Instead of taking away power from users, Firefox should rather inform users.
Let them modify the browser in any way they want. Let them write plugins that can do anything. But inform users who install a third party plugin what that plugin can do. Like "This plugin wants the right to read all data on all websites you visit. In theory it could capture your bank login and transfer it to some other server".
Another bad decision by Firefox to limit users ability to do what they want is that they don't support the File Access API:
As a general thing the average user does not want freedom; they want safety.
What you are saying might still make sense for firefox because they need to find some sort of niche given their falling market share, and power users is a potential one. However i would disagree that in general its a good strategy for winning the market as a whole.
I'm honestly not really sure that's true. People want safety, but I think when it comes right down to it, they want freedom (aka the thing they want to do to just work) even more
Things "just working" isnt really user freedom either. Generally its hard to make something that just works and is also tinker friendly. The more configurable something is the more the user has to deal with configuration.
E.g. Compare just-workingness between emacs and microsoft word.
yes, I agree, that's why I qualified it, because I don't think what users really want is freedom either, unless the lack of it happens to get in their way
> find some sort of niche given their falling market share, and power users is a potential one. However i would disagree that in general its a good strategy for winning the market as a whole.
That's interesting because their whole initial rise to becoming a dominant browser was based on "power" users back in the aughts.
> the average user does not want freedom; they want safety.
More importantly, I want my family member's browsers to provide them safety. Because I'll be the one running malware removal code on their computer, etc. I'm going to lean towards something that is safe for them.
Why is salary for the CEO rising to 3 mil $ a year for the ceo when the marketshare of the browser has been down by more than 85%. How would paying the CEO more salary and laying of the engineers solve low marketshare is beyond my understanding.[0]
Lots of UI changes like constantly changing menu items and losing control of the browser in general. Leaves a bad taste when useful customization options are removed and instead we get color themes sugarcoated as the coolest new stuff (every few years, sometimes its limited edition sometimes some fancy BS term).
Firefox unique ids:
why they leave me a bad taste ? I have added a previous discussion from HN. [1]
I used to love the firefox from around 2017, but every new release removes some configuration options and we lose control. Maybe my views are full of too much rants, but i install firefox as the first thing on any OS for me and my immediate circle followed by ublock origin. I want firefox to thrive and be handled more like linux.
Even though i don't like the current direction of the management, firefox is my daily driver for development as well as browsing.
(You are not alone in believing that the current Mozilla Corp. leadership are incompetent to uphold the original ethos of Mozilla, and worse, are selfishly prioritising their relationship with BigTech than their users).
It's hard to tell from your comment whether you've met any charity CEOs, or worked with them, or tried to solve the problems they've solved and have learned enough from the space to pass such damning judgement so widely.
There are a lot of charities. Some do good work on a hard (I suspect impossible) problem. Some exist to pay the CEO a lot and do as little good work as they can get by with.
As someone working on a massive migration from MV2 to MV3 (where the benefit to the user will be minimal) it does feel like browser makers are waging war against us extension developers. Mozilla's plan is to release a slightly compatible API, but not at all compatible, and yet highly limit what can be done? Do they really think this will increase the effort invested in their ecosystem?
I think ultimately the biggest question people need to ask is: Is it more important for Mozilla to fall in line behind Chromium and track compatibility with their ecosystem or is it better for them to support their own extension ecosystem and not make busy work for extension developers?
Can they maintain MV2 while supporting MV3 their own way for easier porting of Chromium extensions? Or is MV2 support getting dropped and all MV2 extensions have a EOL coming up next year?
I don't think Mozilla has a choice here. Chrome just has the sheer market power at the moment. If Mozilla went down a different path, extension developers would likely just favor Chrome (and MV3) as that's where unfortunately most users are.
Even if they provided an easier path to MV3, developers would likely have hurried up anyway because Chrome won't do so.
At a more detailed level, well... I'm not that familiar with the area, but it seems like it's a bit of a tangled mess.
The big item of disabling non-declarative content blocking seems safe. Firefox appears to have no reason to turn that part of MV2 off.
