muaz-khan / detectrtc Goto Github PK

Do you have the type script code for DetectRTC.

Right now the isMobileDevice script is a bit naive, but it gets the job done for 99% of devices. I propose we upgrade to:

var isMobileDevice = /Android|webOS|iPhone|iPad|iPod|BlackBerry|IEMobile|Opera Mini|Mobile|mobile/i.test(navigator.userAgent)

(source: http://stackoverflow.com/a/25394023/696130)

or something more sophisticated taken from isMobile.js.

Hello.

An error occurred when call function navigator.mediaDevices.enumerateDevices() for Internet Explorer 11 (and higher). This error occurred in this code part (DetectRTC.js):

if (navigator.mediaDevices && navigator.mediaDevices.enumerateDevices) {
    // Firefox 38+ seems having support of enumerateDevices
    // Thanks @xdumaine/enumerateDevices
    navigator.enumerateDevices = function(callback) {
        navigator.mediaDevices.enumerateDevices().then(callback).catch(function() {
            callback([]);
        });
    };
}

Thank for your work.

Chrome , Firefox and Safari and other browser's version is latest or not..

For example ,

1. DetectRTC.browser.isChromeLatest;
2. DetectRTC.browser.isFirefoxLatest;
3. DetectRTC.browser.isSafariLatest;
4. DetectRTC.browser.isOperaLatest;
5. DetectRTC.browser.isIELatest;
6. DetectRTC.browser.isEdgeLatest;

I couldn't find browser compatibility in the documentation. Can you please advise? Very useful library btw, great job!

.

This call to window.indexedDB causes following exception:

InvalidStateError: A mutation operation was attempted on a database that did not allow mutations.

if you have deactivated indexedDB on firefox (not default, but possible by plugins etc).

To replicate:
set dom.indexedDB.enabled in about:config to false and call window.indexedDB from JS console.

if (navigator.mediaDevices && navigator.mediaDevices.enumerateDevices)

in DetectRTC.js?b37f:17

fails in chrome 45

I've been using this library for a WebRTC-enabled chrome extension, and noticed that the Default audio output device is not showing up in DetectRTC, but is showing up when I enumerate the devices on the console.

I'm using latest Chrome.

Can anyone help me understand how to make the Default device available? Thanks!

Hi,

I have tried many method to detect local speakers & checked your given example as well. But nothing is working.

Could you please me to use correct API to detect local speakers?
Thanks in Advance

Thanks,
KrishPS

This is less of an issue with this library (which works nicely) and more of an issue with Twitter's handling of their in-app iOS browser and it's WebRTC support. However, I figured someone might be able to point me in the right direction and greatly appreciate the help. I was very excited when iOS Safari added support for WebRTC and quickly made it a part of my work. However, I noticed that in-app browsers can be hit or miss when it comes to supporting standards. Last time I checked, Twitter and Instagram's in-app browsers did not support WebRTC but Facebook's did. Anyway, it's pretty straight-forward to look at the browser user agent to detect if the user is coming from an in-app browser or Safari directly. However... Twitter's in-app browser has the same exact user agent as the standalone Safari browser... FML. In addition, WebRTC seems to fail silently on prompting and I can't figure out a way to figure out if it's working or not. An easy test is to direct message yourself the following link on Twitter and attempt to open it within the iOS app:

https://www.webrtc-experiment.com/DetectRTC

The only thought I had was to use a timer to detect if either a positive or negative WebRTC action occurred after a specified period of time. If neither occurred, I could assume that something went wrong and encourage the user to try a different browser.

Any help is much appreciated.

