browserless / browserless Goto Github PK

🚨 You need to enable Continuous Integration on all branches of this repository. 🚨

To enable Greenkeeper, you need to make sure that a commit status is reported on all branches. This is required by Greenkeeper because it uses your CI build statuses to figure out when to notify you about breaking changes.

Since we didn’t receive a CI status on the greenkeeper/initial branch, it’s possible that you don’t have CI set up yet. We recommend using Travis CI, but Greenkeeper will work with every other CI service as well.

If you have already set up a CI for this repository, you might need to check how it’s configured. Make sure it is set to run on all new branches. If you don’t want it to run on absolutely every branch, you can whitelist branches starting with greenkeeper/.

Once you have installed and configured CI on this repository correctly, you’ll need to re-trigger Greenkeeper’s initial pull request. To do this, please delete the greenkeeper/initial branch in this repository, and then remove and re-add this repository to the Greenkeeper App’s white list on Github. You'll find this list on your repo or organization’s settings page, under Installed GitHub Apps.

As a consequence of the fact that each browser keeps its own separate userDataDir by default, and since browserless maintains a pool of browsers to serve incoming requests, it seems to me that caching will be somewhat sub-optimal... for N browsers there are N caches to fill, more cache misses, and more cached data.

Is there any way to ensure that the same browser (and associated cache) is used for the same source, so that a consistent cache is used for a particular user? Perhaps based on some ID or key, or even IP address?

Ideally a common cache area would be used for all browsers, but I'm not sure this is possible without horrible conflicts. I did try passing the same userDataDir to all browsers and indeed it didn't work out well. Perhaps using a shared memory mount (/dev/shm) would work better since that is "supposed" to be shared between processes?

The dependency puppeteer was updated from 1.8.0 to 1.9.0.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

puppeteer is a direct dependency of this project, and it is very likely causing it to break. If other packages depend on yours, this update is probably also breaking those in turn.

Your Greenkeeper Bot 🌴

Hey,

I've an issue with a website when I try to load it in my debugger but it works in the browserless debugger

On MY debugger

On browserless.io debugger

I know that's an issue that came from the website, but I'm wondering why I got this error

Hi,

This page explain that is a good practice to close browser connection after some processing.

After 10 connections, Browserless container throw this log :

(node:26) MaxListenersExceededWarning: Possible EventEmitter memory leak detected. 11 SIGHUP listeners added. Use emitter.setMaxListeners() to increase limit

Here is an example :

1. Mount a Browserless container on Docker :
   docker run -p 3000:3000 browserless/chrome:release-puppeteer-1.3.0

2. Exec this code example :

const puppeteer = require('puppeteer');

(async function () {

    // Synchronous Loop
    for (let i = 0; i < 20; i++) {
        const browser = await puppeteer.connect({
            browserWSEndpoint: `ws://127.0.0.1:3000`,
        });

        const page = await browser.newPage();

        try {
            await page.goto('https://www.google.com');
            await page.screenshot({
                path: './browserless.png'
            });
            browser.close();

            await new Promise(resolve => setTimeout(resolve, 5000));
            console.log('close: ' + i)
        } catch (error) {
            console.error({
                error
            }, 'Something happened!');
            browser.close();
        }
    }
})();```

After 10 loops, a warning happened on Browserless container terminal.

The devDependency husky was updated from 1.1.0 to 1.1.1.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

husky is a devDependency of this project. It might not break your production code or affect downstream projects, but probably breaks your build or test tools, which may prevent deploying or publishing.

Your Greenkeeper Bot 🌴

my code is like this:
const request = require('request'); module.exports = async ({page, context, id}) => { };
then I got this error message.
And in your website,you said that "You can currently require 'url', 'util', 'path', 'querystring', 'lodash', 'node-fetch', and 'request' in your functions. Please contact us for adding a module".
So I wander if I write my code in a wrong way?

Hi ! Thanks for this awesome docker image !
How to use with docker-compose ?

Here is my docker-compose.yml.

version: '3' # https://blog.codeship.com/using-docker-compose-for-nodejs-development/
services:
  app:
    build:
      dockerfile: Dockerfile.dev
      context: .
    image: app
    environment:
      - TIMBER=true
      - NODE_ENV=development
    command: node --inspect=0.0.0.0:3001 --require dotenv/config ./dist/index.js
    ports:
      - "8080:8080"
      - "9229:3001"
    links:
      - browserless
    depends_on:
      - browserless
  browserless:
    image: browserless/chrome:latest
    container_name: "browserless"
    environment:
      - DEBUG=browserless/chrome
      - MAX_CONCURRENT_SESSIONS=10
    ports:
        - 3002:3000

What should i put as url to connect to it with puppeteer ?
app.locals.browser = await puppeteer.connect({browserWSEndpoint: 'ws://localhost:3002'}) doesn't work.

