It's also good marketing for regular users. When Chrome introduced this versioning style every other software was stuck in minor versions for years. Regular users started assuming that Chrome must be better since it gets major versions constantly. I don't mind either option but I would love to see some sort of date based version in there, like Firefox 72 (20.01).
After the recent missteps by Chrome (mangling the display of URLs), I decided to switch to Firefox and DuckDuckGo a few months ago and I was pleasantly surprised at how fast Firefox is these days than what I found it to be 10 years ago.
While I still consider DuckDuckGo to be in the "not bad" category, Firefox is in the "seriously awesome" category now.
If you have been away from Firefox for a while like me, give it another shot. It won't disappoint you for sure.
Are you a devtools user? How do the Firefox ones compare? I speak as someone who originally switched to Chrome (having been a Firefox user since the Phoenix days!) specifically for the devtools — it's pretty much the biggest factor in my browser choice.
Sorry, I'm not that useful for your question. My experience is from the reverse switch. I'm used to Firefox (it's my daily driver) and had to get acquainted to Chrome.
I had to get used to Chrome devtools because javascript debugging is a lot better there. For completeness sake, CSS debugging is much better in Firefox. In time, I think both will converge.
I’m a Firefox ddg user, and I would recommend Chrome for devtools still. As of last month, Firefox devtools didnt even recognize meta viewport tags which caused UI discrepancies between testing and production. Chrome devtools does have it.
That being said, I still switched back to Firefox and ddg after I close devtools.
Interesting then that this release enables meta viewport support in responsive design mode - from the link: "Firefox now supports simulation of meta viewport in Responsive Design Mode."
there's some cool stuff in the works but it doesnt sound like there's a plan to internalize the work into devtools but to keep it as a web app and rely on a serviceworker for "localness" - not great, imo.
Sadly noone cares what the performance inspector says. Newspapers just get slower and slower. Our local newspaper takes 15 seconds to first visible paint and another 5-10 seconds to usefulness on my phone...
My impression as a web user is that if any devs actually use the performance inspector it's just to learn their site is slow and then not do anything about it.
Your devs might develop perfectly well behaving stuff just to see it get totally obliterated by loading 3 external tracking scripts through Tag Manager, three different pixels and 7 external contact form/chat bots/newsletter popover scripts from their CRM/ad platform/whatever crap marketing wants.
You get used to it - UI is slightly different in some areas, but it lacks for no major features and performance is good, like firefox in general.
I've found it's extremely valuable to be developing sites in Firefox using the dev tools simply to help catch issues that the rest of the team ignores as they're all stuck on Chrome. Fights monocultures by default!
I remember Firefox being seriously lacking in some WebSocket dev tools that I needed. But I checked again and they're there. So it seems like they are actively improving the dev tools.
This has been a huge Pita, for my team. Good news is it has gotten some of them to switch to Firefox for debugging. It really is awful that bug slipped through.
I am not OP but I'm a FF user. I haven't got around to learning FF's dev tools and still need Chrome(ium) to test that platform for compat anyway. So I just still use Chrome dev tools (which also work for NodeJS instances) but use FF for actual browsing.
I use Firefox as my default browser. But the dev tools is too slow, apparently someone thought it was a good idea to rewrite it in React, causing it to lag even on a few thousand console log messages on a fairly powerful workstation. And there is no search functionality, only filter. So I've started to prefix all my debug messages with whatever I'm working on just so that the logs can be found in the Firefox dev tools console. One good feature though is that you can filter on type of messages, like only show error messages, or warnings. Which is very annoying in Chrome dev tools where it will tell you there is an error, but you can't click on it, so there is no way to see the actual error.
DevTools member here. We have been working on performance over the past release – is there anything specific that you can point out as being slower for us to focus on?
I have not yet tested v72, I'm on v71. The Chrome dev tools Console feels (not measured) about 100x faster then the Console in Firefox. Even Chromium Console can be limiting though. Yeh, I probably should not use console.logs for debugging, but it's much faster forkflow then firing up the debugger. Even with the lag. With the Chromium Console the lag is not that bad, and it has a search function. Both console's have a bunch of cool features, like being able to inspect DOM nodes, explore objects, etc. Witch I think might be a reason for the lag. Most of the time I just want to search the logs to find out what happened. An idea is if you can just get the plain text from the console.logs somehow.
