A better escape hatch is https://mozilla.org

Granted, Mozilla's had their own issues and some apps / sites may have become as dependent on Chrome as sites were on IE way back when. But given Google's flexing their control muscles over the last few years, a clean break is probably the better long term solution.

Question: How do I sign extensions I wrote for myself so I can actually use them, without uploading a copy of the extension somewhere? That's my roadblocker right now. I don't know how to migrate make my own extensions work on my own system without uploading them to other people's servers. (Yes, I want Firefox stable. I don't want to run a buggy browser.)

You can run either Firefox ESR or the unbranded builds [1] if you want unsigned extensions, but I'm curious what your concerns are with the unlisted extensions option from Mozilla?

1 - https://wiki.mozilla.org/Add-ons/Extension_Signing#Unbranded...

> You can run either Firefox ESR

Then I have the exact same problem when support dies in the coming months though.

> or the unbranded builds [1]

Interesting, so does this mean that if I want stable builds, I'd have to manually update to the exact unbranded counterpart of that stable version?

> I'm curious what your concerns are with the unlisted extensions option from Mozilla?

I guess because I really don't see what business they have forcing me to upload my extensions to their servers? Same reason why people don't like uploading Chrome stuff to Google servers?

~~~~~~~~~~~

Edit:

I'm trying out the unbranded builds and running into a problem when I try to use my existing profile: disabling signature checking doesn't work. However, it works on a fresh profile. Not sure what the problem is...

> Then I have the exact same problem when support dies in the coming months though.

Good news! Even Firefox ESR 60 always you to disable the mandatory signature checks by setting the "xpinstall.signatures.required" pref = false. The Mozilla add-on wiki was a little unclear, so I just manually verified that the "xpinstall.signatures.required" pref works correctly in Firefox ESR 60. :)

https://wiki.mozilla.org/Add-ons/Extension_Signing

I know this is exactly what you're trying to avoid, but could you upload an example unsigned extension? I'll try to figure it out, could be a bug somewhere.

Thank you for offering to help! I actually just figured out how to fix it right now myself. The solution is to delete the following lines in prefs.js, then restart unbranded-Firefox twice:

  user_pref("extensions.databaseSchema", 27);
  user_pref("extensions.e10s.rollout.policy", "50allmpc");

The first start takes longer, so I can tell it's resetting something internally (not sure what). However, the second restart is what makes it start working. Not sure why.

Edit:

For anyone else trying to auto-obtain the latest stable Firefox nightly, here's a hacky command (change for your OS as needed):

  curl -s "https://wiki.mozilla.org/Add-ons/Extension_Signing" | sed -n "0,/Windows 64-bit Installer/{s/.*href=\"\\(https:[^\"]*\\.zip\\)\">Windows 64-bit Installer<.*/\\1/p}"

Glad to hear you worked it out, good luck! Looking around there are a few levels of cache around whether addons are enabled, I guess those are what you need to clear it.

It's a temporary hack, but you might also want to look at temporarily loading extensions with about:debugging, this works even in official release.

Yeah the temporary thing was already working in the official release, but it's temporary so I was looking for the normal way. Thanks! :)

You have to use Firefox Developer Edition and change 'xpinstall.signatures.required' to True in about:config

What about the part where I wrote

> Yes, I want Firefox stable. I don't want to run a buggy browser.

? Is the developer edition running the same code as stable?

Developer Edition is essentially Beta with a few changes (which means additional telemetry by default so probably not up your alley):

https://hacks.mozilla.org/2017/04/simplifying-firefox-releas... (edit for better link)

Yeah, I neither want beta nor telemetry...

fwiw even Nightly is quite stable these days, I can't remember when it last kept me from getting work done. And the addt'l telemetry is just a pref away.

I feel this should be weighed into the decision: https://medium.com/@neothefox/firefox-installs-add-ons-into-...

When you're looking for escape hatches, you know the software is working against you. At that point, it's better to move to an alternative browser, it's not worth it to fight it with hacks. If too many people use the escape hatch, they'll remove it.

It seems as though the needed escape hatch is a browser that simply CAN'T send info about you back to home base. Firefox might qualify, but I don't know for sure, and I don't want that to be our only option. There's a lot of great engineering that goes into Chrome.

So I'm wondering: Is enough code open sourced to make it possible for an organization such as Apache, GNU, or whatever, to create a sanitized fork of each new version of Chrome? This would be a version of Chrome with no privileged party other than its user. Every other party would be limited to the same set of features (cookies, local storage, etc.) that apply to all 2nd parties.

You might want to look into the Brave browser (https://brave.com), which is a commercial browser based on the Chromium code, funded by a micropayments idea, which tries to generally be more privacy-sensitive than the Google version.

As for a fully open-source version: here's a fairly recent summary (2017) of what's in Chrome and not in Chromium (the open-sourced subset):

https://www.howtogeek.com/202825/what%E2%80%99s-the-differen...

There are other omissions that could get in the way of general usability; chromium also omits some proprietary licensed codecs, including the mp3 support.

However, Google synch support, etc, is left in -- and if you desire to keep that out, maintaining the patch set that does it without cooperation from upstream could be a hassle. Whenever they refactor of update the integration between synch and the rest of the browser, you need to make corresponding changes in the patch set that cuts all that out -- and at a fairly rapid pace, too, lest you find yourself unable to ship an urgent security update.