Thanks for your help !

Hi,

When I made a screenshot of this website http://fshr-kyoto.com, fonts are missing.
Any idea how to display them?

Hi :)
in my scenario i have 10 "Active" page. in initialize phase i connect with puppeteer to browserless and make a chromium instance.
in iterating loop i want to read each pages Html.
here browserless package.json start configuration :
"dev": "npm run build && cross-env ENABLE_CORS=true MAX_CONCURRENT_SESSIONS=1 MAX_QUEUE_LENGTH=20 PREBOOT_CHROME=true CHROME_REFRESH_TIME=3600000 KEEP_ALIVE=true DEBUG=browserless* PORT=3030 node build/index.js"
as you see i set PREBOOT_CHROME=true and KEEP_ALIVE=true property.
My problem :
as i mentioned in initialize phase with puppeteer.connect i connect to browserless. in first iterate 10 instance of chromium created. in second step (NOT second iterate but still in first iterate) i use browser.newPage to get a page and in second step with page.content i get page content. note puppeteer.connect (first step) is in exported from module and singleton (i think :) )

first iterate complete successfully. but after it browserless closes chromium instances! and in another iterate loops i got >get WebSocket is not open: readyState 3 (CLOSED) error!

what should i do ? can you help me @joelgriffith ?

I've been trying to get capybara working with browserless in a rails app using selenium. During the test cycle capybara will tell the selenium driver to reset cookies. Selenium uses a DELETE request to /session/XXXXXXXXXX/cookie to handle this. It appears that browserless's proxy somewhat naively treats all DELETE requests by closing the session leading to all subsequent requests from capybara/selenium to fail.

Here's a small script that demonstrates the issue:

require 'selenium-webdriver'
Selenium::WebDriver.logger.level = 'debug'

chrome_options = Selenium::WebDriver::Remote::Capabilities.chrome(
  "chromeOptions" => { args: %w( --headless --no-sandbox --disable-gpu ) }
)

s = Selenium::WebDriver::Remote::Driver.new(desired_capabilities: chrome_options, url: "http://chrome:4444/webdriver")

s.manage.delete_all_cookies
s.navigate.to('about:blank') # Will fail

It seems from the documentation that there are a handful of DELETE requests in the spec. I. think the appropriate thing to do is to only close the session when encountering a DELETE to the /session/ID url.

I'm gonna mess around with a PR for this but feel free to comment in the mean time if you have encountered this and feel like my approach is ill advised for some reason.

I followed the Docker Quickstart:

1. docker pull browserless/chrome
2. docker run --shm-size=1gb -p 3000:3000 browserless/chrome
3. Visit http://localhost:3000/ to use the interactive debugger.

I tried this on two separate VPS's, same issue - running Ubuntu Server 16.04.3 LTS.

It's up on one of my test systems: http://144.217.188.229:3000/

Edit: I've also tried the Node Quickstart, same issue.

Hi,

the docker version is working properly but if I try to npm run dev I'll get following error:

[nodemon] failed to start process, "ts-node" exec not found

What it's working in my case:

1. changing port to 8081 in src/config.ts
2. npm run build
3. npm run start

docker image tag: puppeteer-1.9.0
puppeteer-core: 1.9.0
Node.js version: 8.9.1

After a page is open, I tried to run page.$$(selector) . At this point this error message showed up in the logs from the Node app:

Error: Protocol error (Target.activateTarget): Target closed.

What does this error message mean? How can I fix it?

The dependency express-http-proxy was updated from 1.3.0 to 1.4.0.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

express-http-proxy is a direct dependency of this project, and it is very likely causing it to break. If other packages depend on yours, this update is probably also breaking those in turn.

Your Greenkeeper Bot 🌴

I'm unable to install this. Pretty hard for me. Please make this easy to install

Hey,
i'm trying this out as a potential service for web-testing, and one of the first tests I tried was how to handle the file upload.
When connected through puppeteer with browserWSEndpoint, I just can't seem to upload files dynamicaly, while without the browserWSEndpoint, it works fine.
I've even tried passing down the headers --allow-file-access-from-files&--disable-web-security but without any luck.