In chromium all console.logs get sent to stdout, so I can pipe the logs to a file, then use any editor to search it. But I prefer using the dev tools.
Maybe make a perf switch in the console, which dumps the text without formatting or other features. Something that can handle one million lines of logs and search in it. Being able to search the logs is critical. Filtering doesn't work as I need to see the surrounding logs, what happened after. I don't need to see all one million lines at the same time though, it can be buffered to only show 100 lines above it and 100 lines below it. Chromium has such an optimization. But search is a bit slow.
How much better is it than Safari? I find the differences in browser performance so subtle nowadays that I lean towards whichever gives me the most seamless experience—-which, as a Mac/iPhone user, is safari.
For me it's mainly about synced history, open tabs on various devices (phone, tablet, computer), shared preferences, etc. The rendering engine is secondary to that in my list of priorities.
I also gave Firefox another shot a couple months ago and I'm surprised at how much better it is nowadays. It's been such a good experience I'm also giving a go to the Firefox Preview[0] on my Android - even though it's still under heavy development, I like some of the changes to the mobile browser it makes and it'll make in a near future.
The only thing I missed from the ordinary Firefox on Android was that Firefox used to have a brilliant hack with the sharing option where they would show the two most frequently used options (+ a sharing button to access the rest) instead of just one or just the sharing button.
This saved me a few seconds again and again and again.
Yeah, the sharing button thing is something I miss from the ordinary Firefox too. I usually share links with my friends through Whatsapp and its icon is hidden far to the right.
Firefox is definitely an amazing experience, especially if you use Firefox Containers[1]. My only gripe is that the UX of containers isn't quite ready for prime time yet (e.g. your container setup doesn't sync by default).
And there really needs to be an easy way to switch profiles. Firefox has had profiles for basically forever, but it's much easier to switch profiles in Chrome.
I currently go to about:profiles and click "open this profile in a new window", but this is very klutzy compared to Chrome's profile switcher next to the URL bar.
Of course you can also create launchers/shortcuts for Firefox with "firefox -P profilename" but again that's work you have to do on each machine you use.
A classic solution is just one shortcut of `firefox -no-remote -ProfileManager` which semi-replicates one of the old Netscape-feeling behaviors in that such a shortcut will always create a new instance of Firefox and first launch the select profile window.
>While I still consider DuckDuckGo to be in the "not bad" category...
If you're into privacy-conscious search engines and want one that's not susceptible to blanket FISA warrants[0], check out either Qwant[1] or SwissCows[2].
[0] - DuckDuckGo's address is in Pennsylvania - 20 Paoli Pike, Paoli, PA 19301.
wow, those seem quite fantastic at first glance. If I search for my real name I get every single profile I've ever created since the beginning of time - as opposed to google listing a bunch of random garbage. Names there are like John Doe returns famous people named John OR Doe above the person(s) with the actual name.
EDIT: I right click the search box and make qw and sw keywords for them.
EDIT2: I'm just looking for a functional google alternative.
For some reason, your Firefox profile doesn't have a preferred language, but Qwant expects one. If you add one in the Preferences (option "Choose your preferred language for displaying pages"), it should work.
France has a more permissive surveillance regime than the US[0], and as a foreign (US) user I have no recourse if they spy on me.
Furthermore, here are some quotes from qwant:
>We do not use any cookies or other advertising tracers to create your profile for commercial purposes. This means that you can use Qwant with confidence, we will never try to establish your psychological or commercial profile to sell it to third parties, here or elsewhere. [1]
Note the "for commercial purposes" in the first sentence, and the "to sell it to third parties" in the second. These are weasel clauses.
> France has a more permissive surveillance regime than the US[0], and as a foreign (US) user I have no recourse if they spy on me.
Hahahahaha, enough with the jokes. Comparing France to the US... Thinking you actually had some kind or chance of recourse against any of the three letter agencies. You are funny!
> Note the "for commercial purposes" in the first sentence, and the "to sell it to third parties" in the second. These are weasel clauses.
Could be, could not be. With Google you can be sure that they sell your data to _anybody_ that is willing to pay something for that.
Okay, fair point. Googles' business is literally selling access to all the data they have collected from you. So they are not selling your data, only the access to that data, that's right.