So, it's doable -- but at a price in usability, and doing it right requires some kind of continual effort. Which is what Brave is, in effect, trying to arrange with their payments-based funding scheme. And even a very capable person trying to do the project solo, on a volunteer basis, would be putting themselves at risk of burnout.

Does one really want to associate themselves with Brendan Eich though? Brave is what he has been building after being ejected from Mozilla.

While Brendan has built some interesting and useful things including Javascript, he is apparently poisonous enough to cause three of Mozilla's board members to resign, including a founding executive of Yahoo, the former CEO of AVG, and the former CEO of Mozilla Corp (from 2008 to 2010).

His wiki page is a good reference: https://en.wikipedia.org/wiki/Brendan_Eich

You're repeating a lot of misinformation. Here's a better reference:

https://www.cnet.com/news/mozilla-under-fire-inside-the-9-da...

The disagreements were about strategy and mobile experience, so not enough to call someone poisonous. Place Yahoo, AVG, and Moz Corp against Brendan's co-founder status and JavaScript.

The other toxic drama was about a 1000$ political donation. Very sad to see an unpopular view used to paint someone as a monster. If democratic political views are enough to kill someones career, it is more a sign of the toxicity of today's social media and activism than a character judgment.

> Does one really want to associate themselves with Brendan Eich though?

No, not really. Not a good political idea to align yourself with the black sheep. With all the tars and feathers he just looks weird.

Though it also works in his favor: Brave is popular amongst the alt-right and technical-minded early adopters. They see it more as character assassination and SJW corporate culture taking desperate vengeance for Trump's win.

The fact that Eich was pilloried after the fact for what was at the time a mainstream political opinion is absurd.

Brave is fantastic on android though.

> at the time

It was ten years ago. When people fussed about it it was six years old.

How many years back do political actions count as relevant?

> Is enough code open sourced to make it possible for an organization such as Apache, GNU, or whatever, to create a sanitized fork of each new version of Chrome?

Yes: https://wiki.qt.io/QtWebEngine

> Qt WebEngine uses code from the Chromium project. However, it is not containing all of Chrome/Chromium: Binary files are stripped out. Auxiliary services that talk to Google platforms are stripped out. The codebase is modularized to allow use of system libraries like OpenSSL. We do update to the latest Chromium version in use before a Qt release.

Here's a browser: https://www.qutebrowser.org/ I haven't tried it yet.

[edit]

oh, it's already in the apt repos on debian testing :)

If it's connected to the internet, then it can send info to home base. The only way around this is code that's simple enough to audit and understand.

That ship sailed long ago for browsers -- if you care about auditability, you can't trust the web.

Exactly. I'll go further in saying a system or software should be considered insecure if it connects to Internet and wasn't designed/implemented with strong security. That's most stuff.

That's a red herring. We're not discussing a security bug -- the system is working as designed. The problem is that the goal is not desirable, and the standards and code are complex to the point that an independent implementation is intractable.

This doesn't answer your question, but I thought I'd mention: A lot of great engineering goes into Firefox too. I really think you ought to try the version of Firefox that enables WebRender by default, when it comes out. I've been using it in Nightly on Mac, and it's seriously worth trying.

> a browser that simply CAN'T send info about you back to home base

Unfortunatly that requires a large sacrifice: no Turing complete languages.

https://news.ycombinator.com/item?id=15708099

That was also my argument against all the "hacks" people posted online against Windows 10 tracking. Most of them were short-lived because Microsoft had an incentive to stop them from being used in the next update of Windows.

> When you're looking for escape hatches, you know the software is working against you.

I would say that when you're looking for escape hatches, you know the defaults don't match your personal preferences. I understand why some people don't like combined login. However, I personally prefer it, and I guess that the decision to add it was made on the basis of believing that most people would feel like I do, or at worst feel neutral.

Based on all the discussions I've had or seen with Google developers, they added it based on the basis of believing that most people should feel that you do, or at worst won't notice. It's unlikely that they considered privacy concerns, above and beyond "we can be trusted so it's fine".

I don't know who you are or whether you have had conversations with the people who are responsible for the feature. All I can say is that that does not match my perception of the typical way of doing things at Google, as someone who has been working there for a while. That said, I do not sit close to where this feature was built, so maybe they do things differently over there.

I know. I was just trying to help another poor soul like myself in the interim of figuring out how to switch. For me at least, it's very nontrivial work.

That is how I feel about my use of youtube-dl.

But youtube is so damn entrenched nowadays...

Maybe they've revised the flag, but the flag did not disable the feature as much as the text might lead you to believe. My comment from a couple weeks ago... https://news.ycombinator.com/item?id=17943524

Thanks for letting me know! For some reason I don't see that on my toolbar, though I haven't tried creating a new profile. I'm on Chrome 68 on Windows, if that might explain anything?

This flag is not really working. Just checked it (chrome 69.0.3497.100)

Aw man. It's working on 68! Guess I have to avoid upgrading for now. Thanks.

So you're choosing a less-secure version of Chrome over just switching to Firefox? What makes you so loyal to Chrome?

Your comment is aggressively presumptive (and wrongly so) so I'm not inclined to respond directly. You can browse my last few comments if you're genuinely interested.

Doesn't work for me.

What version?

I tried 69 at the time. Haven't checked for a few days.