Jan 19, 2026, 7:58 PMRE: https://social.vmbrasseur.com/@vmbrasseur/115922995190919975
Mozilla wants your input. Here's mine:
Mozilla should be doing two things and two things only:
1: Building THE reference implementation web browser, and
2: Being a jugular-snapping attack dog on standards committees.
3: There is no 3.
Mozilla should have NOTHING to do with AI. Nobody wants it. Stop forcing AI into every corner of every project because your VC-brained management have completely lost the plot.
@jwz My reply
@jwz Related: if a browser truly wanted to re-establish trust with end users, and make it clear that they are in fact a "user agent" rather than a tool of the adtech industry, you know what the single biggest way to speak with their actions would be?
Ship an adblocker out of the box, by default. Everyone using the browser would have it on, with a reasonable default set of rules, unless they explicitly disable it or added custom rule sets.
An open declaration of war on the adtech industry.
@azonenberg indeed. we built adblock directly into konqueror.
Jan 23, 2026, 3:20 PM@azonenberg funnily enough Mozilla has one for Android, and it's called Firefox Focus.
Granted, the adblocking is simplistic compared to the dedicated browser extensions, but it suffices for my day to day browsing.
It's the default browser on my phone, not just because of this, but also because it clears all cookies after each session ends, which is ideal for opening random links. Of course it also means that those annoying cookie policies keep popping up, which it sadly doesn't block.
@jwz
@FiXato @azonenberg @jwz Love Firefox Focus. It's just fast and hardly any webpage breaks. FF is set as my default browser. Thus, anything first opens in FF. Any webpage I trust enough to have a login for, I will open in Firefox Full-Fat. (FF has a handy "Open in..." menu item that reloads the site in the browser of your choice.
Jan 23, 2026, 4:27 PM@jwz @azonenberg I can take or leave ad blockers. If I (miraculously) had the budget to coordinate the development of a browser ex nihilo, the design principle throughout would be making it monetization-hostile. This would be a higher priority than standards compliance, as large portions of the standards have been drafted by the IP or adware industries to serve their own interests.
Particularly, the Javascript runtime would be highly user customizable to make broadly defined categories of events loggable, bypassable, etc.
I would include opt-in telemetry, but specifically to enable users of the browser to "compare notes" (in an automated way) on the multitude of domains in the asset pipelines of the websites they visit.
Feb 1, 2026, 9:05 PM@azonenberg @jwz @onepict I get the impression that Firefox does SOMETHING toward this general idea, since I'm continually shamed by sites for having an ad-blocker, despite never having knowingly installed an ad-blocker in Firefox. 🤔
@jwz I told them to be LibreWolf. Get rid of the scammy money-making "responsible ads". Ublock origin, no telemetry, no AI. I donate money because I want there to exist a libre open standards browser. Not because I want a corporate brand
@jwz how mysterious that they neglected to include an "other" option only in the AI question... fucking weasels!
Jan 19, 2026, 8:17 PM@jon_ellis I used the other Others for additional Anti-AI.
@jon_ellis @jwz Exactly... where's the "burn it all, burn it all to the ground" option... or even just "other" so I can type that in.
FFS.
@ivor @jon_ellis @jwz Does the Butlerian Jihad count as "government regulations for AI"?
@ivor @jon_ellis @jwz I interpreted "Holding corporations accountable for their AI products" as including Mozilla.
@ivor @jon_ellis @jwz you can check nothing on this field
@jwz Told them something along those lines too. Maybe they'll get the message if there's enough answers in that vein.
I don't have high hopes, though.
Jan 20, 2026, 7:17 AM
@jwz I would be happy if they just fixed memory leak issues and be faster than chrome
@impactology @jwz Why are you running the nightly version if you don't want any bugs in your software?
Jan 20, 2026, 7:18 AM@impactology
The fast loading was obviously a bug.
@kneoghau @jwz
@jwz Urgh this survey is a trainwreck:
Some real "begging the question" with the second option (if I understand the phrase "begging the question correctly.) (And a whole bunch of questions on AI later that do not include "Protecting the open web from AI"
[X] Done
I really love firefox and it makes me cry to see, what is happening with it.
@jwz are we sure this is from Mozilla? (I’m just not familiar with tfaforms.net.)
mine were like yours
except I also added
3. moving their activity away from the USA to the rest of the world (not just Europe)
@jwz Did my part. Also told them to not screw up Thunderbird since the email client ecosystem is just as broken.
@jwz I would happily pay the Foundation IFF they do those things you list. I want my browser to be my tool, not someone else's. I never mind paying for things of value. But I *won't* pay for things which victimize me to enrich someone else.
@jwz My response in the third block:
---
There is no #3. GET THE FSCK OUT OF THE LLM CESSPOOL!
"But everyone else is adopting AI..."
Oh, so I suppose if everyone else jumped off the Empire State Building, *you* would have to jump off the Empire State Building.
LLMs are the 3D movies of tech. A novelty. It *seems* cool, but in truth it's a *dead end.* Save your reputation. Save your *MONEY!* Get off the bandwagon before it sails off the cliff.
Jan 19, 2026, 9:56 PM
I forgot to add: get a decent spell-check. 5/6 suggestions are just letter-salads, not words.
@jwz The only thing I would add to that list is to provide industry leading developer tools built in to the browser that allow developers to build the most robust, performant, beautiful and standards compliant web apps.
