Fourth [LWN.net]

Fourth

Posted Dec 16, 2025 20:40 UTC (Tue) by pizza (subscriber, #46)
In reply to: Fourth by mb
Parent article: Mozilla gets a new CEO: Anthony Enzor-DeMeo

> It simply doesn't work correctly with many modern web sites anymore.

That's because sites only bother testing against Chrome now. And maybe Safari.

Not much Firefox can really do about that other than making Firefox just another reskin of Chromium.

to post comments

Fourth

Posted Dec 16, 2025 20:57 UTC (Tue) by davecb (subscriber, #1574) [Link] (3 responses)

Yes, this is called the "modem testing problem"*.

Were I working on Firefox (I'm not), I'd build a mechanism with something like Playwright to download a url and compare the two files, alerting on things that responded differently. Then I'd review the difference and decide if it deserverd to be changed in Firefox.

A customer used to do something quite similar, when they needed their program to work on numerous browsers.

--dave
[https://leaflessca.wordpress.com/2017/01/03/the-modem-tes...

Fourth

Posted Dec 16, 2025 22:35 UTC (Tue) by pizza (subscriber, #46) [Link] (2 responses)

> Then I'd review the difference and decide if it deserverd to be changed in Firefox.

The answer will _always_ be "we must be bug-for-bug compatible with Chrome"

(Because even if Firefox is "correct", that doesn't change the indisputable fact that it "works just fine" in every other browser)

How to be compatible with Chrome?

Posted Dec 17, 2025 15:32 UTC (Wed) by DemiMarie (subscriber, #164188) [Link] (1 responses)

The question is how to achieve this.

Should Firefox be another Chromium-based browser? Or is there a way to make this work with Gecko?

How to be compatible with Chrome?

Posted Dec 18, 2025 8:56 UTC (Thu) by taladar (subscriber, #68407) [Link]

If they decided to replace Gecko with the same engine every other browser is using they would lose most of their remaining user base. Their one major advantage right now is that they are preventing a total browser engine mono-culture.

Fourth

Posted Dec 16, 2025 21:10 UTC (Tue) by mb (subscriber, #50428) [Link] (26 responses)

Yes.
But I'm wondering what Firefox developers actually do when they hit these problems by themselves.
They must surely see these problems frequently, too, right? I assume they actually all use their own product.

What are they going to do if they hit such a problem?
Are they going to fire up Chromium to get the thing done quickly, because they must get back asap to implementing the next "AI" feature for Firefox?

I do get it that website developers don't test on Firefox anymore.
But then the only chance for Firefox to survive is to become bug compatible with Chrome.

Priorities are set incorrectly.

Fourth

Posted Dec 16, 2025 21:19 UTC (Tue) by josh (subscriber, #17465) [Link]

Yeah. The only way to fix this is to become more popular. Mozilla might think the way to do that is "AI", but then they're in a feature race with one of the primary AI companies, and they're going to lose that race. Meanwhile, they're bleeding trust every time they do something like this.

Fourth

Posted Dec 16, 2025 21:24 UTC (Tue) by ptime (subscriber, #168171) [Link] (12 responses)

It’s quite possible they aren’t hitting them. I use Firefox for a lot of browsing, and I can’t tell you the last time I encountered a broken page that wasn’t just being broken by pi-hole or ublock

Fourth

Posted Dec 16, 2025 21:29 UTC (Tue) by pm215 (subscriber, #98099) [Link]

Mmm, I think I encountered a hotel website three years ago that seemed to need chrome, but that was it.

It probably depends very heavily on your browsing habits -- I expect 90% of mine is the same quite small set of websites, and I don't have a job that requires me to deal with a lot of corporate or internal-IT web apps. There could be lots of chrome-assuming websites out there that I simply never see...

Fourth

Posted Dec 16, 2025 22:11 UTC (Tue) by dskoll (subscriber, #1630) [Link] (8 responses)

Yes, I haven't encountered a website that refused to work with Firefox in ages.

Actually, one exception was my bank website. They did user-agent sniffing and I used a spoofed user-agent that they didn't like, so I had to un-spoof it for them. But that wasn't a Firefox-specific thing,

Fourth

Posted Dec 17, 2025 6:32 UTC (Wed) by ebee_matteo (guest, #165284) [Link] (7 responses)

As somebody using Firefox with Microsoft 365 and MS Teams daily, there are several small and annoying issues. Even to the point that editing a Word document jumps backwards in the text while auto-saving during typing. Or audio breaks down in calls for no reason. Or email writing in Outlook is janky.

