Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Mozilla's open source AI strategy

blog.mozilla.org

Post reply on HN
Print view View original

71–80 of 213 posts

Re: Mozilla's open source AI strategy

#71

Post by LunaSea »

jampekka wrote: post #58
LunaSea wrote: post #48

Earlier quoted context omitted.

> what exactly are the complaints with Firefox? If you are a (the) leading browser like Firefox once was, the "what are the complaints?" is the right question. If you are a minor browser like Firefox currently is (~2.5% market share), the "what is it doing better?" is the correct question.

The Firefox market share was eaten largely by the enormous and legally dubious marketing campaigns by Google and Microsoft. Hard to see how Mozilla could compete with constant forced nags and defaults in the most widely used websites and operating systems.

The Firefox market share was eaten by being worse than Chrome, especially around the developer tools and extensions market places at the time.

Re: Mozilla's open source AI strategy

#72

Post by LunaSea »

duped wrote: post #59
LunaSea wrote: post #48

Earlier quoted context omitted.

> what exactly are the complaints with Firefox? If you are a (the) leading browser like Firefox once was, the "what are the complaints?" is the right question. If you are a minor browser like Firefox currently is (~2.5% market share), the "what is it doing better?" is the correct question.

> what is it doing better adblock is the single most important feature of a web browser to me. Firefox has the best adblock support.

I agree, although Chrome has extensions like uBlock Origin Lite and Privacy Badger which are decent enough for most people and uses.

Re: Mozilla's open source AI strategy

#73

Post by drnick1 »

freedomben wrote: post #44
drnick1 wrote: post #38

The open source community will start taking Firefox seriously again when all the AI shit is removed for good and real improvements to performance and privacy are made. Despite all the posturing about "respecting your privacy and freedom," the stock configuration of Firefox is trivially fingerprintable. At the very least, a privacy-focused browser should adopt the Tor patches and report standardized spoofed values for…

I'd definitely love to see them take some steps, but at the points where a minor increase in privacy leads to a much worse UX for average people (aka, "why doesn't this site work in FF, FF sucks!" because they don't know they have to enable something). If Firefox becomes a browser that is harder to use then it will only ever be used by the extremely small niche of people that care about that. That will only further l…

> I'd definitely love to see them take some steps, but at the points where a minor increase in privacy leads to a much worse UX for average people

Disabling Javascript or even just third party scripts does lead to major breakage, but reporting spoofed values for identifiers like Tor does not. The Arkenfox user.js does all of this and more, but these options are not enabled by default. This shows that Firefox does not care much about privacy in practice.

The only "breakage" that I have encountered with such a hardened configuration is related to the spoofing of the time zone. But the fundamental issue is that Javascript/browsers should have not been designed to allow websites to extract this kind of personal information in the first place. But even that is not enough and users are still fingerprintable. In an ideal world, the only thing a website should see is the originating IP and nothing else.

If anything, Brave has done more to harden Chromium than Mozilla has with Firefox, even though Brave comes with its own set of problems (scammy crypto integrations, AI, VPN and other stuff).

Re: Mozilla's open source AI strategy

#74

Post by EvanAnderson »

Larrikin wrote: post #24
nusl wrote: post #12

I guess replies on this thread are evident that Mozilla has lost much of the trust and goodwill it once enjoyed. Admittedly I am also very skeptical that Mozilla has the ability or genuine interest to make this work.

There is always a pile on on Firebox for not being perfect. Sometimes with valid complaints. But if you dig deeper nearly always the commenter is using a version of Chrome and justifies it over Firefox for a very shallow or outdated reason. Firefox would do well to listen to some of the criticism about the browser and ignore the noise about anything else There's also the cohort of bad web developers that only test on…

> There is always a pile on on Firebox for not being perfect.

I don't have a problem w/ Firefox not being perfect. I have a problem with the Mozilla Foundation spending money on seemingly random other stuff and not on Firefox.

Re: Mozilla's open source AI strategy

#75

Post by devsda »

dralley wrote: post #69
devsda wrote: post #62

Earlier quoted context omitted.

It is more easier to secure revenue/funding from Google once they retain existing market share and gain more. They need to improve the product for that to happen. With all the distractions they are abandoning their primary product and they are bleeding whatever miniscule market share they have. This means Google has more leverage over them and can eventually stop the funding once their market share drops beyond a thr…

If we're being completely honest, improving the quality of the product would not meaningfully improve their market share. That worked in the early 2000s when the competition (internet explorer) was utterly stagnant and the internet-using population was composed predominantly of techies willing to try new things. Browsers are commodities now, and most people aren't going to try a new browser when they're already using…

End users are easily influenced but they could have targeted developers.

I think they should have pushed for a gecko based electron alternative. End user dont really care if their favourite markdown editor or notes software is based on electron or gecko but it would have made sure that developers do not target, develop and test for only chromium based browsers.

Re: Mozilla's open source AI strategy

#76

Post by striking »

bondarchuk wrote: post #17
striking wrote: post #11

Earlier quoted context omitted.

Maybe, but I would argue that some of these features are genuinely useful and important. Take translation, for example. It's not great to have to send off a page that potentially contains identifying content to Google, but it is the easiest way to handle the matter. Firefox uses local AI to perform a decent translation relatively quickly, and I'd like them to work on improving that capability.

Many things that are not browsers are genuinely useful and important, this alone doesn't mean Mozilla should be doing them.

Translation is a necessary part of the web browsing experience for many people.

Re: Mozilla's open source AI strategy

#77

Post by LamaOfRuin »

diffeomorphism wrote: post #37
philipallstar wrote: post #14

> Mozilla was born to change this, and Firefox succeeded beyond what most people thought possible — dropping Internet Explorer’s market share to 55% in just a few years and ushering in the Web 2.0 era. Is this true? I can see from here[0] that its peak was 32%, as IE was really on the back burner but before Chrome had fully risen to dominance, but I wouldn't claim that it was responsible for IE's market share drop. […

When do you think the "web 2.0 era" was? Web 2.0 is around 2003 or so and chrome would not even exist for another few years. Giving Firefox/phoenix/Netscape the majority credit for the first fall of IE seems accurate. The rise of chrome happened afterwards and by then IE also fell much deeper than 55%.

Opera was also essential at this point, not in terms of market share, but of innovation in the browser space with features that would eventually spread to everything else.

Re: Mozilla's open source AI strategy

#78

Post by dralley »

devsda wrote: post #75
dralley wrote: post #69

Earlier quoted context omitted.

If we're being completely honest, improving the quality of the product would not meaningfully improve their market share. That worked in the early 2000s when the competition (internet explorer) was utterly stagnant and the internet-using population was composed predominantly of techies willing to try new things. Browsers are commodities now, and most people aren't going to try a new browser when they're already using…

End users are easily influenced but they could have targeted developers. I think they should have pushed for a gecko based electron alternative. End user dont really care if their favourite markdown editor or notes software is based on electron or gecko but it would have made sure that developers do not target, develop and test for only chromium based browsers.

Developers are no longer a significant fraction of the pie, and a significant fraction of those are web developers or do web development, and those users will in all likelihood primarily use what their users are using, which isn't Firefox.

Re: Mozilla's open source AI strategy

#80

Post by morcus »

LunaSea wrote: post #72
duped wrote: post #59

Earlier quoted context omitted.

> what is it doing better adblock is the single most important feature of a web browser to me. Firefox has the best adblock support.

I agree, although Chrome has extensions like uBlock Origin Lite and Privacy Badger which are decent enough for most people and uses.

> decent enough for most people and uses.

Except for the very big use case of mobile browsing, where only Firefox allows extensions.

Post reply on HN