Seems unlikely they'd want to sabotage Firefox.

Google is Firefox' primary source of income, and Firefox is strategically crucial token competition that stands between Google and an antitrust lawsuit of the sort Microsoft faced in the early 2000s (where Microsoft were basically a technicality away from being broken up).

It does seem unlikely, but playing with the example UA in curl (as noted in https://news.ycombinator.com/item?id=38925015) seems to show it's the combination of "Android" and "Firefox" that causes this behaviour. I wonder if any other mobile browsers are broken?

A far more likely explanation is simply that two features or workarounds are enabled, maybe "Android" triggers a prompt to install Chrome or something and "Firefox" triggers some workaround code. However these interact poorly and cause a crash.

If they wanted to sabotage Firefox I doubt they would choose to send a blank page. They would probably send infinite captchas or just disable most features (which they already do).

Some middle manager out there was making this decision. Not strategy heads

Strategy heads come up with the strategy - friction. Enough friction against Firefox will let the browser still exist (essential for the claim that there is competition) but without actually being able to compete.

Middle management finds a tactic that implements the friction strategy - small "random" breaks and persistent performance issues.

An engineer would find the precise measure to implement - break the UA string sniffing targeting a specific browser.

> Enough friction against Firefox will let the browser still exist (essential for the claim that there is competition) but without actually being able to compete.

Interesting theory, but that's exactly what M$ got dinged for in the antitrust suit re: Netscape.

Google doesn't care, they knowingly implemented the mother of all illegal antipattern cookie banners and only changed it after getting fined over $100M.

If the risk is a lawsuit after years of anticompetitive behaviour - if they're unlucky - that's absolutely worth it for Google. Chrome is now THE web browser platform with the only two remaining notable exceptions being competitors in name only, and no single lawsuit can just reverse that.

MS waged a lot more open war against Netscape, and documented their desire to "extinguish", "smother", and "cut off Netscape's air supply".

Google is doing more "eternal irritation by 1000 papercuts". Never one big blow, never aiming to kill Firefox, they're financing it after all. Firefox needs to exist but never be attractive to users, especially on the phone where tracking is next level.

I think plausible deniability does a lot of the heavy lifting here. But statistically speaking Firefox being the one browser hit constantly by a stream of random Google issues can't be random.

The legal and regulatory landscape changed a lot since then too, with Big Tech slowly but constantly lobbying and pushing a lot more than just these tactics into normalcy. A lot of what's normal today was outrageous in the '90s.

Sometimes! And sometimes a middle manager just does what he thinks clears the ticket quickly enough and then someone yells at him.

Assuming competence and intention is foolish.

> Middle management finds a tactic that implements the friction strategy - small "random" breaks and persistent performance issues.

> An engineer would find the precise measure to implement - break the UA string sniffing targeting a specific browser.

You got the strategy right but the implementation is laughable, sorry :-)))

The implementation is: "We have a budget of N story points this sprint to resolve bugs, let's prioritize them. Let's prioritize by impacted audience size."

The audience size will make 99% of Firefox specific bugs be deprioritized out of the current sprint. And the next one. And the one after that.

And unless a senior engineer stands up to update the prioritization criteria, plausible deniability forever.

The aim was to lay out at which level each stage happens. It's not middle management taking a random decision. The fish rots from the head but the tail certainly doesn't smell rosy either. Dieselgate was also just a random deprioritized code bug until it wasn't.

Repeated incompetence from one level is actual maliciousness from the level above, all the way to the top.

They want two things: 1) Firefox needs to exists as a hypothetical option 2) People shouldn't really use Firefox. Intermittent oopsies breaking key services are doing wonders towards goal #2

Google is currently facing an antitrust suit where their payment to Firefox is a crucial part of the evidence against them. And it's over their search engine, which is far more important to them than Chrome.

They pay Firefox as they're buying a valuable product. Which is why they also pay Safari.

Are you certain that search is more important than chrome?

Sure search was their first product, but they have long since pivoted to being an ad company. And while yes, they can show ads along side search, the real cash cow is all the juicy data they can exclusively hoover up through Chrome to better target ads all across the web not just search.

Nope, search ads are still where they make at least 2/3rds of their money. The rest of their ad inventory has substitutes, search really doesn't (because they basically invented the category (grumble grumble something Overture)).

Chrome is okish for collecting data. But they have Android which probably provides 100x the data.

Only on poor people! (I mostly jest, but having talked to a bunch of former Google ad people, it was a real concern for them that most of their high revenue users were moving to iPhone).