Downvoters care to explain? I was born and lived in France for 20+ years, I think I've a pretty good idea about the government, freedom and privacy in that country. I've seen Qwant being shilled countless times, starting from the french government pushing a "privacy focused search engine" a lot of school use this search engine by default. Despite all of that I have yet to see code or any actual proof they respect user privacy.
I tried Swiss Cows just now, it appears effective at first glance but a little longer and realized I’ve never seen a more sanitized and curated search.
I don’t necessarily mean that good or bad. Just interesting how... different it is. Seems good if you know your thug exists already and someone approved it for their results.
> After the recent missteps by Chrome (mangling the display of URLs)
Me too! I'm not as bothered by software-politics as many of the commentators on here (I don't hate Facebook, and worked there for four years; I don't hate ads; I don't worry too much about trafficking and privacy...) but Chrome fucking with the URL was a bridge too far. I switched to Firefox and when Google walked back their hiding-the-URL business, I never bothered switching back, because I didn't trust them anymore.
Companies take note: users' trust that you won't arbitrarily change their workflow in annoying ways is hard to win back.
The problem is it's impossible to separate "this changed it sucks" from "this is actually bad" since the messages are always the same. E.g. in this case you didn't even bother explaining why the new way works worse for you you just said "it changed, I left! Make note!" and I mean that's fine if that's what you actually wanted them to make note of but no company with competition is just going to keep every workflow the same for over a decade and wait for someone else to come out with a better product to replace them.
- Do they really need to see* http:// or https:// if immediately to the left is the icon which indicates secure/insecure (which is more than just "protocol is https or not")?
- Do they really care to see* if they are at www.site.com or m.site.com or site.com or do they just want to know they are at site.com which they expect to be as they make their purchase or log in?
* noting that bookmarking, copying, or editing the URL operate on the full URL and this is just a visual change otherwise
And voila now the leftmost (first) text they read is not a bunch of jargon it's the thing people look at the URL for the most.
That you couldn't think of any reason tells me you didn't try not that you disagreed with the reasons and had an actual lesson to share about why. Again: "it changed, change bad" people impossible to distinguish from "it changed, actually bad".
> Do they really care to see* if they are at www.site.com or m.site.com or site.com or do they just want to know they are at site.com
www.example.com == example.com is an assumption that has never been true and I have seen it screw over regular users countless times. Quietly rewriting the URL and hiding that fact will just make that worse.
As for m.example.com, that is even further from standardized. Me and and at least one person I know have personal websites with short aliases like mike.lastname.tld == m.lastname.tld, not to mention the plethora of sites that don't have matching paths on mobile vs desktop. Even if copying the URL quietly switches it out for the real one (which is bad on its own), that break the surprisingly common workflow of writing down the URL you see on a piece of paper.
Thank you for talking about a real use case :). The m. causing a new workflow issue to be introduced is exactly why they decided "m." should not be treated as a trivial subdomain when they rolled it out, it was only that way during testing.
As you said www.example.com being functionally different from example.com was already a broken workflow, users weren't differentiating it. Continuing to display something users haven't understood for 20 years was not considered a strong enough reason to simplify the URL display now in a way that causes the same errors users were commonly making anyways.
While somewhat a bad form for general web browsing, its definitely possible for www.site.com and site.com to be entirely different hosts. In newer versions of Chrome, you would never know which one you were actually on unless you select in the address bar.
Maybe dropping the protocol definition in the URL is debatable, as probably 99.99% of the time in Chrome a user is going to be either http or https, but dropping a part of the hostname is unacceptable in my book.
Would dropping .com be acceptable? All a user wants to see is that they went to Google, to an average user the .com could be seen as redundant clutter as well. Might as well have the address bar show "google" or "cnn" or "facebook" in it, as clearly that's what a user cares about.
Possible and reasonable are different things. If www.google.com gave you gmail and google.com gave you search users were confused before www was hidden post load.
You can't really get rid of the TLD without breaking the security model of the web (e.g. a lock icon and google.com is different than a lock icon and google.gtld). Would it be nice to refactor that? Probably. Is it reasonably possible at this point? No.
You mean the strawman repeatedly trotted out by people defending shitty changes as justifications for why their shitty change is necessary? Ever wonder why so many things suck nowadays?
Nope, just mean "not the guy that writes the webapp who needs to see if he is currently connected via HTTP or HTTPS".