Here's how i'm trying:

const browser = await puppeteer.connect({
    browserWSEndpoint: 'ws://localhost:32769',
    headless: true
})
const page = await browser.newPage();
await page.goto('http://localhost:8080', {waitUntil: 'networkidle2'});

(...)

let testUpload = async () => {
    const upload = await page.$("input");
    await upload.uploadFile('test.jpg');
    await page.screenshot({path: 'test_s_'+Date.now()+'.png'});
}

(...)

I'm 100% sure that all the elements exist, being the DOM elements or the files.
Any ideas or is this a limitation?

Cheers

You have other documentation on https://docs.browserless.io but you don't point that out in the README or anywhere else in the repo. I was having an issue with the debugger session disconnecting and was digging into whether it was in issue with my proxy server but it turns out all I needed was to know about the CONNECTION_TIMEOUT env variable mentioned here. So I think it would make sense to specifically call out the additional documentation somewhere obvious such as the README or maybe to combine all the docs for the open source project and separate them from the service docs.

I also just want add a 👏👏 great job on the project. I've been trying to get something setup with puppeteer and headless chrome for awhile and this is super useful and much appreciated. And the web debugger is 🔥😍.

First of all, thank you for this great project, i'm working on another project to show puppeteer action steps by chrome inspect and i got lots of ideas from your project.

Now i'm stuck by an very weird problem, once i append the puppeteer code with chrome inspector (i load a frame which url like https://chrome-devtools-frontend.appspot.com/serve_file/@7f3cdc3f76faecc6425814688e3b2b71bf1630a4/inspector.html?ws=127.0.0.1:3000/devtools/page/(CA197D0A33141BE44B19C0603ADD9E7C)&remoteFrontend=true ), the await page.click('dom-selecter') action alway blocked, but when i run puppeteer without chrome inspect, everything works fine.

there is no error show up. and i tried debug page.click function, the code run to below:

send(method, params = {}) {
    if (!this._connection)
      return Promise.reject(new Error(`Protocol error (${method}): Session closed. Most likely the page has been closed.`));
    const id = ++this._lastId;
    const message = JSON.stringify({id, method, params});
    debugSession('SEND ► ' + message);
    this._connection.send('Target.sendMessageToTarget', {sessionId: this._sessionId, message}).catch(e => {
      // The response from target might have been already dispatched.
      if (!this._callbacks.has(id))
        return;
      const callback = this._callbacks.get(id);
      this._callbacks.delete(id);
      callback.reject(e);
    });
    return new Promise((resolve, reject) => {
      this._callbacks.set(id, {resolve, reject, method});
    });
  }

it return a promise but this promise never resolved, so the code after page.click never run,

i tried figure out this problem for several days, but still no any progress , so i add an issue here to find any help from you, Do you have any idea about this problem?

i tried below code in browserless, and sometime this problem show.

await page.goto("https://weidian.com/item.html?itemID=1692458617")
await page.click("#buy_now");
await page.waitForSelecter('#item_control');
await page.click('#sku_ul > li:nth-child(2) > a')

I emailed earlier but wanted to create an official issue about adding support for cookies in Chrome. This could be accomplished by passing a cookies object.

Let me know if you have follow up questions.

I have installed the browserless/chrome docker image (as a service), and have opened port 3000, so that it should in theory be accessible via browserless.mysite.com:3000 (like lots of other web apps I've deployed in the same way). But when I browse to browserless.mysite.com:3000 I just get:

This site can’t be reached
browserless.mysite.com unexpectedly closed the connection

Any ideas?

Hi,

(it's me again)

I guess this error is probably due to the way the website has been developped but I'm wondering if there a fix for this.
When screenshoting this website I got an error that prevents to load the website
Uncaught (in promise) DOMException: Failed to load because no supported source was found.

https://chrome.browserless.io/?script=await%20page.goto(%27http%3A%2F%2Fhki.paris%2Fhome%27)%3B

With dumpio, I got more info :
[0926/072802.585761:INFO:CONSOLE(12)] "The Web Audio autoplay policy will be re-enabled in Chrome 70 (October 2018). Please check that your website is compatible with it. https://goo.gl/7K7WLu", source: http://hki.paris/build/desktop.js?v=1 (12) [0926/072810.183949:ERROR:render_media_log.cc(30)] MediaEvent: MEDIA_ERROR_LOG_ENTRY {"error":"FFmpegDemuxer: no supported streams"} [0926/072810.416384:ERROR:render_media_log.cc(30)] MediaEvent: PIPELINE_ERROR DEMUXER_ERROR_NO_SUPPORTED_STREAMS [0926/072810.921032:INFO:CONSOLE(0)] "Uncaught (in promise) NotSupportedError: Failed to load because no supported source was found.", source: http://hki.paris/home (0)

I tried to install FFMpeg on my server but It doesn't changed anything.