@jwz it's already done and detached if you look closely. speaking as someone that had their name under yours on that monstrous Makefile https://servo.org/
+1
@jwz @gdinwiddie NoBoDy WaNtS iT 🤬
@jwz Firefox market share is <6%. Zen is trying to port the much-loved features from the now stagnant Arc, but the result so far is clunky and I've returned to Arc because it's more perfect right now for how I work and play.
Mozilla needs a mission bigger than just a standards-based browser. It should be experimenting with new UX. Not Liquid Glass UI/UX, but Arc-like UX.
Perhaps they should be experimenting with AI, with the goal of making AI invisible to the user without boiling the oceans.
Jan 19, 2026, 11:41 PM@jwz @n3bulous I can live without a new UX. I can even live with incomplete implementation of the standards.
astoundingteam.com/2020/04/21/…
What I really miss is the independence. When Firefox was first airdropped into the public domain, legions of indy websites advertised it, for free. You literally can't buy that kind of enthusiasm and loyalty, but Mozilla sure knows how to piss it away.
Jan 19, 2026, 11:43 PM@n3bulous NO
@jwz What is accomplished with an irrelevant Mozilla? What are FF and Mozilla in 10 years being only a standards based browser?
My original post agreed with points 1 and 2, but I trimmed them due to length. I believe those are noble goals, but you are effectively killing FF. Maybe you want EFF but for the Web.
That said, I didn't really agree with any of the question choices so I posted only a comment.
Jan 20, 2026, 12:03 AM@n3bulous How you judge the success of a commercial product is by counting users and counting ad revenue.
How you judge the success of a reference implementation is by standards compliance and code reusability.
I guess you think Mozilla should be a company making a product rather than a nonprofit whose mission is the open web, with a reference implementation existing only in service of that goal.
@jwz I see where you are coming from. I was operating from where they are now, and you are saying they shouldn't be doing most of it? That would be fair.
At one point, as you know, people were concerned with being compatible with Firefox/Navigator. My perspective is that was because a significant marketshare used it. Right now, it seems most companies target only Chrome and hope for the best with others (while forcing excessive amounts of React down our tubes).
@n3bulous @jwz I think you have a point about user base being the basis for influence on standards but the wrong proposal. Web browsers as a product are mature, and adoption is mostly driven by platform control. UX changes and random bolt on features won't make much difference to adoption because most people don't think about browsers, they just assume the internet button will be available and work. All is not lost. Embedding in apps, app shells like Electron, PWAs, still ways to grow.
@n3bulous @jwz Expanding on that slightly, those are all areas where tool choices are made on a partially technical basis and a a better answer than Chromium will be welcome. Users may not know why their computer is slow but at least some product managers care if its their app causing the problem. Robust, predictable behavior is a big plus when it's code becoming a part of an app you're responsible for.
Jan 20, 2026, 12:17 AM@jwz @n3bulous EFF but for the Web is basically IceCat. I like IceCat, even though it's basically unusable. IceCat is unusable because of how websites are implemented, not because of how IceCat is implemented. I like to fire it up every now and then just to update my mental friend/foe list. Privacy Badger (also supplied by EFF) is also very educational in that way. The Tor Browser has a similar but different usability/unusability pattern. Retail websites block Tor exit nodes about as aggressively as news media websites accuse tracker blockers (such as Privacy Badger) of being ad blockers (which it absolutely is not).
People speak of tools like Privacy Badger "breaking" certain websites. The message discipline I use reverses this. I describe tracker-dependent sites as "breaking" Privacy Badger. With such tools amplifying the contrast between "figure" and "ground" (breakage and non-breakage), you can practically see the outline of various business models. It's as entertaining as it is educational.
What I really want is the pre-Web Internet. Do the young people even know that the Internet is older than the Web?
Jan 19, 2026, 11:33 PM@jwz dear heavens that survey is as biased as anything Sir Humphrey could've dreamed up 😬
Correct on all counts.
Jan 20, 2026, 1:24 AM
Well said. Can't think of anything I would add or change.
But yeah, inherent AI functionality is the LAST thing I need or want. Keep it out of the core build. Make AI plug-ins for those that desire it maybe..
Cheers.
@jwz But look you can run it locally on your machine, and it helps to code better and faster, and it really doesn't consume as many resources as you think, I have the maths on it that I copied off a neutral website, and why would you not use it come on, do it, do it ?
@jwz submitted my 2c. Just focus on the core browser.
Mozilla wants your input.
https://mozillafoundation.tfaforms.net/201
Here's mine:
@jwz Make everything AI free Mozilla. We already have human intelligence. AI is simply regurgitating stolen human intelligence
Jan 20, 2026, 5:24 AM@jwz
I might sound a bit underducated on this, but I also wish they could make the Gecko rendering engine more independently usable. It pisses me off that every new browser is based on Chromium. imo if we make the rendering engine more popular, it will get way more attention and people jumping in to fix issues lying around for years because of lack of staff.
@jwz I don't disagree really, but it's not necessarily clear cut. Where's the line for what just a browser is? Does that include password management? Extensions? Themes? Tabs? Translation? Developer tools? Bookmarks?
Jan 20, 2026, 6:05 AM@SonnyBonds Side quests worth considering after declaring the main mission well under control.
@jwz I'm saying I think what the "main mission" of a browser is can be debatable. Having a single window that just takes a single URL input and displays what's on that page is the core browser experience I guess. At what point is that 100% complete enough to add tabs, etc etc.
I'm just saying I don't think there's a clear cut line and different things could be argued to be fundamental for a reference implementation of a browser.
