Announcement

Collapse
No announcement yet.

NSS Updated On Ubuntu 12.04/14.04 To Allow Netflix Support

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Originally posted by anda_skoa View Post
    I am pretty sure Netflix engineers know that. The problem here was that Netflix decided it would be a a good idea to have a non-technical person, most likely from the PR department, post to the Ubuntu development list.
    And that person, due to lack of understanding any of the involved technology, made it sound like the user agent check was related to the NSS update.

    Cheers,
    _
    I read the message, and there is no way to get your view of the situation out of that message.
    What the person from netflix was talking about, is unblocking LINUX once they are able to *assume* that a proper version of NSS is installed.
    The user agent hack goes along with a, typically manual, update of NSS to the required version.

    My user agent on chrome:
    Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/37.0.2062.120 Safari/537.36

    Hack user agent to enable netflix:
    Mozilla/5.0 (Windows NT 6.3; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/38.0.2114.2 Safari/537.36

    Note that there in neither mention of NSS, nor mention of WHICH DISTRO.
    It happens to be Fedora 20, which is READY to netflix out of the box (nss ver 3.17.0).... EXCEPT that they apparently are considering Ubunto to be the "consumer" distro that they need to ensure is supported.

    Once they can ASSUME NSS >= 3.16.2, then no longer need to *blacklist* the *operating system*, and then the user agent hack is no longer needed.

    Worth pointing out... "I can make a case here to lift the User-Agent filtering" -- it is not a "done deal". Just somebody AT netflix who wants to make it happen.
    Last edited by droidhacker; 09-23-2014, 09:16 AM.

    Comment


    • #32
      Originally posted by droidhacker View Post
      I read the message, and there is no way to get your view of the situation out of that message.
      What the person from netflix was talking about, is unblocking LINUX once they are able to *assume* that a proper version of NSS is installed.
      The user agent hack goes along with a, typically manual, update of NSS to the required version.
      As you wrote yourself: the user agent does not in anyway reflect which version of NSS or any other library is installed.
      Therefore the mentioning of the user agent string in the context of the NSS update was worthless.

      They could already remove it if they wanted to. It is pretty apparent that they don't want to for other reasons.

      A technical person would not have made the mistake of bringing that up, only someone clueless chasing buzzwords would.

      Cheers,
      _

      Comment


      • #33
        Here is the article that explains what Firefox is working on:

        https://hacks.mozilla.org/2014/05/re...n-and-w3c-eme/

        Comment


        • #34
          Originally posted by anda_skoa View Post
          As you wrote yourself: the user agent does not in anyway reflect which version of NSS or any other library is installed.
          Therefore the mentioning of the user agent string in the context of the NSS update was worthless.

          They could already remove it if they wanted to. It is pretty apparent that they don't want to for other reasons.

          A technical person would not have made the mistake of bringing that up, only someone clueless chasing buzzwords would.

          Cheers,
          _
          They have to life the filter to make it work, technical person or not. Besides is it even possible to detect the version of NSS from JavaScript, this is probably why they ended up filtering the UA, because the only other way is to have the DRM module fail at playback.

          Comment


          • #35
            Originally posted by AJenbo View Post
            They have to life the filter to make it work, technical person or not.
            I just found it strange that a company would let a non-technical person contact the core technical channel of another company and write things that the technical people on that channel must have immediately categorized as just playing with buzzwords.

            Normally companies do not try to make themselves look incompetent.

            Originally posted by AJenbo View Post
            Besides is it even possible to detect the version of NSS from JavaScript, this is probably why they ended up filtering the UA, because the only other way is to have the DRM module fail at playback.
            The point is that the UA check is worthless. If they wanted to check for the DRM functionality, they could have.
            The UA in this case would have told them that it does, yet they treat it like it does not.

            Cheers,
            _

            Comment


            • #36
              Originally posted by anda_skoa View Post
              I just found it strange that a company would let a non-technical person contact the core technical channel of another company and write things that the technical people on that channel must have immediately categorized as just playing with buzzwords.

              Normally companies do not try to make themselves look incompetent.

              Typically it's developers to managers to marketing to managers in other company to developers in other company unless there's subcontracting or something else special in place to allow direct developer communication

              Comment


              • #37
                Originally posted by anda_skoa View Post
                The point is that the UA check is worthless. If they wanted to check for the DRM functionality, they could have.
                The UA in this case would have told them that it does, yet they treat it like it does not.
                The function might very well be there, but not function properly (not showing video, but neither giving and error). This makes a function test worthless, and leave a UA check as the only option.

                Comment


                • #38
                  Originally posted by AJenbo View Post
                  The function might very well be there, but not function properly (not showing video, but neither giving and error). This makes a function test worthless, and leave a UA check as the only option.
                  Even if we assume that can't detect a failing key exchange, which I am pretty sure they can as this is required for normal operations anyway, it still doesn't make the UA check any more worthwhile.

                  The UA string simply does not contain any information regarding the problem.

                  The UA check would make sense to detect browser too old to even have the functionality, but since it works by simply changing the platform part of the UA string, it obviously can deliver.

                  Cheers,
                  _

                  Comment


                  • #39
                    Originally posted by anda_skoa View Post
                    Even if we assume that can't detect a failing key exchange, which I am pretty sure they can as this is required for normal operations anyway, it still doesn't make the UA check any more worthwhile.

                    The UA string simply does not contain any information regarding the problem.

                    The UA check would make sense to detect browser too old to even have the functionality, but since it works by simply changing the platform part of the UA string, it obviously can deliver.
                    If you know that Windows older then 8.1 and Mac OS older then 10.9 are broken and anything else has a more then 50% of being broken, and you can detect this using the UA, then that is not worthless information. In fact it solves the problem for all Windows and Mac users and only gives you a false negative less then 1% of the time (assuming 2% are non Win/Mac/ChromeOS users). The announcement they gave means that they are going to change it so that it will instead give a less then 1% false positive, as more then 50% of the "other" users can be assumed to work. This is also better as we can assume that the false positive would not be able to use the alternative (silverlight) any way.

                    Comment


                    • #40
                      Originally posted by AJenbo View Post
                      If you know that Windows older then 8.1 and Mac OS older then 10.9 are broken and anything else has a more then 50% of being broken, and you can detect this using the UA, then that is not worthless information.
                      Right, I agree. If such information were present in the UA string, then of course that would be valuable.
                      We are, however, talking about the Chrome UA string on Linux, which does not contains such detailed version information.

                      Cheers,
                      _

                      Comment

                      Working...
                      X