"The browser must not proxy or alter the network communication. Your browser must not do any of the following:
* Rewrite HTTP headers
The browser must have a reasonably complete implementation of web standards and browser features. You must confirm that your browser does not contain any of the following:
* Headless browsers
* Text-based browsers"
Sure, I see the "increased security" goal of protecting HTTP headers and allowing images and Javascript in the context of the "sign in" process that Google has implemented. However I also see the goal of not impeding Google's online ad services business which, at least in part, relies on images, Javascript and blocking automation after the user signs in.
I fail to see the benefits of these requirements outside of Google's sign-in process.
HN does not impose such restrictions. It is no less useful than Google, IMO. Imagine if HN required "a reasonably complete implementation of web standards and browser features" just to sign in.
I once read that Marissa Mayer, former Google VP, still uses Pine.
Bias disclosure: I am a text-only browser user; I prefer text-only software.
Chrome on iOS isn't even a real thing, it's just an app embedding Safari like every other app with a webview (because Apple's draconian policies forbid alternate browser engines). All the other Google apps also have to use Safari for any web stuff.
They might hide the URL, so I can't see where the link led me. Or they might disable the usual browser menu, preventing me from bookmarking a page. Or they might have the "close tab" interaction be in a completely different location, breaking any muscle memory.
If I'm following an HTTP(S) link, it means that I want to view it in a browser. I don't want to have some in-app view that can only return me to gmail, or to discord, or to google handouts, just because that happened to be where I clicked the link from. I don't care in the slightest whether the rendering engine is the same as the default browser. I care whether the user interactions are the same.
On iOS, yes, since Apple's rules forbid alternate browser engines. On Android, it's not unknown for apps to embed an alternate browser engine in it's entirely.
Even worse when the embedded web view is not properly detecting the user-agent and puts up a banner... prompting you to download the app (looking at you, NHL).
yeah and in the case of Fairemail on Android, the embedded browser is dumbed-down and lacking most features... and the option to disable this "feature" is kind of hard to find
This is regarding man-in-the-middle attacks. There is not attack on Lynx. Many sites do not functionin without JavaScript. Sad, but that's the way it is.
How can a prohibition on "text-based browsers" have anything to do with MitM attacks? The browser's chosen rendering method (text or GUI) is irrelevant.
Google is actually misrepresenting the danger in their blogpost, because they write "One form of phishing, known as man-in-the-middle,..." - this statement is clearly false because MITM is not phishing. Not in any stretch of the definition. So who knows what is the real justification.
I agree that the post is not very well written, but if you are a bit generous you can read them saying "phishing sites are doing MITM attacks that we have a very hard time distinguishing from CEF logins, so we are axing CEF logins altogether".
I'm appalled of the general direction Google is pushing the web to (an AD haven), but some of their points make sense.
I feel like detecting these environments is directly at odds with user privacy and anti-tracking; but I guess google has never been anti-tracking, so that's not too surprising. Still, I'm incredibly disappointed that they'd essentially require clients be fingerprintable to auth. I feel like this is just codifying an arms race between strengthening requirements and JS environment checks, and hostile embedders ability to emulate a real runtime, and taking legitimate embedders with less incentive to participate in the race down as collateral damage.
This has nothing to do with security and everything to do with banning tools like youtube-dl, wget and others; from the post:
> The browser must identify itself clearly in the User-Agent. The browser must not try to impersonate another browser like Chrome or Firefox.
> The browser must not provide automation features. This includes scripts that automate keystrokes or clicks, especially to perform automatic sign-ins.
Edit:
I feel like mentioning youtube-dl was a mistake. It's not just about youtube-dl.
This policy bars the use of all text-based browsers, headless browsers, browsers without javascript, and browsers with automation when accessing Google's services. It bans browsers that contain Node.js, even.
> The browser must identify itself clearly in the User-Agent.
I wonder what they'd think of my proxy which I have setup to (among other privacy respecting features) rewrite the User-Agent to "By allowing me access, you waive all rights and policies regarding my access." This is basically my form of EULA.
> The browser must not provide automation features.