SAP is another thing that only works in Chrome/Edge.

The majority of employees at European companies rely on Microsoft products for their business. That's million of users, 8 hours per day.

Sure, as a Linux embedded dev I am more than willing to take the pain to have a browser that doesn't resell my data. I can work around problems. I am also in the 1%. Corporate? They just want to reduce calls to the help desk hotline.

Fourth

Posted Dec 17, 2025 8:58 UTC (Wed) by eru (subscriber, #2753) [Link]

All features of Office365 pages really work only on Microsoft's Edge (to the extent they work on Linux at all). No surprise there. Edge is the new IE. Leopard can't change its spots...

Fourth

Posted Dec 17, 2025 13:08 UTC (Wed) by ptime (subscriber, #168171) [Link]

I use O365 on Firefox Nightly every day and don’t have any issues

Fourth

Posted Dec 17, 2025 21:10 UTC (Wed) by mathstuf (subscriber, #69389) [Link] (4 responses)

I have definitely noticed jank in Teams on Firefox, but I'm going to blame it all on Teams. I don't know what it is like recently because I've been using a macOS machine for video calls (still Firefox though; everything I've tried works fine) because Dell "wisely" used IPU7 camera hardware in this model which works poorly on Linux.

My observations were once:

- everything except video working with uBlock enabled
- if there is a whiff of a Microsoft login cookie, I was blocked because "Teams was disabled for my account" (nevermind that anonymous users could join the call…)
- in private browsing, the login page would infinitely redirect
- in a normal session (fresh-ish profile), it mostly worked (I think background blurring/replacement wasn't supported at the time)

Fourth

Posted Dec 17, 2025 21:21 UTC (Wed) by mb (subscriber, #50428) [Link] (3 responses)

> but I'm going to blame it all on Teams

Yes. That's the problem.

Firefox used to work perfectly fine between say about 2010 - 2020, roundabout.
Now we are back to the bad old times where websites only work correctly with certain browsers.

The answer to that is *not* to just ignore it and say it's somebody else's fault.
Even if in fact it *is* somebody else's fault.
That doesn't help Firefox.
It kills it.

And to the people saying that it's plugins/ublock that break sites:
No, it's not. Almost never does disabling ublock unbreak sites.
This is a fundamental problem.

Fourth

Posted Dec 17, 2025 23:33 UTC (Wed) by ptime (subscriber, #168171) [Link]

I was trying to rent a storage unit today, the site was “broken,” and turning off UBO fixed the site.

Fourth

Posted Dec 17, 2025 23:39 UTC (Wed) by pizza (subscriber, #46) [Link] (1 responses)

> The answer to that is *not* to just ignore it and say it's somebody else's fault.

Pray tell, WTF is Mozilla (or indeed, any of us) supposed to do about Microsoft deliberately choosing to degrade functionality on browsers (and/or platforms) other than their own? What will making Firefox into yet another Chrome reskin accomplish?

None of us use Teams _willingly_. Instead it's a requirement for $dayjob. Complain and at best you'll be either ignored -- more likely you'll just be told to "use the app or other supported platform" [1] and/or get written up for violating IT policies. [2]

[1] my new employer's SSO requires Chrome or Edge along with a couple of plugins in order to even _log in_.
[2] processing company data with unapproved software

Fourth

Posted Dec 18, 2025 9:37 UTC (Thu) by tzafrir (subscriber, #11501) [Link]

For me video calls are broken with Firefox and work with Chromium: both on Teams and on Jitsi (and on a trivial html demo file I tried, to see using a camera). This started recently, not sure exactly when. Maybe related to playing with libcamera, and maybe related to another upgrade. Have not had the time to debug this further. Video calls have generally worked better with Chroium than with Firefox: this is not just an issue with Teams.

Fourth

Posted Dec 17, 2025 11:42 UTC (Wed) by mote (guest, #173576) [Link] (1 responses)

I have a persistent problem using https://www.heb.com in Firefox (typically private browsing but persists to non-private). The home page loads, you can load a single product once, then refresh once or twice and you're not blocked from the entire site.