Getting this issue on Internet Explorer and Safari

Unable to get property 'prototype' of undefined or null reference

on Line no 840 approx

if ('createMediaStreamSource' in window[item].prototype) {
webAudio.isCreateMediaStreamSourceSupported = true;
}

and then I get

ReferenceError: 'DetectRTC' is undefined

curious
In the index.html test, why is it 2 tests using isWebSocketsSupported ?
Line 234: appendTR('Browser Supports WebSockets?', printVal(DetectRTC.isWebSocketsSupported), 'isWebSocketsSupported');
and
Line 250: appendTR('Browser Supports WebSockets API?', printVal(DetectRTC.isWebSocketsSupported), 'isWebSocketsSupported');

seems like this error is already fixed in WebRTC-Experiment, would be nice if you could update the repo

muaz-khan/WebRTC-Experiment#298

Hello!
DetectRTC is not working on Xiamo MI browser which is using chromium 61.0.3163.128 under the hood.
Reproduction:

DetectRTC.load(() => {
       // this is never called
});

Any ideas if this can/will be fixed ?
Thanks.

Can you please put a new release on bower that includes the latest bug fixes?

I see that https://www.webrtc-experiment.com/DetectRTC/ loads https://cdn.webrtc-experiment.com/DetectRTC.js, which says it has version 1.3.6.

We are using the released version 1.3.5 from npm. When we use this version on Chrome / Android, it says hasSpeakers is false. While the testpage on webrtc-experiment says that hasSpeakers is true. When I manually added the code of DetectRTC from the master branch, things started to work (hasSpeakers === true).

Can you please make a release 1.3.6, and publish it on npm?

1. Is DetectRTC.DetectLocalIPAddress supported to work with MS Edge browsers in a default configuration? When I test with Edge 16 it does not return anything, but the same code returns the local IP in Chrome.

For instance, DetectRTC.DetectLocalIPAddress(function(ip) {alert(ip);}) doesn't show any alert.

Running your live demo shows the same issue: https://www.webrtc-experiment.com/DetectRTC/ (local IP returned in Chrome 62, but not Edge).

2. Likewise when running in privacy mode in Edge the sample shows Nope for privacy mode whereas it should show yes (but does get it right in Chrome).

Thank you.

Perdon for my english.

I use Ipad Air with Safari, Chrome and Firefox. When I execute the DetectRTC.browser.isSafari sentence, in the three browsers it gives the same result => true.

Thank you very much for the great job you have done.

Once we remove the camera and try to load DetectRtc still it is giving the previously removed camera in the list until we close the tab r browser.This is happening only in chrome latest versions above 70, in firefox it is giving empty results which is working fine.

Hi!

First of all: Great tool! Unfortunately I cannot get it to work.

Szenario:
Page opens. I'm calling DetectRTC.load() in my HTML right in the header after including the script and then initializing all WebRTC related stuff when $.ready is called. But when I want to use the variables like "hasWebcam" or "hasMicrophone" they're telling me that nothing was detected.
Strangest thing: when I'm debugging your lib to see what's going on, all variables are getting initialized (correctly!) and when I reach the point where it usually fails: All variables are available. So I'm heaving some kind of asynchronous issue.

In short:
without breakpoints -> not working
with -> working

Can you help me?

Thanks a lot in advance!

If you're using SSR, the environment lacks DOM globals, but the dependencies are treated like in the browser environment. In your code, you're assigning an empty object to the document:

that.document = {};

which is a serious hazard and can simply crash any code depending on this. Here's an example of one of my dependencies (store.js):

Seems like React's SSR uses node-process which sets the browser property on process object:
https://github.com/defunctzombie/node-process/blob/master/browser.js#L156

Based on this fact I've prepared a quick & dirty fix:
feimosi@6a58481
Would be great if you could incorporate it into the next release or suggest a better solution.

Hi Muaz,

First, thanks for your work.

I've remark that in my application i got the "Please invoke getUserMedia once" only for the microphone list on my Android Firefox browser (v56.0).
I thougth i was doing something wrong, but if you go on this page https://www.webrtc-experiment.com/DetectRTC/ on Androïd Firefox browser you will have the "Please invoke getUserMedia once".

How to fix this issue ?

Regards,
Loïc.

Hi , I am using DetectRTC for checking microphone and camera permissions.

I need to check for if camera permission is set to "Ask(default)". The value of "isWebsiteHasWebcamPermissions" key is false even when it is actually "Ask(default)". It is not blocked nor allowed. How can I get this using DetectRTC?
Please guide me on this.
Thank you.