Can "the larger group" be used as an excuse for a bad change? Sure! But so far the only reason mentioned as to WHY this change is bad is "www.site.com and site.com could technically be different sites" and I'm not sure how that is supposed to be clear regardless if www shows post load or not or how that's supposed to outweigh the advantages.
> Nope, just mean "not the guy that writes the webapp who needs to see if he is currently connected via HTTP or HTTPS".
HTTP currently shows "Not secure" to the left of the URL. It's still easy to differentiate.
(I do think hiding www is stupid, but hiding http vs https is reasonable so long as they're still marked differently. Ideally http should be a more noticeable error, like it is if there's any user input box, or better yet a massive error like it is for self-signed certs and HTTPS.)
In general I prefer to use tools that show me real data rather than transforming it in “smart” ways that are poorly documented and not under my control.
If other people have different preferences, that’s fine! They can keep using Chrome.
Also, even if that weren’t true, change is a cost. It’s perfectly reasonable to complain about change as being intrinsically bad without giving any other reason.
Any tips for transitioning? I've seen people recommend various stuff like containers and other stuff but I never saved their recommendation as I was happy with chrome at the time.
I stopped using Chrome about 2-months ago. It was difficult at first, simply because I was so used to it. But you are 100% right; I couldn't imagine going back to Chrome now that I've made the switch.
Agreed. I use it all the time. There is an issue where the video in the pop-up can be slightly "laggy" for some reason, not sure why that is. Otherwise great feature.
Open in new window. Resize window. Right-click title bar, select "Always on Top".
App developers spend a lot of their time making up for the poor decisions of OS developers. A good window manager would give you tabs for free, PiP (aka an always-on-top window) for free, dark mode for free. Instead OS devs go overboard in simplifying things to the point that app devs have to pay the cost of delivering what users want.
It can create, for example, a pop-out video overlay, which can be moved around and resized. It then sits over top of the other content on the page (and remains over other tabs) and will remain fixed as you scroll down the page and similar. On a site like YouTube, you could pop the video out and proceed to scroll down and interact with comments while the video plays to the side or above.
Yep. On gnome you can set it. That doesn't appear to be the case on osx. And it seems that for PIP, being sticky should be the default. I'm currently keeping safari open just to have a pip sticky window of youtube.
Beware though - NextDNS is good but their servers being hosted in France the geolocation gets you IPs in france (when accessing Google properties for example) which slows things down a lot - 30-50ms vs 130+ in my case.
They should have thought about this - it is possible to force resolve the individual domains in the NextDNS UI but that gets problematic fast.
Did you create your personal configuration? When you do, their server should be detecting the geolocation of your IP and assign the nearest ones (what you see at "DNS Servers") to that configuration like how CDN works. You can also check the assigned region at https://ipv4.dns.nextdns.io/ping.
If you want it to just always use what you've configured and not have to follow the changes they make you can set network.trr.mode to 5 in about:config. 5 signifies "off by choice"
If you just want to know how to toggle it normally in case something goes weird with one of the changes it's in the network settings part of the settings page.
My favorite since the 2000's (idk when I started using it, mom didnt let me install it on the home computer for ages) I tried Chrome earlier on but (and I've said this on HN before) it didnt have adblock, then when it did, it was crippled. I never went back.
I used it since the betas. I'm early Mozilla user too. But I quit using it when Google Chrome shipped its first beta. Then, it was matter of time I started to use it. Remember FF was so slow between 2008-2015.
I really wish they'd add HDR support so my monitor would stop flickering but other than that I don't really miss much else from Chrome. Maybe the audio being better on sped up videos but that's about it.
You'll be happy to hear that we've recently dramatically improved the audio quality of sped up (and slowed down) videos, here is the ticket for the details: https://bugzilla.mozilla.org/show_bug.cgi?id=1427267. This will be release in Firefox 73, in about 4 weeks (give or take a few days maybe).
HDR is very much in the list of things we want to do (especially for videos), but as you can imagine, it's not particularly trivial, and will require platform specific work, and deep changes into Firefox. On the other hand, hardware is shipping (we got some devices so that developers can start assessing the situation), and HDR content is becoming increasingly available, so we're trying to figure out what it will take to do it properly.
I have been using it since 2005 (since 1.0.X), and while Firefox at times were slow during those years I always preferred its UI and its extensibility.