The various security tightening measures are another story. My guess is that with Chrome removing higher-risk APIs like dynamic code evaluation, Firefox is using it as an excuse to do the same in order to not have to deal with the issues that arise from them. Which isn't great from the point of view of extension capability. Personally I'd prefer to err on the side of scary permissions notifications for that sort of functionality.
Or at least some of it—I'm not convinced that allowing an extension to execute arbitrary code loaded over an insecure HTTP connection is ever a good idea, except for some temporary development mode thing. (Sure, that's exactly what happens with a random web page containing JS that's served over HTTP. But there's no need to support that case for extensions that are active across all of your browsing activity.)
But it sounds like the removals go well beyond that.
> I don't think Mozilla has a choice here. Chrome just has the sheer market power at the moment. If Mozilla went down a different path, extension developers would likely just favor Chrome (and MV3) as that's where unfortunately most users are.
That's be true for some extensions, but it would be a big, big selling point for Mozilla to be able to say: our ad-blockers are better than yours.
There's also some incorrect binary thinking here. Why must Mozilla support only one extension standard? Mindless application of some software engineering principle? It would make total sense for them to support two: one with the features they want, and another for Chrome compatibility.
> DNR is not yet available in Firefox. Firefox retains WebRequest blocking in MV3, which can be used in place of DNR. When DNR is available, simple request modifications can be moved over to DNR.
I know a lot has been said about DNR for Chrome, and while retaining WebRequest is a good step, I wonder how much Firefox's DNR implementation will differ from that of Chrome. As an extension developer, I really like the idea of write-once extensions that run everywhere, and unfortunately today, a lot of that simply means building for Chrome-* browsers and managing Firefox as a "special case" :/
What? Mozilla is building the best open-source browser on the planet. Sure they sometimes make mistakes (who doesn't? and who can please all users?), but a "war against Firefox users"?!
A "war against Firefox users" is maybe a bit much, though MV3 is definitely a slap in the face, and is a battle in the ongoing war on general-purpose computing[0]. But then, "they sometimes make mistakes" isn't also an accurate description of the mess Mozilla is these days. In the last years, they've been routinely wasting money and alienating the userbase by doing lots of weird side projects nobody asked for, instead of focusing on improving the software people actually use.
Well, as a Firefox user, I would say that Firefox has become way better over the last few years. It's super fast. It actually blocks ads. It syncs beautifully with Firefox on Android.
Some side projects felt useless. Some feel quite relevant to the goal (e.g. Mozilla VPN), some grew to become huge (Rust). Some were maybe an amazing idea at the wrong time (Firefox OS?). I wouldn't want Mozilla to stop doing any side projects.
What's so bad about MV3? Would love to hear your take. Mozilla made it clear it will not adopt changes that prevent extensions like uBlock Origin from working. Many other changes seem to be focused at user security (which is good), and other seem to just follow the "standard", which one can argue is quite important from the perspective of users too, because the fact that extensions can be easily built for Chrome and Firefox together is effectively _the reason_ some of the extensions I use are even available on Firefox.
> What's so bad about MV3? Would love to hear your take.
It's a trust issue. Mozilla burned a lot of it in recent years.
> Mozilla made it clear it will not adopt changes that prevent extensions like uBlock Origin from working.
That's the thing: I don't trust they'll keep good on that promise for more than a year or two. I may be misremembering things, but I'm pretty sure they initially weren't going to adopt MV3 at all, then they weren't going to add DNR, now they're adding DNR but "not removing request blocking, pinky swear". Once DNR lands, I worry they'll start thinking about removing request blocking, under the guise of "security".
I guess you can't argue with someone who ignores everything good Mozilla is doing and everything good they promise to keep doing because they don't "trust" them. Meanwhile I've updated my FF Dev Edition to the new beta version and I'm very happy with the work being done.
It had a reputation for being slow and bloated at the time which prompted the creation of a slimmed down browser, Phoenix which eventually became Firefox
Oh wow, I’m not even sure if I ever used that. I do remember Phoenix, but this barely rings a bell.
And towards your question: Different times. Both hardware and browser capabilities were vastly different back then, so some technological choices made in the early days might have made more sense.