Running DetectRTC.isSctpDataChannelsSupported && DetectRTC.isWebRTCSupported && DetectRTC.isWebSocketsSupported on a webkit based browser (Apple Safari, Epiphany) throws a type error:

[Error] TypeError: undefined is not an object (evaluating 'window[item].prototype')
	(anonymous function) (DetectRTC.js, line 840)
	forEach ([native code], line 0)
	(anonymous function) (DetectRTC.js, line 832)
	global code (DetectRTC.js, line 1026)

Failing code is at if ('createMediaStreamSource' in window[item].prototype)

    // --------- Detect if WebAudio API are supported
    var webAudio = {
        isSupported: false,
        isCreateMediaStreamSourceSupported: false
    };

    ['AudioContext', 'webkitAudioContext', 'mozAudioContext', 'msAudioContext'].forEach(function(item) {
        if (webAudio.isSupported) {
            return;
        }

        if (item in window) {
            webAudio.isSupported = true;

            if ('createMediaStreamSource' in window[item].prototype) {
                webAudio.isCreateMediaStreamSourceSupported = true;
            }
        }
    });
    DetectRTC.isAudioContextSupported = webAudio.isSupported;
    DetectRTC.isCreateMediaStreamSourceSupported = webAudio.isCreateMediaStreamSourceSupported;

@muaz-khan do you happen to know how to implement ICE Servers (STUN and TURN) it doesn't matter if the language but if its in golang it would be awesome!

In IE11, the browser version is reported as 5. Even though IE doesn't support WebRTC, it would be nicer to make sure the version reported is correct.

Looks like you have 'hasSomethig' properties always true regardless of actual present devices because of code:

    var hasMicrophone = canEnumerate;
    var hasSpeakers = canEnumerate;
    var hasWebcam = canEnumerate;

Hey Muaz,

I am very happy to see this library here! In order to lock it's dependency to a specific version it would be great if you could create a release version here on GitHub.

Thanks for the bower support!

Benny

If I dismiss the permission dialog using the close-button (which turns out to be equivalent to selecting "not now" from the popup menu) then the entire results section never shows at all.

If I explicitly choose "don't share" from the popup menu then the results section correctly shows the number of mics and cams, but all of them have the description string "Please invoke getUserMedia once.".

DetectRTC v1.27
Mozilla/5.0 (X11; Linux x86_64; rv:44.0) Gecko/20100101 Firefox/44.0 Iceweasel/44.0.2

I run an interval function that check the status of these two values (inside a DetectRTC.load call) every three seconds.

But here it comes the problem: isWebsiteHasWebcamPermissions and isWebsiteHasMicrophonePermissions are updated only if they become TRUE (i.e., when the user grants mic and camera permissions), but they don't update to FALSE if the user revoke these permissions.

In other words, after that DetectRTC finds that isWebsiteHasWebcamPermissions and isWebsiteHasMicrophonePermissions became true, it seems these two values aren't updated anymore, regardless any user interaction.

The issue occurs on both Firefox and Chrome.

Thank you in advance for your support.

isSetSinkIdSupported is returning false in new firefox version 64

it should be true, thank you

https://github.com/muaz-khan/DetectRTC/blob/master/dev/DetectRTC.js#L97

if (DetectRTC.browser.isChrome && DetectRTC.browser.version >= 46 && location.protocol !== 'https:') {
    DetectRTC.isGetUserMediaSupported = 'Requires HTTPs';
}
DetectRTC.isGetUserMediaSupported = isGetUserMediaSupported;

In DetectRTC info page you say that:

Can you please provide an example? I think the browser will never let you do so without a an user permission...
Tried with Chrome and Firefox with this snippet:

only my first feed is shown...
nothing happens with track.applyConstraints()

Here's the output:

When i visit the demo page (https://www.webrtc-experiment.com/DetectRTC/,), it always shows the permission check.