Any idea?

Thanks!

Edit:
I'm using chrome & not chromium

Edit 2 :
Here the list of args that I'm launching puppeteer with :
args: [ '--disable-gpu', '--no-sandbox', '--disable-setuid-sandbox', '--disable-translate', '--mute-audio', '--hide-scrollbars', '--disable-translate', '--ignore-certificate-errors', '--ignore-certificate-errors-spki-list' ],

The dependency ts-jest was updated from 23.10.0 to 23.10.1.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

ts-jest is a direct dependency of this project, and it is very likely causing it to break. If other packages depend on yours, this update is probably also breaking those in turn.

Your Greenkeeper Bot 🌴

I encountered an issue in our gitlab-ci pipeline, that sometimes connection reset error occured when two or more docker containers communicate with each other. I found out, that to solve this issue, I should set the server host to bind to IP 0.0.0.0. I updated the IP address and it seems to be working without any issue now. We created our own image with this change, but it would be great, to have this also integrated in the official repo.

The dependency lighthouse was updated from 3.1.1 to 3.2.0.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

lighthouse is a direct dependency of this project, and it is very likely causing it to break. If other packages depend on yours, this update is probably also breaking those in turn.

Your Greenkeeper Bot 🌴

Similar to the /pdf route, we should also allow for sending in a content parameter so that users don't have visit a live site for screenshots

In order to get caching of service workers working in browserless (and puppeteer generally) I find that it's necessary to set a userDataDir via the --user-data-dir passed to puppeteer.launch(). This may be necessary to get plain http caching working too... not sure.

You can see (for service worker caching) this via the following... when the flag is passed via the ws, the browser is able to cache the content in the service worker, and subsequently (after the content has been cached) you'll see "retrieved from service worker cache". Without the flag, you just get "retrieved from network" no matter how many times you request the content.

(async () => {
  const browser = await puppeteer.connect({
    browserWSEndpoint: 'ws://browserless.mydomain.com:3000?--user-data-dir=/tmp'
  });
  const page = await browser.newPage();
  const url = 'https://cloud3squared.com/files/sw-stackoverflow-demo/index.html';
  await page.goto(url, { waitUntil: 'networkidle0' });
  await page.screenshot({ path: './temp.png' });
  browser.close();
})();

This is fine when only one browser has been instantiated by browserless under the hood. The problem comes when multiple browsers are all passed the same --user-data-dir and are all trying to access and write to the same storage area.

At least, I'm seeing a few issues when using browserless which I think result from this. It's partly educated guesswork at this stage.

I wonder if it's necessary to be able to somehow use a different --user-data-dir for each different browser that is instantiated, so that there is no conflict?

Or maybe it's possible to use /dev/shm (if it's big enough and hasn't been disabled) for --user-data-dir (I think that /dev/shm is "supposed" to be shared between different processes) ... I haven't yet tried that ... mainly because I can't since browserless disables /dev/shm by default.

P.S. the page rendered by browserless when "fetched from network" in the above example seems to be missing (unable to render) this character... the "tick" is rendered fine (for "fetched from service worker cache")

If you use PREBOOT_CHROME & MAX_CONCURRENT_SESSION > 10 then you get a MaxListenersExceededWarning for a bunch of events:

% docker run -e "PREBOOT_CHROME=true" -e "MAX_CONCURRENT_SESSIONS=11" -p 3000:3000 browserless/chrome:latest
(node:27) MaxListenersExceededWarning: Possible EventEmitter memory leak detected. 11 exit listeners added. Use emitter.setMaxListeners() to increase limit
(node:27) MaxListenersExceededWarning: Possible EventEmitter memory leak detected. 11 SIGINT listeners added. Use emitter.setMaxListeners() to increase limit
(node:27) MaxListenersExceededWarning: Possible EventEmitter memory leak detected. 11 SIGTERM listeners added. Use emitter.setMaxListeners() to increase limit
(node:27) MaxListenersExceededWarning: Possible EventEmitter memory leak detected. 11 SIGHUP listeners added. Use emitter.setMaxListeners() to increase limit

From Node documentation:

By default EventEmitters will print a warning if more than 10 listeners are added for a particular event.