Can't wait to find out if this release fixes a bug (https://bugzilla.mozilla.org/show_bug.cgi?id=1557160) that was preventing correct rendering of extension popup windows/context menus when the APZ zooming was enabled (which allows pinch-to-zoom gestures, e.g. on the MS Surface).
Update: Just got to try it out, seems like it indeed works now. Yay Firefox 72!
> Firefox replaces annoying notification request pop-ups with a more delightful experience, by default for all users. The pop-ups no longer interrupt your browsing, in its place, a speech bubble will appear in the address bar when you interact with the site.
This is fantastic! I had finally figured out I could turn this off in settings a while ago, glad it's now a default. I get so annoyed by this, annoying indeed!
In a webapp where you present user a button to activate notifiactions, when the user clicks the button seemingly nothing happens in FF72 (user is focussed on a big enable notifiactions button in the web app and may not notice that some tiny gray icon wiggled a little in the address bar).
On a big screen a button in the middle of the screen is so far away from the address bar, that you don't see any change in the address bar at all in the peripheral vision.
So yeah, web apps that don't try to force the user to enable notifications are now punished for good behavior again.
How about web apps too not show annoying pop-ups which are pretty useless in most cases? Websites which show two pop-ups (custom pop-up followed by browser pop-up) are worse.
My website doesn't show any popups. I have a settings dialog that user can open and setup email and other notifications.
This will just not work anymore. I'll have to add some long winded FF specific explanation on what to do after clicking the button.
Anyway the FF UI is broken if user physically can't see any reaction from the browser after he requests notifications by clicking anywhere on the page.
If your users are clicking a button then it should actually show the permission prompt, unless you're losing the user interaction somewhere in the callback (by doing something async first)
I guess this is then happening because I'm first checking if there's a subscription via `pushManager.getSubscription` before creating a new one in the event handler.
Yep, dropping getSubscription makes the popup appear again.
I receive these popups multiple times per day. It is rare that I actually want notifications from your web site (or mobile app). Messaging is one of the few use cases that comes to mind but even then I don't always want to be notified.
Do you mind me asking what your notifications are for?
What do people use for “Profiles”. I need ability to switch easily between profiles (different identities) without going to command line. I also don’t want to install a third party extension for that if possible
Firefox has a Container Tabs extension which does just this - the defaults are profiles for Work, Personal, Banking, and Social, and you can add more. You do have to install it but I think the extension is from Mozilla itself, so it's not exactly third-party. It's just features that are already in the nightly builds as I understand it [0]. Personally I've really enjoyed it so far!
The extension also allows you to set that a particular site be assigned to, and automatically open in, a given container. I don't believe this particular feature is built into Firefox.
Indeed, it's just the containerisation functionality and assigning a tab to open in a Container that's built in; actually initiating the opening is done by the extension, e.g. when you navigate to a specific origin.
On macOS, instead of running Firefox via the app icon, I run a an AppleScript[0] that forces Firefox to launch with the Profile Manager window. I have it bundled as a one-step Automator application, which I run from Spotlight but which can easily be added to the Dock or whatever.
I used to have this script built as an app bundle via Platypus, but that stopped working somewhere between Yosemite and Mojave (likely due to increasingly strict OS file permissions or something). The Automator app doesn't seem to suffer from this issue...I think I just had to grant it a couple permissions the first time I ran it.
I wasn't able to figure out how to set a custom icon from within Automator, but it's not too hard to do if you're comfortable mucking around inside app bundles. The icon file is stored under <app-bundle-root>/Contents/Resources/, and its name (sans extension) is referenced in <app-bundle-root>/Info.plist.
"Firefox replaces annoying notification request pop-ups with a more delightful experience, by default for all users. The pop-ups no longer interrupt your browsing, in its place, a speech bubble will appear in the address bar when you interact with the site."
That should pretty much kill off a lot of the notification request crap, especially if Chrome follows suit. I can envision the conversion rate massively falling off when it's no longer something right in your face.
Slack has started disabling unrelated features if you disallow notifications. For instance, you can’t add channels to the “ignore @here” list if notifications are off. I’m sure this’ll get worse over time.
FF needs a “disable notifications, but lie to the website and say they’re enabled” button.
I feel like a good extension to make would be a lying extension. Make it able to lie to a site about any API and have set configurations to deal with specific types of behavior.