For me personally, I know that back in the day I was a lot more forgiving of delays than I am now.
I'm getting tired of this manifest v3 drama on HN. Why don't people who oppose it band together and fork Chromium and maintain that fork. The software is open source at the end of the day.
Because it's not really about MV3 but about the largest ad and tracking network on the planet changing the most popular web browser to cripple ad blocking under the disguise of security.
What's happening is step 1 - Firefox buying into MV3, even if with its own tweaks, and mentioning that DNR is a thing that will happen eventually. Step 2 - once people get used to it and drama dies down - the implementation will likely become very close to Chrome, under the guise of not fragmenting the dev experience and similar nonsense excuses.
These moves keep happening in tech all the time. Slippery slope fallacy doesn't work here - if you see something that looks like slippery slope in a tech business, it almost definitely is one.
But they never mention removing WebRequest, only having DNR in addition where it makes sense for performance reasons (and for compatibility with chrome), not having it would be stupid. I’m all for slippery slope arguments, if they are well-founded, not if it sounds like wild conspiracy theories.
I'm puzzled why you think supporting DNR is a bad thing. I get the argument that it feels like it moves closer to abandoning blocking web requests, and that's probably more a question of trust than anything.
But there are valid technical reasons to support DNR. It really does make more sense from a performance point of view, if it happens to support what you need. If not, then it's good to have blocking web requests available. And there's good reason to be unhappy if Firefox drops them. I don't see any strong incentives for doing so, though.
> I'm puzzled why you think supporting DNR is a bad thing.
You actually got it in your first paragraph - for me, it's a trust issue. I don't trust Mozilla to retain web request blocking, particularly when DNR becomes established in Firefox and most extensions migrate: because then, the argument for removal web request blocking will be, "look, there's very few legitimate extensions using this, and it's a Huge Security Hole, and everyone standardized on DNR already".
> Last fall [i.e. 2019], we announced our intention to support add-ons in Mozilla’s reinvented Firefox for Android browser (...) In the next few weeks, uBlock Origin will be the first add-on to become available in the new Firefox for Android.
Sounds to me like uBlock is literally top priority for Mozilla, so I don't see why they would do that.
> Is your claim that firefox and chrome are joining forces as part of a conspiracy to prop up advertisers?
No, because there is no conspiracy needed. Chrome is a product of an adtech company, and Firefox is funded by the very same one. Mozilla is at best fighting a losing battle here, forced to make concessions to survive. Or, it could be argued that Firefox exists only to serve as an antitrust shield for Google.
If Mozilla: Do you think that Mozilla would think of their users' privacy if it meant alienating Google and loosing their money? I'm not picking a side here - I have no idea so I'm just asking.
> The software is open source at the end of the day.
Then you can point to the source of widevine? Something that chrome developers called a core component when they where caught side loading it on dedicated open source distros a few years ago and initially refused to supply an option to disable it?
It always said chromium, but I find that people actually mean "chrome but open source" when they say "chromium", when that's not really the case at all.
I’m no expert, but I think the main moat that Chrome has here is its web store where users find extensions and developers submit their extensions. There’s a review process which whilst far from perfect is labor intensive and still helps catch rogue extensions. Hard to open-source this part of Chrome.
Exactly. So get behind the wheel and do something (eg. fork Chromium without MV3 or something like that). But if you're not willing to spend your own time and energy then what's the point of constantly complaining about conspiracies surrounding MV3 on HN. It's tiring to read.
Really? That's pretty sad. Greasemonkey is one of those extensions that opened my eyes to the power of browser extensions. To lose its abilities would be sad.
Let them modify the browser in any way they want. Let them write plugins that can do anything. But inform users who install a third party plugin what that plugin can do. Like "This plugin wants the right to read all data on all websites you visit. In theory it could capture your bank login and transfer it to some other server".
Another bad decision by Firefox to limit users ability to do what they want is that they don't support the File Access API:
https://web.dev/file-system-access/
This gives Chrome users the ability to build simple web pages that can be used like native applications.
An empty html page with a textarea, load and save buttons and you have a text editor!
I hope Firefox will notice that they lose market share because of restricting users freedom. And turn the ship around.