LOL. This was obviously written by some tech illiterate law type, perhaps a first year law student? I fear to think what incompetent engineer working at google of all places would have come up with that verbiage . . .
Yeah, the browser itself is an automation feature. It automates the downloading of a file over HTTP, downloading of any dependent resources, managing of caches, and rendering of the result. That these are so frequently done that this automated pipeline is referred to as "a browser" doesn't mean that this isn't an extremely automated system.
> To protect our users from these types of attacks Google Account sign-ins from all embedded frameworks will be blocked starting on January 4, 2021.
(emphasis mine)
So I don't follow how this would have anything to do with banning youtube-dl, which doesn't require login? And as the blog post mentions, you can still bootstrap auth through a normal web browser, and pass the auth token to your command line / less secure browser / ... app.
(Disclaimer: I work at Google, not on anything related to this blog post or to your hypothetical scenario.)
Which would still be fine. The only thing that'd be blocked is obtaining those OAuth tokens by passing your Google username/password to a browser automation tool.
And it's also ineffective, because any full GUI browser can be automated with extensions or userscripts, so banning useragents or non-javascript browsers won't actually prevent automates sign-ins.
So they're breaking mountains of automated test and scripting tools? Nice to know.
>Headless browsers locked out (thanks for nreaking test automation)
>Node.js (riiight)
>Text-based browsers (Going aft Lynx? Of all things?)
>JavaScript MUST be enabled
Totally not a power grab here folks. Totally not a company known to value reaping user data in any form possible up to any kind of user hostile behavior, or exercising undue influence over the character of the Net.
That’s a bash script that runs via cron.
One thing to note: this uses the cookies from a logged-in browser session because at some point YouTube blocked password log in from youtube-dl. This was is a bit of a pain to set up, and I wish it was not the case, but it mostly works.
And you claim that doing more to stop people from giving their google account password to "random apps" (I personally trust youtube-dl a lot too, but "random apps" is what it comes down to) and forcing those apps to use OAuth to obtain scoped tokens has "nothing to do with security"?
That's only true if you assume the user is perfectly capable of evaluating the trustworthiness and quality of the software they want to use. It's understandable that that's not the assumption Google designs their security under. Yes, that sometimes somewhat sucks for us power users.
If that were all that they were doing I might agree; but they are blocking browser identity misrepresentation and automation, as well; it also requires that all "browsers" have a complete implementation of web standards.
It explicitly blocks "headless" browsers.
> You must confirm that your browser does not contain any of the following:
Google’s control...over the security of Google accounts?
If you are worried enough about Google’s dominance over the Internet to be upset by this particular practice, it is unlikely you have (or should maintain) a Google account.
I’m not a “Google stan” by any means, but to say that they want to take over the Internet is just not true.
Well I guess my unofficial YouTube chat bot won't work anymore. The YouTube API is awful compared to the Twitch one for bot creation so it is easier to get the functionality you want using Selenium.
Why not? This restriction was added on Android years ago and it basically means that you retrieve the OAuth2 token with a normal browser and then send it to your automaton script/app.
It blocks auth to prevent phishing, not actual access.
You'll need to find another provider which doesn't care that much about preventing phishing attacks. Google accounts are a big target so it makes sense you move away from the masses.
In practice it just means faking the user-agent and other fingerprinting more enthusiastically. I'm not sure how google can win that without resorting to the same anti-cheat measures as games companies.
You'll try, but the first time you won't know what fingerprinting tests they are going to do. After a few iterations you'll succeed, but it will be obvious to Google that the account you've just been testing it on belongs to someone trying to break their auth restrictions...
OAuth tokens used in automation tools will continue to work. Entering in username & password through auth, to automate an OAuth flow (or any other traditionally manual flow) will stop working. Breaks some puppeteer scripts too - but those have been getting flaky for a while now.
Yes I would agree with that, except that if you change a password you know the scripts will fail, but if an OAuth token gets invalidated by the system and not you, then it will fail without warning.
And if your password gets reset by the system and not you, same story.
What makes you say oauth tokens are any less robust? Aside from the fact they usually have an expiration attached to them, there's not much difference.