They used to just hit you with a network denial type message, but now they emit JSON blaming your browser (incorrectly in my Firefox case, disabling ublock/etc. doesn't do anything).

```
hostName "www.heb.com"
errorCode "15"
description "This page could not load. It looks like an ad blocker, antivirus software, VPN, or firewall may be causing an issue. Try changing your settings, switching to a different device, or use the My H-E-B app"
```

I have to use Chromium to use HEB's website even to browse anonymously.

Fourth

Posted Dec 17, 2025 12:06 UTC (Wed) by farnz (subscriber, #17727) [Link]

I've just tried it in Firefox 146 (from Flathub), with uBlock Origin installed and active, and I have no problem browsing, up until the point I need an account to continue (add to cart/add to list); I've not bothered with account creation since I'm an ocean away from my nearest H-E-B branch.

It continues to work just fine if I disable uBlock Origin, too.

Fourth

Posted Dec 16, 2025 23:05 UTC (Tue) by jond (subscriber, #37669) [Link] (10 responses)

> But I'm wondering what Firefox developers actually do when they hit these problems by themselves.
> They must surely see these problems frequently, too, right? I assume they actually all use their own product.

I’m not a Firefox dev, but I am a Firefox user and I never have a problem where a website doesn’t work with Firefox and I have to launch another browser. I’m not sure how common that issue is.

Fourth

Posted Dec 17, 2025 4:03 UTC (Wed) by xanni (subscriber, #361) [Link] (9 responses)

I've encountered quite a few of these problems, but just to pick one that has a happy ending: Patreon discussion groups used not to load in Firefox, and the image carousels used to display 50% offscreen. Both of those issues persisted for a year or two but have recently been resolved and now work perfectly, though I don't know how much is Firefox improvements and how much is site updates. Patreon is fairly widely used!

Fourth

Posted Dec 18, 2025 11:50 UTC (Thu) by kpfleming (subscriber, #23250) [Link] (8 responses)

Portions of the Monoprice website have been broken in Firefox for a few years at least, but work fine in Chromium.

For example, go to this page - https://www.monoprice.com/product?p_id=2268. After the page has finished loading, click a radio button to switch to a different color. The page content will switch to the new selection, but the 'overlay' which dims the page and shows a spinner never goes away, so you cannot interact with the page. This functionality works as they intended in Chromium and other browsers.

Fourth

Posted Dec 18, 2025 12:46 UTC (Thu) by karkhaz (subscriber, #99844) [Link]

For cases like this, you can right-click on the modal spinner and then press "Inspect" (or "Q"), which displays the developer tools showing you the code for that element. In this case, the class of the modal spinner is "mp-spinner-overlay", so you can use the Stylus extension to remove it from the web page:

.mp-spinner-overlay {
  display: none;
}

This permanently resolves the problem for Monoprice. It's good for modals that break interaction, but also popups like newsletter signup forms, "Sign in with Google/Facebook/etc", and so on.

Fourth

Posted Dec 18, 2025 13:01 UTC (Thu) by farnz (subscriber, #17727) [Link] (4 responses)

This is another case where Firefox 146 from Flathub doesn't reproduce the problem - it works fine for me with, or without, uBlock Origin active on the page.

I wonder if you and mote are using a build from the same supplier? I'm not seeing the reported problems in Firefox from Flathub, and haven't in previous versions, either - which to me suggests that there might be an underlying difference worth tracking down.

Fourth

Posted Dec 18, 2025 13:13 UTC (Thu) by jzb (editor, #7867) [Link] (3 responses)

I'm using the Fedora build of FF and I see the problem as well. If I run a Firefox build from Mozilla, I don't see the problem. Probably worth filing a bug against the package.

Fourth

Posted Dec 18, 2025 13:18 UTC (Thu) by kpfleming (subscriber, #23250) [Link] (1 responses)

Bizarre... I am also using the Firefox build from Mozilla (Debian packages), and I do see the problem. Presumably that means the problem is caused by some interaction between Firefox and the extensions that are loaded, or something else in my profile. I'll try again with a fresh profile to see if it persists.

Fourth

Posted Dec 18, 2025 15:51 UTC (Thu) by johill (subscriber, #25196) [Link]

Indeed. I'm also using Fedora build and it works fine :-)

Most likely different extensions loaded for all of us, and even in the different builds jzb tried?