Maybe browserless should increase the max listener allowed to the MAX_CONCURRENT_SESSION ?

Per our conversation @joelgriffith, I am submitting this request to be able to disable the front-end debugger altogether. Maybe HEADLESS=true would just off the web ui?

Thanks much!

The dependency @types/node was updated from 10.11.5 to 10.11.6.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

@types/node is a direct dependency of this project, and it is very likely causing it to break. If other packages depend on yours, this update is probably also breaking those in turn.

Your Greenkeeper Bot 🌴

Wanted to reach out and gather thoughts on bundling the last 2-5 version of Chrome in single image vs maintaining numerous image's with the only difference being the Chrome version.

PROS

• Less docker and related built maintenance
• Easier to understand build tools
• Quick way to test different versions for performance and other intricacies
• Can apply semver builds in docker (vs confusing release-puppeteer-x.x.x)

CONS

• Bigger image bundles
• Slightly more complex runtime (have to specify the puppeteer version in the URL or something similar).
• Having to download several version of Chrome, and building relevant tooling for each in the debugger (all the tooltip stuffs).
• Things like pre-booting Chrome become harder since you'd have to pre-boot one version only vs many.

Thoughts? Comments? Concerns?

Hey, I can connect to my https://browserless.site just fine and the wss inititiates just fine. However, when I'm using the wss://browserless.site I get an error that returns DEPTH_ZERO_SELF_SIGNED_CERT. Any idea how I can fix this?

I did have it working over ws:// as the Getting Started recommends but I'd like to have my connection encrypted if possible.

I have followed the straightforward instructions from docker and was doing proxy via nginx, also configured the ssl as well via certbot. But it says debugging connection was closed. Below was the url I am trying to achieve.

https://browserless.ngelmat.me

The dependency @types/node was updated from 10.11.0 to 10.11.1.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

@types/node is a direct dependency of this project, and it is very likely causing it to break. If other packages depend on yours, this update is probably also breaking those in turn.

Your Greenkeeper Bot 🌴

🚨 You need to enable Continuous Integration on all branches of this repository. 🚨

To enable Greenkeeper, you need to make sure that a commit status is reported on all branches. This is required by Greenkeeper because it uses your CI build statuses to figure out when to notify you about breaking changes.

Since we didn’t receive a CI status on the greenkeeper/initial branch, it’s possible that you don’t have CI set up yet. We recommend using Travis CI, but Greenkeeper will work with every other CI service as well.

If you have already set up a CI for this repository, you might need to check how it’s configured. Make sure it is set to run on all new branches. If you don’t want it to run on absolutely every branch, you can whitelist branches starting with greenkeeper/.

Once you have installed and configured CI on this repository correctly, you’ll need to re-trigger Greenkeeper’s initial pull request. To do this, please delete the greenkeeper/initial branch in this repository, and then remove and re-add this repository to the Greenkeeper App’s white list on Github. You'll find this list on your repo or organization’s settings page, under Installed GitHub Apps.

When converting a HTML file that contains emojis, my PDF becomes extremely large. Without any emoji characters, the email is around 1.5MB. With emoji characters, the email becomes around 8MB in file size. Other than these emojis, the file is nothing special, just some plain text.

I saw that one of the features/advantages of Browserless it that fonts/emojis are supported out of the box, so this seems an unusual issue. I can print the file as PDF and it will be less than 1MB, so browserless seems to make this file exceptionally large.

Hello! First of all, thanks for the amazing project. It helps a lot.

My problem is: I'm running browserless (hosted) in a ECS cluster. So my scraping jobs just connect to it (on the launch method) and get things running.

I need to keep a log-in session alive for several hours - without actually needing to keep a Chrome instance alive for that long (most of the time I'm just waiting for things to happen). So I'm saving the userDataDir somewhere and, when I need to get back to that session, I launch a new Chrome using this previously saved dir.

But I'm saving the userDataDir from a different machine than the one I run chrome.

Can I pass, somehow, this dir to browserless (without being way too hacky)?