However, I will say to counter my own point: it’s not all roses. Using password by necessity means that your password is now stored somewhere that is likely more easily compromised (In my case: secure passwords are stored in 1Password, but passwords for script usage are either stored in an ENV or in the script itself, neither of which are great from a security standpoint)
> The browser must not provide automation features. This includes scripts that automate keystrokes or clicks, especially to perform automatic sign-ins.
If a web developer knows what they are doing they are using the standard web APIs supplied by the browser in an efficient way, designed to be invisible to accessibility for accessibility test automation, and thus this control from Google is largely unenforceable. As such I believe this is just a block against incompetent forms of automation that probably shouldn't be there in the first place.
The whole premise is a mistake, not the mention of youtube-dl. This policy doesn't ban the the things you claim it does. It is not 'everything to do with banning wget'. It's just a strange and mistaken conclusion you arrived at, seemingly by very selective reading.
Once upon a time Google would've been applauded for forcing people to improve their security. Like when they made https a ranking factor for sites and overnight forced all the laggards to move off http.
Now, people just scream "monopoly" at everything google does, good or bad and boy is it getting tedious.
Before, it stated that the company must not be evil. Now, it is for the employees not to be evil somewhere in the end. In the meantime, if you are high up the management ladder, you can fuck the subordinates and get away with millions.
The reason Google is requiring you to have JavaScript enabled, potentially opening you up to running malicious code, is because they want to use feature detection to prevent people from using competitors' Google-unapproved browsers.
They are not improving security, just making it more difficult to scrape google, who ironically is 100% based around scraping and living off other peoples content and creations in general.
Most of these controls are a blessing. They are blocking gross incompetence from front-end developers who don't know ow to do their jobs. I say this as a front-end developer.
If browsers which does "server-side rendering" are blocked from accessing my Google Account, I lose my access to all Google services requiring sign-in like Gmail etc.
I don't have the privilege to own a desktop, laptop, or even smartphone. I am using a J2ME enabled feature phone with Opera Mini to access the internet. Most websites requiring "modern browsers" are out of reach from me. Thanks to all the people who maintain the websites that are functional without JS or upto ES5.1 (last JS version supported by Opera's server rendering powered by Presto) or less. Only Google Search and Gmail works in Opera Mini, other Google services don't.
Rent a vps and run WRP [0] on it. WRP is basically a proxy that use chromium and render the pages as imagemap html pages that compatible with older/simpler browsers. It should works on opera mini. Hopefully google won't block it outright.
If I could rent a VPS, I could buy a smartphone instead. sigh
My j2me phone's screen resolution is 320x240, and Opera's server does a respectable job in transforming webpages to fit into that small screen size. It also uses a binary file format named OPML to encode the transformed page. With my 2G internet connection, it'd take much more time to load a page and cost me more to load images.
It meshes nicely for Google, because they want to use feature detection to detect whether you're using a Google-approved browser and not a competitor's unapproved browser.
This is why they state that JavaScript must be enabled, because that's how they do feature detection:
> The browser must have a reasonably complete implementation of web standards and browser features. You must confirm that your browser does not contain any of the following:
The only reason chrome isn’t mandatory is that there are still a few hold out browsers they can’t force out of the market.
Also, the requirement that the browser not lie about its identity in the UA means that the existing UA tests that google properties deploy everywhere means that those “acceptable” browsers may still be “accidentally” blocked.
It would be nice if people would start to acknowledge that chrome is the new IE and Google is the new MS.
Actually arguably worse: in addition to using free services subsidized by their primary advertising business. Once that business is gone they start charging.
All the while they grossly destroy user privacy, and come up with new specs that just happen to accidentally make tracking users easier. Generally poorly thought out ones to help single teams at google without any thought of what the general problem is.
Welcome to Google's private World Wide Web. Please ensure that you use the one and only official Google WWW client (others exist, but they are just for show). Unauthorized alteration of its configured operation will result in user termination.
