set_cpu_flags_mask':
  
  libavutil/cpu.c:101: error: 'cpu_flags' undeclared (first use in this function)
  
  libavutil/cpu.c:102: error: 'memory_order_relaxed' undeclared (first use in this function)
  
  libavutil/cpu.c:265:5: warning: "HAVE_WINRT" is not defined
  
  libavutil/cpu.c:268:5: warning: "HAVE_SCHED_GETAFFINITY" is not defined
  
  libavutil/cpu.c:275:7: warning: "HAVE_GETPROCESSAFFINITYMASK" is not defined
  
  libavutil/cpu.c:279:7: warning: "HAVE_SYSCTL" is not defined
  
  libavutil/cpu.c:285:7: warning: "HAVE_SYSCONF" is not defined
  
  libavutil/cpu.c:287:7: warning: "HAVE_SYSCONF" is not defined
  
  libavutil/cpu.c:289:7: warning: "HAVE_WINRT" is not defined
  
  make[1]: *** [libavutil/cpu.o] Error 1
  
  make[1]: Leaving directory `/usr/src/ffmpegscript/mplayer/ffmpeg'
  
  make: *** [ffmpeg/libavutil/libavutil.a] Error 2
  
  cp: cannot create regular file `/usr/local/cpffmpeg/etc/mplayer/codecs.conf': No such file or directory
  
  Installation of mplayer.tar.gz ....... Completed
  
  
  
  
  
     Mplayer installation Failed :( ,  please contact  professional support sales@syslint.com

• Calling pkill to pause a process with subprocess inside a class in python suspends the python script

  31 mars 2017, par muammar

  I have a class that contains a "blocking method". Inside that method, I am listening to keys to perform some actions. The action in question is pkill. Below I am pasting it :

  def show_control(self, control):
  
      """docstring for control"""                                                                                                                           
  
      if control == True:        
  
          from mkchromecast.getch import getch, pause
  
  
  
          self.controls_msg()    
  
          try:                   
  
              while(True):       
  
                  key = getch()  
  
                  if(key == 'p'):     
  
                      if self.videoarg == True:
  
                          print('Pausing Casting Process...')
  
                          subprocess.call(['pkill', '-STOP', '-f', 'ffmpeg'])
  
                      ...

  It turns out that the ffmpeg process is paused, but the python script gets suspended ?. I don’t understand why that is the case. If one creates the same function in a regular script (not inside a class to be clearer) this does not happen. I have tried using multithreading and multiprocessing modules without success. What am I doing wrong ?. Thanks.