Fourth

Posted Dec 18, 2025 15:49 UTC (Thu) by rbtree (subscriber, #129790) [Link]

Report these issues to Mozilla when you see them. "Menu: Help → Report broken site", or "hamburger: Report broken site". It takes very little time.

Fourth

Posted Dec 18, 2025 16:54 UTC (Thu) by excors (subscriber, #95769) [Link] (1 responses)

That one happens when Firefox's enhanced tracking protection is set to "strict", and you don't enable "fix minor site issues": https://bugzilla.mozilla.org/show_bug.cgi?id=1962420

You should be able to add per-site ETP exceptions by clicking the shield icon in the address bar. Or turn on "fix major/minor site issues" globally in the settings (which slightly reduces the tracking protection, but appears to fix Monoprice).

Strict mode is advertised as "stronger protection, but may cause some sites or content to break", so the breakage is not considered a Firefox bug - it's a deliberate tradeoff for privacy.

Fourth

Posted Dec 18, 2025 16:57 UTC (Thu) by kpfleming (subscriber, #23250) [Link]

Thank you! I've had ETP enabled for so long I wasn't aware that those 'fix issues' checkboxes even existed.

Fourth

Posted Dec 17, 2025 3:20 UTC (Wed) by dvdeug (guest, #10998) [Link]

I'm going to join the people who say they don't see those problems. I run Firefox and I don't recall a site that I've had trouble viewing. There are maybe a few that I blamed on lacking an ad-blocker that might be better under Chrome, but the sites I use, Google Docs, LWN, Google Mail, Wikimedia, iNaturalist, Archive.org, GitHub, Ollama, they all work fine under Firefox.

Fourth

Posted Dec 18, 2025 11:28 UTC (Thu) by nye (guest, #51576) [Link] (1 responses)

> That's because sites only bother testing against Chrome now. And maybe Safari.

Sure, if it's actually broken, that either means it wasn't tested or some higher-up made the decision that it wsn't worth the time investment to support FF. But...

> Not much Firefox can really do about that other than making Firefox just another reskin of Chromium.

This isn't really true, and all the talk about being "bug compatible" with Chrome really misses the mark. The principal thing they could do is *fix the known bugs*. Updates to web specs, especially CSS, commonly add new features that developers have been crying out for for years, and when they finally become available, people have been waiting patiently for years and really want to start using them.

FF is often the last to implement new features, eg :has() which is probably the greatest CSS improvement since its inception, and where FF lagged by years. When it does get new features, they often have trailing bugs that can take a long time to get fixed, eg https://bugzilla.mozilla.org/show_bug.cgi?id=1882408 .

I know I've personally had several occasions where I was about to submit some new frontend updates, did a final round of cross browser testing, and then get hit with the crushing discovery that something is badly broken in Firefox because of a bug in how it implements some feature that's theoretically been usable for a couple of years. It's true that there are *some* times where it comes down to "there is no single unambiguously correct way to implement this, and FF and Chrome made different choices" (eg the phrasing of the file upload button, which is hardcoded by the browser and which leads to bug reports from Dragon Naturally Speaking users which are unfixable except by giving up and using a fully custom upload control), but those are by far the minority IME.

Supporting Firefox in web apps

Posted Dec 19, 2025 11:37 UTC (Fri) by jch (guest, #51929) [Link]

To be fair, it's not completely trivial to make a web app compatible with Firefox and Safari.

I've been maintaining a videoconference system <https://galene.org>, and for ideological reasons I'm trying to keep it compatible with Firefox. I've had to work around a number of issues in Firefox's implementation of WebRTC; off the top of my head:

- Simulcast is broken with VP9 <https://bugzilla.mozilla.org/show_bug.cgi?id=1633876>
- Firefox fails to negotiate a trivial simulcast envelope <https://stackoverflow.com/questions/71550933/limiting-web...>
- the `sendEncodings` field is ignored in `sendTransceiver`, you need to manually hack things with `setParameters`;
- the Picture-in-Picture (PiP) API is not implemented.

Now perhaps that's specific to WebRTC, but some of these issues have been reported years ago, and there's still no fix.

I've also had a number of issues with Chrome (incorrectly implemented data channels, for example) and with Safari (overly restrictive autoplay, no filter property on Context2D), but I've found Firefox to be more painful to support than the other two.