One might wonder how that can accompany all the talk about open standards, and multitude of devices implementing different subsets, and responsive/adaptive/semantic design, etc. Then you realize that you don't really need, say, user-agent sniffing if you are already in position to dictate what browsers will and will not do, so into the trash it goes. You don't need interoperability hacks if you've stopped having interoperability problems.
Not much. If you're not using app passwords and your client wants to authenticate using Google auth (e.g. Thunderbird), it has to open the user's browser and setup the oauth flows instead of embedding the browser directly in the app.
It was recently (Oct 8) announced that that Google would provide a 12-month heads-up for stopping less secure app access, so it's my understanding that IMAP is not affected at this point.
I fail to see how they can possibly do this in a way that isn't trivially worked around by just embedding the same code as the full browser.
I guess their best bets are detecting non-fullscreen screen sizes on mobile, requiring Widevine or requiring Chrome and adding some proprietary authentication code, but all these are problematic and can be worked around.
Also of course both Firefox and Chrome support automation via WebDriver and WebExtensions so not quite sure what they plan to do with "The browser must not provide automation features".
However, if just to login in some application, it would be awful UX if going to the login step in an application triggers an unwanted load of 3 desktops full of 20 browser windows and a few hundred tabs, and some minutes delay while they all start up.
So if I'm not already running the "full browser" required for auth, ideally for authentication I'm going to want it to launch an "alternate profile" instance of my full browser which doesn't include all the other tabs or normal user info.
I.e. the browser should somehow be able to load just one special window for this application, and remember that it hasn't actually loaded my regular profile and saved state yet.
Clicking on any links for info that is logically "outside the application", that's what should probably lead to a regular full browser being started.
In the end, this ideal browser behaviour in response to an application requesting Google auth is much the same as using an embedded web view - except running separately from the application for security purposes so that it's UI isn't subject to application interference.
Given that's just a web view with security properties, why not instead allow auth to launch a "security instance" version of an embedded web view, one that is subject to guarantees from the OS/GUI security systems that it is running independently from the application which triggered its launch?
On Android, there's a feature called Chrome Custom Tabs (despite the name, it works with other browsers as well) which basically opens the default browser window in a restricted UI without most of the chrome and tabs. It shares the state and extensions though and it's meant as a replacement for these exact banned flows (on Android, webview logins are banned for years now).
I wonder if such interface could be exposed for desktop browsers.
If you don't like that google does this: stop using their products. Make the effort to choose, use and promote a service you think is doing a better job. If you so deem it necessary, tell Google why you're switching.
If people actually did something instead of just complain, companies like this would think pretty hard about their actions since it would harm their bottom line.
Same here. It's like I made a sharp, long-term investment. As the months pass, I enjoy the payoff: watching Google get worse and worse without it touching my life.
Google is becoming increasingly user-hostile, which is the wrong business phase to be in while your competition is on the rise. Other email providers are (finally!) getting almost as good as Gmail. Bing is an okay substitute for Google Search. YouTube has been strangled by subservience to advertisers and people are moving to Twitch and other places.
> The browser must not provide automation features. This includes scripts that automate keystrokes or clicks, especially to perform automatic sign-ins.
So... banning password managers? I’m not seeing how that’ll improve security.
Also, I wonder how they plan to enforce this. Presumably impacted browsers will just spoof the user agent, etc.
Haven't heard of any large scale phishing operations on CEF/Electron/whatever apps. Then again I'm not keeping up with infosec news. Are they a big problem?
I wonder what is the plan for MFP to email scanning. There's a potential of bricking hardware in this case (or not using Google Workspace, formerly knows as G Suite).
Disabling less secure apps has been postponed though because of covid.
So, uh, does this mean I won't be able to use IMAP with Gmail anymore...? It already complains at me about this for being less secure than webmail, but that doesn't seem to be covered in this announcement.
They need non-Chrome browsers to exist, to avoid accusations of a monopoly. Chromium is arguably a strategy around this, where you can have a bunch of browsers using the same (Google controlled) infrastructure.
Safari is an exception, since Apple won't accept giving that up in their products.
>Next step: The browser must not be anything but Chrome.