I'm wondering, would there be any timeout issues using this technique ?
For example if I'd launch a big crawl for +24 hours

I'd say no, but maybe you pushed it to its limits :)

Thanks

The dependency lodash was updated from 4.17.10 to 4.17.11.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

lodash is a direct dependency of this project, and it is very likely causing it to break. If other packages depend on yours, this update is probably also breaking those in turn.

Your Greenkeeper Bot 🌴

There is an issue "ignoreHTTPSErrors is not working when request interception is on" which is fixed with puppeteer v1.8.0. Can you move the project onto the latest version, Thanks Glenn

puppeteer/puppeteer#1159 (comment)

According to this it is necessary to pass --shm-size=1gb to docker run.

The problem I have is that I would not be firing browserless up via docker run but rather docker service create... and it's not possible to pass --shm-size as an option to that. There is a possible workaround to pass --mount instead, but to cut a long story short... that's not possible for me.

I have encountered this shm issue before, when deploying headless chrome to docker in a different way, and for that I am passing the option --disable-dev-shm-usage to puppeteer.launch()... based on this advice.

These two different options (i.e. --shm-size=1gb passed to docker run and --disable-dev-shm-usage passed to puppeteer.launch()) seem to be addressing the same problem, so...

Is it possible to pass a custom option (and in particular, --disable-dev-shm-usage) to puppeteer.launch() in browserless?

I think, if that were possible, it would solve my problem!

Would it anyway be better to have browserless pass --disable-dev-shm-usage to puppeteer.launch() by default... since there is then no longer any need to advise users to pass --shm-size=1gb to docker run? Just a simple docker run -p 3000:3000 browserless/chrome.

But that, as well as the ability to pass other custom options, would be nice.

EDIT OK now that I look at the browserless launch code, I see that you do already pass disable-dev-shm-usage to puppeteer.launch()... so I suppose that the above question then becomes... is it actually still necessary to pass --shm-size=1gb to docker run if you're also passing disable-dev-shm-usage to puppeteer.launch()... aren't they there for the same reason?

Having read README.md, it is unclear how to start a docker container with customer arguments.

I have tried:

$ docker run -p 3000:3000 browserless/chrome --proxy-server=127.0.0.1:8050

But that does not work.

Furthermore, even if this did work, how would one then acquire the host IP address? That requires checking ifconfig on the container.

Perhaps whats needed is an initialisation script (JavaScript) that could be loaded to the container and used to configure the instance.

With Puppeteer 1.5 connecting fails with:

     Error: Protocol error (Target.getBrowserContexts): 'Target.getBrowserContexts' wasn't found undefined
      at Promise (node_modules/puppeteer/lib/Connection.js:86:56)
  From previous event:
      at Connection.send (node_modules/puppeteer/lib/Connection.js:85:12)
      at Function.connect (node_modules/puppeteer/lib/Launcher.js:257:50)
      at <anonymous>

Hey!

One can set a proxy like:

  const browser = await puppeteer.launch({
    headless: false,
    args: [
      '--proxy-server=PROXY_URI'
    ],
  });

Is there a possibility to set different proxies for different pages in the same browser instance?
Like:

browser1, page1 => proxy1
browser1, page2 => proxy2

Cheers

Hello,

I was wondering if it would be possible to create a tag on Docker Hub for the build with puppeteer-1.1.1
I'd prefer to use this over latest to prevent potentially breaking changes being pulled when I'm rebuilding my service.

Thanks a lot,
Andy

Repro steps:

• Render a page containing some Chinese characters such as the following: 中文樣本文本
• Load page with browserless + capture as pdf or png

Expected: Characters rendered correctly in output

Actual: Chinese characters rendered as missing glyph "box"

Is there a way to configure a proxy used by Chrome to make all the HTTP requests?

Hi,

As puppeteer doesn't support mp4 codec by default and the only way to fix that is to add a path to Chrome in puppeteer.launch. (see here and here)

How could it work with browserless?
Is there a way to launch browserless with a specific Chrome app?

Thanks!

The dependency joi was updated from 13.6.0 to 13.7.0.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

joi is a direct dependency of this project, and it is very likely causing it to break. If other packages depend on yours, this update is probably also breaking those in turn.

Your Greenkeeper Bot 🌴

Hi, just want to say you really are an inspiration towards an indiehacker path. Always good to see someone who’s hustling and making it work and living life on their own terms.