Puppeteer Connection and Crash Problem

[kalimeraa]

Hi..

we have two issues one of them when we opened the firefox we had a crash page in order to skip crash page we changed to them user_pref("browser.tabs.remote.autostart",false); user_pref("browser.tabs.remote.autostart.1",false); user_pref("browser.tabs.remote.autostart.2",false); user_pref("browser.sessionstore_resume_from_crash",false);

then we skipped crash page.

when we had connected to puppeteer that lauched but thereis no going anywhere and then

we saw TimeoutError, timed out after 3000ms while trying to connect to firefox

[plaperdr]

Hey, Thanks a lot for looking into FPRandom! Are you using a recent version of Firefox (i.e. did you recompile a recent version of Firefox with the patch)? Since FPRandom was built 3 years ago, I'm not expecting it to work with Puppeteer.

[kalimeraa]

thanks for answering but i im using recent version firefox on my ubuntu os and i setted puppeteer executablepath your firefox but I did not recompile and I don't know how can I compile @plaperdr

[HMaker]

@kalimeraa the last release of patched Firefox found in this repo is based on Firefox Nightly 56 which had no support to a subset of Chrome DevTools Protocol (CDP) required by puppeteer to connect and control it. Now the latest Firefox Nightly releases supports that subset of CDP, called internally "Remote Protocol". You need to take Firefox's source code, apply the fprandom.patch with git-appy for example and recompile it.

[HMaker]

@plaperdr Could not that anti-fingerprinting patch to Firefox be done with JS?

[plaperdr]

@HMaker the protection provided by FPRandom could be done in JS but the biggest downside is that it is often detectable. That's why a browser version is preferable.

[HMaker]

@plaperdr I think property enumeration order can't be changed from JS. I never compiled firefox, does it takes much time?

[plaperdr]

Ah, you are definitely right, I simply forgot about that. I'd say that it takes about an hour to compile Firefox from scratch.

[HMaker]

Ah, you are definitely right, I simply forgot about that. I'd say that it takes about an hour to compile Firefox from scratch.

Later I found it's possible with JavaScript Proxy, but requires every navite object to be wrapped by a proxy which can make browser slower. Also I found known pixels read-out techniques are used to detect spoofed canvas, like draw black canvas and retrieve the pixels later to test if it's really black, is FPRandom able to bypass that kind of tests?