I feel like if a specific browser, ie. Firefox went down this path, it wouldn't take long until websites stopped supporting that browser. It'd take browser wide action to take on big players such as Slack imo.
This is a common error people make. It’s transferring their emotions to inform what some (often companies or countries) should do.
In the specific example, it makes no sense for any single website to forgo, say, 5% of revenue to “teach browsers a lesson”, especially since any change would benefit them accrue mostly with their competitors.
See also: “Google/FB/... should leave the EU to protest these privacy laws...”
It may not; you'll just get fake notification prompts from the page that will morph intro instructions on how to enable notifications when you click "accept".
"OneSignal Prompts are a 'soft request', meaning that they are not invoking the 'hard request' of the browser's Native Permission Prompt. This is important because if a user denies the native prompt, the developer is unable to prompt the user again, unless the user goes through a multi-step process to re-enable these permissions. On the other hand, if a user dismisses a Soft Prompt, the app or website can still present them the option later on."
Speaking for myself, showing such prompts repeatedly is a very good way to ensure that I'll never visit your website ever again. No, I don't care about your mobile app, anymore so than I did two days ago.
To provide some context on this, we provide clients with 4 built-in prompting options and we do our best to encourage them to use the one that is best for their user experience.
We definitely want to discourage people from asking for notification permission in an obtrusive way. It's not good for anyone when that happens. We wrote a blog post with some recommendations here: https://onesignal.com/blog/web-push-permission-prompting-cha...
I'm not sure how to put this, but your product is "abused" to such an extent that I personally took the time to pop open the Web Inspector, find the CSS selector for the notification, and use it to look up your company and mentally blacklist it and make a point of bringing it up by name when relevant: https://news.ycombinator.com/item?id=21848866. Your product is seemingly fundamentally misaligned with how users want to interact with the web, to the point where browsers are implementing behaviors that specifically target your implementation. Your blog post gives me very little confidence in your commitment to making this experience better:
> OneSignal encourages all websites to adopt a two-step prompting system if they don't already do so. This will help prevent the website from being penalized and forced to show a quieter permission prompt if too many users have denied the traditional native prompt in Chrome.
Websites are being penalized for a reason, and you know this. Simply adding another modal is just making the problem worse.
> The easiest prompt to transition to is OneSignal's Slide Prompt. Ideally, however, we encourage users to use the Custom Link prompt.
The Slide Prompt should not even be offered as a transition. It's completely at odds with the intent of this browser change, which is all about user intent. Providing facilities to present annoyances that bypass user interaction is the complete opposite of this.
> Offering a coupon in exchange for users opting-in to notifications.
…no?
Please, please, please reconsider how you are implementing these. If you are genuinely unaware of the widespread misuse of your tools, I'd be more than happy to direct you to numerous examples of zero-click, near-immediate, largely irrelevant faux notification requests coming from websites using your product.
I keep trying Chrome out every so often, but there's little usability issues that keep bringing me back to Firefox. For example, Chrome never wants to bring me to the correct sites when you start typing letters. It's always a site I maybe visited once (like nypost instead of news.ycombinator.com)
That's because one of Chrome team's KPIs is how many Google searches are performed through the Omnibar, and better tab search would decrease that KPI and decrease Google ad revenue.
Think about it: it's a web browser literally built by the biggest search company in the world. The latest version of Chrome will even add Google Drive files to the omnibar search for G Suite users.
There is no excuse that they have a worse omnnibox search than Firefox.
My biggest gripe with Firefox is how it opens tabs. If I open a new tab in Chrome, it opens just to the right of the current one which is nice because it keeps related tabs together. In Firefox it opens to the far right which is annoying. I tried installing an extension to fix this but it was buggy. Only recently did I discover that I could dig into about:config to enable the behavior I consider expected (and I'm sure most people agree with me given how many Chrome users there are).
You have a typo there, it's browser.tabs.insertAfterCurrent and browser.tabs.insertRelatedAfterCurrent (Current, not Content), just in case someone tries to copy and paste those into about:config
Okay, so I don't really post to HN and this may seem like a silly complaint but Firefox's management of theming just seems off to me. I use macOS, I don't fiddle about with my wallpaper very much and so this stains Firefox's top bar with an ugly brown tint. I could switch to the light theme, however when I do so Firefox no longer respects my system change to dark mode.
Does anyone else get bothered by this? Is there any way around this? I'm not an aesthetics person by any means but this is quite annoying. It's a fantastic browser other than this, I've been a user for as long as I can remember.