After accepting the camera and microphone, "Website has webcam permissions?" returns a yep. ""Website has microphone permissions?" on the otherside always returns a Nope. Therefore there might be an issue here with the isWebsiteHasMicrophonePermissions check.

Here is a short video of the issue.

SCRIPT438: Object doesn't support property or method 'indexOf'

        // trim the fullVersion string at semicolon/space if present
        if ((ix = fullVersion.indexOf(';')) !== -1) {
            fullVersion = fullVersion.substring(0, ix);
        }

if (DetectRTC.hasSpeakers === false && (DetectRTC.broser.name === 'Chrome' || DetectRTC.broser.name === 'Edge')) {
alert('Oops, your system can not play audios.');
}

Should read browser, not boser

If you bower install with tag 1.0.9 it will fail on IE10.
Please create a new tag with the updated master branch.

Hey there. I'm currently working on a product that's using the RTCOutboundRTPStreamStats API. The API has undergone a few iterations and different browsers have implemented it differently in the past - specifically the way framesEncoded works.

Right now we're just browser detecting but it might be nice to feature detect this. Is this something that DetectRTC could/should do, or is it a little out of scope for this lib?

A bit of context:
General browser support with vendor prefixes, etc: http://caniuse.com/#feat=rtcpeerconnection
Discussion around the spec + implementation for Firefox: https://bugzilla.mozilla.org/show_bug.cgi?id=1348657

On Safari, the version reported is actually the Webkit version rather than the Safari version. For Safari version 10.1.1, it's currently reporting 603.2.4, which is inconsistent with how other places report Safari versions.
The cause is that the current mechanism we have for detecting Safari doesn't support Safari 10+.

On Edge browser, DetectRTC is not reporting the major version number.
E.g. on Windows 10, I expect the reported version to be 15.15063, and yet it's only reporting 15063.
This appears to be an intentional choice.

This is not consistent with how Edge versions are normally reported. In the following links, Edge version is identified by major version number only.
https://www.netmarketshare.com/browser-market-share.aspx?qprid=2&qpcustomd=0
https://analytics.usa.gov/data/
http://caniuse.com/#feat=rtcpeerconnection

Suggested fix: uncomment line 80 and remove line 81.

See https://www.webrtc-experiment.com/DetectRTC/#isWebSocketsBlocked.

I'm seeing:

To be clear, the LABEL says Nope, as well as the value.

Label device didnt show, but permision granted

Hey!

When calling DetectRTC.load() and webcam is plugged, DetectRTC.hasWebcam value is set to true. But when unplugging my webcam and calling DetectRTC.load(), DetectRTC.hasWebcam value is still set to true.

I purpose you to set hasMicrophone/hasSpeakers/hasWebcam to false as you set audioInputDevices = []; ... in the function getIPs(callback). What do you think ?

Trying this on an android phone with two cameras. I can see both cameras if I manually enumerate devices but DetectRTC only shows the one. See below:

Hi

At DetectRTC.js:598 you return early when canEnumerate is false, but here you don't call the callback. This is different from line 611 where you do call the callback when navigator.enumerateDevices does not exist. This causes the callback of DetectRTC.load to never be called in some browsers (specifically IE10).

'webkitAudioContext' is deprecated. Please use 'AudioContext' instead.

Version 46.0.2490.86 (64-bit) 
Google Chrome is up to date.

the problem is related to function

function afterCreateOffer() {
  var lines = pc.localDescription.sdp.split('\n');

  lines.forEach(function(line) {
      if (line.indexOf('a=candidate:') === 0) {
          handleCandidate(line);
      }
  });
}

https://github.com/muaz-khan/DetectRTC/blob/master/DetectRTC.js#L647

I think it should be

function afterCreateOffer() {
  var lines = pc.localDescription.sdp.split('\n');

  lines.forEach(function(line) {
      if (line && line.indexOf('a=candidate:') === 0) {
          handleCandidate(line);
      }
  });
}

https://github.com/muaz-khan/DetectRTC/blob/master/DetectRTC.js#L33 defines window as a property of global, which causes checks like typeof window === 'undefined' to run browser code in node environments.