It seems like that'd be difficult without somehow dealing with Apple first, maybe by getting the government to force them to allow Chrome. Which could happen. Some of the "antitrust" stuff getting tossed around is already starting to get exploited by entities like advertisers, and not just big ones like Facebook, there were those EU ones recently. Like all power, Apple's focusing of its user's collective power can be used not just for bad stuff but for very good stuff as well. But that nuance doesn't seem to be present in a lot of the last year's discussions, and of course lobbyists will use the opportunity if they can.
Without that though or another big disruptive shift Apple misses, can even Google afford to give up on the entire iOS market? Even the Mac market perhaps, if Apple really wanted to push back against Google there they've certainly got the potential capability.
Restricting to just Chrome/Safari(Apple webkit) would still be really bad though. Even if they still allow Firefox, that would further formalize just 3 browsers with minimal further experimentation still possible. That'd be a real shame.
> It seems like that'd be difficult without somehow dealing with Apple first, maybe by getting the government to force them to allow Chrome. Which could happen.
But that would also imply they'd have to allow Firefox, and Brave, and Tor Browser. Which would certainly be worth the "cost" of allowing Chrome.
> Some of the "antitrust" stuff getting tossed around is already starting to get exploited by entities like advertisers, and not just big ones like Facebook, there were those EU ones recently.
All political coalitions work like this. If you're against DMCA 1201 then commercial pirates will be on your side. That doesn't mean they're your friends. They're not, and in fact are costing your side goodwill, even if your side is right in the end.
> Like all power, Apple's focusing of its user's collective power can be used not just for bad stuff but for very good stuff as well. But that nuance doesn't seem to be present in a lot of the last year's discussions
Because it's true of anything. Dictatorships are a wonderful thing if you're the dictator's friends, but that's hardly making a strong case for dictatorship.
FUD? You still do not see the frog boiling after years of pushing Chrome and mindwashing developers into creating sites that only work in Chrome?
Google is the one "spreading FUD". Just look at the downvoting going on in the discussion and you'll see some pretty obvious attempts to silence anti-Google criticism.
No way. (Theoretically on Android, but let's be real, they won't).
No way they are giving up marketshare on Safari or on corporate boxes with IE. They've paid so much to get onto iPhones, there's not a chance they'd risk any marketshare erosion.
This is pretty much the current step already, not the next step, with broad ban on anything that isn't a lot like Chrome and anything that they simply don't want to allow. In the last thread the narrative was hijacked with bullshit "security" justification, while in the blog post they ban much more broadly, any automation, text-based browsers, etc.
Somewhat unrelated, but Google already blocks me from my account all the time because they don't recognize my device because of privacy settings in my browser... they need a lesson about fingerprinting I guess.
I really dislike this notion of many internet companies of their own self-importance. To me the obvious example is a website that requires you to set up a very strong password and link a phone number. A user account is a two way street, the website should give you the tools for good protection, and you should use them if it matters. If it doesn't matter to me let me use a weak password. If it doesn't matter to me let someone hack my account, what do I care. And if the user doesn't care why does the website owner? Why should hackernews care more about my user account than myself for example? It could be argued this position of security maximalism is due to cutting costs on customer support, for account recovery, but as I understand it Google doesn't have customer support already.
Frankly, this is just wrong. Maybe for some circumstances, but in Google's case, they provide email.
When it comes to things like email, your account being compromised doesn't just affect you. Google let people send out emails from those accounts, so if a compromised account is used for spam, it hurts them reputationally as they are actively facilitating harm.
You might not care if that account is compromised, but they should.
1. Many uses are not computer experts and don’t realize they’re at risk. They won’t adopt extra security measures unless they need to.
2. No company wants to announce that a bunch of accounts were hacked. The excuse that “our users don’t care” would be widely criticized.
3. Well yes, of course companies want to reduce customer support costs, but guess who else benefits from not needing customer support? The customers. It’s better to avoid a problem in the first place than to have great mechanisms for resolving it.
The problem is that you have to have an account on google to participate in a number of communities. Because of this they have social scaling problems that might be fundamentally unsolvable and in their attempt to find a solution they've done things like this.