I'm sure you can probably modify userChrome and all this and I've not really tried because I'm not THAT bothered by it, but surely there should be a way to set your preferred themes for light & dark?
If I'm reading your correctly you'd like to set FF to the light theme (or any theme) but continue to have dark mode pages? about:config browser.in-content.dark-mode true should do that.
Re: picture in picture - anyone else here think that this feature doesn't belong to a browser? It just seems so... erm... random, basically.
Looks like something that was cloned from some other product and rather crudely shoehorned in. At the very least it should have been introduced after an update and given an option to opt-in to using it, rather than automatically enabling it without any notice.
I couldn't get it either then tried it with a Twitter video just yesterday.. which made me realise what use it might be, were it not for Twitter muting videos that are not in focus.
It's built in because the browser is best equipped to provide this feature universally for HTML <video> elements and (if necessary) lie to the page about whether it's currently active.
I was kind of skeptical about it but given that in practice all it does is display a small popout icon on video mouse-over, it's not really that annoying.
I absolutely love the ability to do PIP. sometimes while working I'd want to view a stream in a small window (always on top) in a small corner of my screen without all the extra options the browser page has. Its a great feature and im happy FF gives me the option
I think it's a sensible feature but should have been added just as a button on the standard video controls and as part of the right click menu. If a website goes out of the way to abstract both of those things into site controls then the browser shouldn't be adding additional menus on top to try to get around it it should let the site decide to add the button or not.
A great example is Twitch (which I would have assumed was one of the prime use/test cases). Twitch has these things they call "extensions" that streamers can add to the player of anyone watching the stream... and they pop out on hover from the right when you want to access their main settings. So instead of clicking the button to modify the extension suddenly the video is in the corner of my monitor.
Meant to imply "just as a" for the right click and add to the standard video controls. For now I have media.videocontrols.picture-in-picture.video-toggle.enabled false which does half of this.
The use case of PiP is legit but I agree it doesn't make a lot of sense for the browser. However, many users do not have proper native media players that would provide the same functionality so I can see why browsers might want to implement it for the casual users.
A native media player with youtube-dl is still king for power users.
I'm way into it. I was using the feature in Safari to have some low-key video game streams on while I work. I suppose it's not for everyone, but I'm stoked about Firefox having this feature.
I find this feature extremely useful in Chrome where I can run a YouTube video in PiP mode, then have it sitting nearly over a second terminal pane in my IDE while I'm developing.
It's actually quite convenient to watch a video while you're browsing elsewhere (e.g., a passive video you aren't intentionally watching all of but also trying to do other things). I didn't think I'd use it but every once in a while it's nice.
I love it and use it multiple (25+) times per day. My only wish is that it kept the interface from whichever player the videos comes from (eg. YouTube's, Netflix's, etc., especially for subtitles & platform-specific controls like Speed)
That would be a tough one. There's no information which part of HTML is controls. Maybe they may hardcode this information for some popular players, but that would break as soon as for example Youtube decided to change the layout a bit.
This is a web standard thing (though I believe FF's implementation may be proprietary.) Safari, Chrome, etc. have it, why would you not expect Firefox to? It's very useful for browsing while you watch a video open in another tab.
I absolutely love this feature and the fact that I don't have to use a 3rd party extension is even better. It's also practical for websites that autopause videos as soon as you switch to another tab which I find ridiculous.
I'd love for them to add a progress bar and subtitle (think youtube, netflix or prime video) in the PIP window.
I've been trying out the Firefox picture-in-picture feature for a few months now and its evolved into something I use every day.
I don't know whether it's right for a browser to add that feature rather than the OS, but Ubuntu doesn't have any plans to add similar functionality so I don't really care.
As far as making it opt-in I dont really understand your reasoning. If people want to use the feature then it's there, if not then they probably won't notice anything is even different. Do you feel similarly about things like reader mode?
Does fingerprint blocking just use a blacklist, like ad blockers? Seems like a pretty lame approach. Shouldn't they focus on approaches that work for all future scripts? I.e. making browsers seem less unique.
It's complicated. There is in fact a hidden option that makes the web browser less unique, but using it has side-effects such as retrieving pages in English with UTC timezone, noticeably reducing the browser view point size, breaking WebGL, break some JavaScript using <canvas> and most notably, making the web browser noticeably slower. It's not something that a web browser can afford to have without informing the user of the consequences. Not to mention without additional measures, this is unlikely to help much against fingerprinting.
