My RTC session was started with text only. And video is added by user when needed (renegotiation)
navigator.getUserMedia({ video: true, audio: false }, function (myStream) {
localVideo[0].srcObject = myStream;
myConn.addStream(myStream);
}, function (error) {
console.log(error);
});
When user do not need the video session anymore, I remove using:
var tracks = localVideo[0].srcObject.getTracks();
tracks.forEach(function (t) {
t.stop();
});
myConn.removeStream(localVideo[0].srcObject);
localVideo[0].srcObject = null;
Everything is working fine, until I try to add the video again I noticed that the createOffer() request size is getting larger and larger.
Seems to me that WebRTC didn't forget about the previous stream, and is adding to the offer again and again. Or maybe my way of removing a video stream / track is wrong?
This is a known issue see this thread on the W3C list.
The best way to get around this is to use replaceTrack and is suggested in the thread.
Note: It is still possible to prevent the list of transceivers from growing
by *manually* recycling them using transceiver.sender.replaceTrack() and
transceiver.direction, but that still wastes resources on transceivers
currently not used, and implies you probably shouldn't use
transceiver.stop() in most cases.
Also see the "Unified Plan" Transition Guide
Related
I am trying to play a video when developing locally with VueJS 2.
My code is the following :
<video class="back_video" :src="`../videos/Space${videoIndex}.mp4`" id="background-video"></video>
...
data :
function() {
return {
videoIndex:1
}
}
...
const vid = document.getElementById("background-video");
vid.crossOrigin = 'anonymous';
let playPromise = vid.play();
if (playPromise !== undefined) {
playPromise.then(function() {
console.log("video playing");
}).catch(function(error) {
console.error(error);
});
}
This code is causing the exception given in title. Tried in several browsers, always the same.
If I change the src by :
:src="require(`../videos/Space${videoIndex}.mp4`)"
it works.
But in that case building time is very long as I have many different videos in my videos directory, because adding require() will force to copy all videos in the running directory at build phase (vue-cli serve), and this is really annoying. In other words I want to refer videos that are outside the build directory to avoid this (but also to avoid having videos in my git).
It is interesting to note that when I deploy server side, it works perfectly with my original code
:src="`../videos/Space${videoIndex}.mp4`"
Note also that if i replace my code with simply
src="../videos/Space1.mp4"
it works too. So the video itself, or its location, are not the source of the problem.
Any clue ?
You can host your videos on a CDN to have something faster and easier to debug/work with.
Otherwise, it will need to bundle it locally and may take some time.
I use the dynamic source for vue-webpack images in nuxt :src="require('path/to/image' + dynamic.variable)" in my project navbar. If the users substitute their image through a form which refetches their information and deletes their previous image I get a webpack error module (img) not found (it does not find the new one): is there a way to solve this, like wait for webpack HRM to finish?
I tried setting up a setTimeout() of one second before user re-fetch and it works, but I don't like a random waiting, I'd use a promise or a sync dynamic, the point is webpack hot reload is not controlled by my functions.. I also tried with setting the dynamic path as a computed: but it doesn't fix.
My image tag:
<img v-if="this.$auth.user.image" class="userlogo m-2 rounded-circle" :src="require('#assets/images/users/' + this.$auth.user.image)" alt="usrimg">
My Useredit page methods:
...
methods: {
userEdit() {
//uploads the image
if (this.formImageFilename.name) {
let formImageData = new FormData()
formImageData.append('file', this.formImageFilename)
axios.post('/db/userimage', formImageData, { headers: { 'Content-Type': 'multipart/form-data' } })
// once it has uploaded the new image, it deletes the old one
.then(res=>{this.deleteOldImage()})
.catch(err=>{console.log(err)})
}else{
this.userUpdate() //if no new image has to be inserted, it proceeds to update the user information
}
},
deleteOldImage(){
if(this.$auth.user.image){axios.delete('/db/userimage', {data: {delimage: this.$auth.user.image}} )}
console.log(this.$auth.user.image + ' deleted')
this.userUpdate() // it has deleted the old image so it proceeds to update the user information
},
userUpdate(){
axios.put(
'/db/user', {
id: this.id,
name: this.formName,
surname: this.formSurname,
email: this.formEmail,
password: this.formPassword,
image: this.formImageFilename.name,
})
.then(() => { console.log('User updated'); this.userReload()}) // reloads the updated user information
.catch(err => {console.log(err)} )
},
userReload(){
console.log('User reloading..')
this.$auth.fetchUser()
.then(() => { console.log('User reloaded')})
.catch(err => {console.log(err)} )
},
}
...
the problem happens after "console.log('User reloading..')" and before "console.log('User reloaded');", it is not related to the file upload nor the server response. I broke a single function in many little ones just to check the function progression and its asynchronous dynamics but the only one that is not manageable is the webpack hot reload :/
I'd like the users to upload their images and see their logo in the Navbar appear updated after submitting the form.