And indeed, we are talking about Firefox on android here.

It's never been clear to me that Google is actively trying to take-out Firefox. Firefox's user base is pretty small - so if they succeed in reducing it they'd gain very little but lose some shielding from antitrust accusations.

I suspect that these problems are more to do with neglecting testing, and just not caring very much about non-Chromium browsers

(Disclosure: I'm a long-time Firefox user)

The thing about "high quality malice" is it's indistinguishable from an error.

A bug that breaks others' tools but not yours is not implemented intentionally, but it's not fixed intentionally, so it looks like a benign error.

At the end of the day, you use the chance you get, and it makes real damage while you sip your coffee...

It’s pretty clear to others. Chrome is supposed to be The Web in the same way that Google is search.

Antitrust is just money to pay for Google. Firefox is cancer to Google.

I do not know it that is true. Cancer tends to grow... ;)

well played, well played :)

I get why people want to assume bad faith but it's pointless, it's hard to intentionally sabotage Firefox when nobody tests shit in Firefox at Google. There's (or at least was) still an internal group of Firefox users a few years ago, but it's small, and internal tools may have poor or no support for Firefox at any given time. When I worked there I made some attempts to fix things (e.g. once Google Cloud was accidentally broken on Firefox during a time when many people were out...) but the bigger problem is that most people don't use Firefox because it's inconvenient or they don't care to and a lot of the tooling only works on Chrome/only works well on Chrome.

There's a silver lining though: it really doesn't matter if it's due to negligence or malicious sabotage, and if they (where "they" is hypothetical leadership within Google that is trying to eliminate Firefox) thought that pleading negligence instead of sabotage was going to be a good defense, we'll see how that holds up to future scrutiny and regulation. I mean hey, Microsoft tried to pull a lot of shit too, and they were probably more competent at it if the Epic Games lawsuit is any indicator of Google's competence right now.

To me, the fact they don't run automated checks on Firefox is in itself malicious and anti-competitive.

Yes, deciding not to test in Firefox is a decision to break Firefox. Sites with a tiny fraction of Google's resources manage to test Firefox. There's no way they just forgot.

Similar to the google image search on Firefox mobile, which has been broken for how many years now?

Well image search is broken in general.

Reverse search is now completely useless for example.

There's a much easier explanation.

They didn't test on it at all. It's a browser with sub-5% market share.

There are folks in Google who personally test on firefox and believe in it ideologically as an important peace of the web ecosystem, but that's not a company policy and detecting Firefox bugs does not gate feature release for most projects.

That's just sabotage with extra plausible deniability.

Is it "sabotage" when you don't intentionally devote resources to making some random third party's browser work? What obligation do they have?

Mozilla has cachet with nerds like us because of its history. Not because it's particularly better than alternatives or because it has enough of a user base to throw its weight around.

In contrast, they do test on Safari. They'd lose double-digit percentages of users if they didn't.

The ecosystem, unfortunately, generally has room for three top browsers. I'm not even sure Firefox is number four these days.

Except that firefox is particularly better than alternatives. It isn't sending my browsing history to Google, and actually lets me block ads. Firefox provides a level of customization and options to protect your privacy that chrome will never match.

> It isn't sending my browsing history to Google

For obvious reasons, I don't think Google is devoting any resources to optimizing the experience of navigating to google.com for people who don't want any history generated on their use of google.com.

I agree with you that the other features listed are positives, but they don't appear to be positive enough to push use of Firefox psst Safari, Chrome, or Edge numbers.

Google isn’t sitting around randomly breaking things for Firefox users in this obvious way.

I really feel like HNers often forget how minuscule the scale of their usage patterns are.

Google doesn't need to intentionally break things, they just have to not test on anything but Chrome and say "oops" once a critical mass of complaints come in. Now you could say that they don't have to support the competitor, which is fair, but the web is based on standards, and in this case the site breaks based on the user agent and not the lack of some technical capability.

How convenient!

The title is bad, this only affects Firefox mobile that AFAIK has a really low amount of users (BTW, I am one of them).

Not saying that they shouldn't have detected this (i.e.: where is the automated testing for those things?), but I don't think this is Google screwing up Firefox on purpose.

And if Google really wanted to screw up Firefox, they would probably do a better job than User-Agent sniffing. Firefox already has an internal list where it applies fixes for websites (e.g.: setting a custom User-Agent), and the bug report actually comes from it. You can see it by going to `about:compat` page in Firefox.