First of all, as somebody told you in the comments, webpack hmr shouldn't be used for production.
In Nuxt, everything that you reference from the assets folder will be optimized and bundled into the project package. So the ideal use case for this folder is all assets that can be packaged and optimized, and most likely won't change like fonts, css, background images, icons, etc.
Then, require is called only once by webpack when it is either building the site for local development or building the site for generating a production package. The problem in your case is that you delete the original file while you're in development and webpack tries to read it and fails.
In the case of these images that the user uploads, I think you should use the static folder instead and instead of using require you'll have to change the :src with
:src="'/images/users/' + this.$auth.user.image"
Let me know if this helps.
Okay, I probably solved it.
HMR: you are of course right. Thank you for pointing out, I am sorry, I am a beginner and I try to understand stuff along the way.
Aldarund, thank you, your idea of not changing the path and cache it client side.. I am too noob to understand how I could implement it ( :) ) but it gave me a good hint: the solution was to keep the image name as the user id + the '.png' extension and to manage the image with jimp so that the image name, extension and file type are always the same, and with or without webpack compiling the new path, I always have the correct require().
Jair, thank you for the help, I didn't follow that road, but I will keep it as a second chance if my way creates errors. Just to be specific: the error comes when it does not find -and asks for the name of- the NEW image, not the OLD one, as I wrote in my question: it happens because the fetchUser() functions reloads the user information including the new image name.
Do you guys see any future problems in my methodology?
Really thank you for your answers. I am learning alone and it's great to receive support.
I have a file uploader that handles very large files. Lots of times the users will cancel an upload. There does not appear to be any way to make the browser cancel the XHR POST that is handling it. Consequently, progress and complete events fire much later and the upload actually completes. Presumably there must be an XHR object embedded in the uploader somewhere that I could call abort on, but I see nothing in the API docs, or in a console dump of the uploader object.
If your goal is to support only the HTML5 uploader (and not the Iframe or the Flash version), then you can do something like that:
Create a new uploader widget with the following (code not tested, might need some adjustment):
define(['dojo/_base/declare', 'dojox/form/Uploader'], function(declare, Uploader) {
return declare([Uploader], {
xhrRequest: null,
createXhr: function() {
this.xhrRequest = this.inherited(arguments);
return this.xhrRequest;
},
cancel: function() {
if (this.xhrRequest && this.xhrRequest.abort) {
this.xhrRequest.abort();
}
}
});
});
Use your new widget instead of dojox.form.Uploader
Call the cancel method when you want to cancel the upload.
But remember: this will work only with the HTML5 uploader.
I'm trying to automate an application that uses form security in order to upload a file and then scrape data from the returned HTML.
I started out using the solution from this question. I can define my steps and get through the entire workflow as long as the last step is rendering the page.
Here are the two steps that are the meat of my script:
function() {
page.open("https://remotesite.com/do/something", function(status) {
if ('success' === status) {
page.uploadFile('input[name=file]', 'x.csv');
page.evaluate(function() {
// assignButton is used to associate modules with an account
document.getElementById("assignButton").click();
});
}
});
},
function() {
page.render('upload-results.png');
page.evaluate(function() {
var results = document.getElementById("moduleProcessingReport");
console.log("results: " + results);
});
},
When I run the script, I see that the output render is correct. However, the evaluate part isn't working. I can confirm that my DOM selection is correct by running it in the Javascript console while on the remote site.
I have seen other questions, but they revolve around using setTimeout. Unfortunately, the step strategy from the original approach already has a timeout.
UPDATE
I tried a slightly different approach, using this post and got similar results. I believe that document uses an older PhantomJS API, so I used the 'onLoadFinished' event to drive between steps.
i recomend you use casperjs or if you use PJS's webPage.injectScript() you could load up jquery and then your own script to do form input/navigation.
I am looking fore some code allowing me to launch automatically a video after another one.
I'am using the great video.js library, which has a quite complete API. I found some snippet to get an event listener working at the end of the 1st video, but then I cannot launch the second one.
This is working, displaying an alert at the end of the 1st video :
_V_("intro").ready(function(){
this.addEvent("ended", function(){
alert('foo');
});
});
And this is also working, launching a video in fullscreen on page reload :
_V_("leader").ready(function(){
var leader = this;
leader.requestFullScreen();
leader.play();
});
But I can't get the 2nd video launching in fullscreen at the end of 1st video...
Last subtility, I would like to entirely build the 2nd video with javascript, not having to write it and just hiding id with CSS.
Thank you folks !
Elliot
You can simply use the provided 'src' method in the Video.js API, if you want to play a second video right after the first one finishes it would work like this:
_V_("intro").ready(function(){
this.addEvent("ended", function(){
this.src({ type: "video/mp4", src: "http://path/to/second/video.mp4" });
});
});