Yes, it’s a blacklist. However, there is ongoing work to improve privacy.resistFingerprinting (enabled in about:config) which makes the browser more similar to other users with this setting enabled.
TLDR: if js gets direct control of GPU, it can read/modify present and past screens.
Maybe after Wayland implementations get good enough, distributions will make hw acceleration work securely with wayland compositors, probably years? Just guessing.
Playing video on youtube is easy, but performance sucks due to bloated youtube website. Performance of the same video URL in empty page with video player is much better, close to that in video player with CPU only.
If you need GPU acceleration, you can use something like youtube-dl or mpv to play videos on a URL in a video player.
I do wonder how much could be saved on energy bills for one of the most common uses of the net (watching videos), especially for folk in developing nations.
Oh dear. Video acceleration means GPU is doing what CPU would be else doing. Which means user threads get less indirect access to GPU. Which is a potential security risk, because GPUs and GL are designed to perform, not to secure data.
VA-API is not mentioned in the question at all. It is the most common api for video acceleration, but there are others - direct vdpau, or nvdec.
We're currently rewriting most of the graphics rendering of Firefox (with what is called the WebRender project). This is the first time Linux users will have hardware accelerated rendering by default, and is a pre-requisite to take real advantage of hardware-accelerated video decoding. It's enabled on some configuration in Nightly builds [0], and it's being worked on actively. If you want to know more, the graphics teams has a blog where updates are periodically posted about what is happening, often with interesting insight about the innards of a web browser rendering stack [1].
Once rendering is hardware accelerated, it makes sense to look into hardware video decoding. Doing hardware video decoding without hardware rendering roughly means we need to read back from GPU memory into main memory, composite the image in software, and then upload back to the GPU to display, which is super super inefficient.
- Open new windows more quickly. Firefox feels sluggish (on Mac) even though it isn't, simply because it opens new windows far more slowly than Safari or Chrome.
- Use the platform native key store. I don't want my passwords stored unencrypted on disk. But I don't want to enter a separate master password either. I do want to use fingerprint/face unlock on mobile to reveal passwords.
- Give me a setting to autoconfirm all cookie consent requests and lobby for a legally binding do-not-track header. Cookie consent was well meaning, but it has turned out to make things worse. Let's move on.
> - Give me a setting to autoconfirm all cookie consent requests and lobby for a legally binding do-not-track header. Cookie consent was well meaning, but it has turned out to make things worse. Let's move on.
Most of those are implemented in JS, so NoScript goes a pretty long way to fixing this. It would be nice if Firefox somehow auto-detected them, and other useless pop-overs, and nuked them.
Unfortunately, browsing with JavaScript disabled on mobile, I've noticed quite a large number of sites (usually WordPress blogs) that have the opt out button in the HTML, controlled server-side, so you have to enable JavaScript at least once to get rid of it.
Given that I prefer brave or bromite custom tabs (via Lynket) for short sessions on mobile (and prefer not to have long sessions on mobile at all), I generally just deal with it.
> Open new windows more quickly. Firefox feels sluggish (on Mac)
I assume you mean new tabs?
That's very quick on Windows using FF 72. Interestingly that's been a long-time annoyance of mine when using Chrome on Mac - when compared to Safari which shines here.
Many of my workflows begin by opening an new empty browser window. I also tend to open links in new windows instead of tabs. I'm not a huge fan of tabs generally.
Whilst there are the "every tab in o e window open forever" people, there are plenty of "window for task" people.
Personally, I have been a window for task person 'since forever' (tabs are context depth in said task, hence "close tabs to right" being one of the most useful features).
> - Give me a setting to autoconfirm all cookie consent requests and lobby for a legally binding do-not-track header. Cookie consent was well meaning, but it has turned out to make things worse. Let's move on.
This can be done with uBlock Origin and an "annoyances" filter list such as EasyList Cookie. It doesn't actually give websites consent to use of cookies, only hides the consent form.
Windows open more quickly now than before. A few versions ago, it had become unbearably slow (I'm an FF user since Camino died), and I stuck to using tabs only, but 72 seems to have caught up again. I estimate that it's around 0.3s slower than Chrome on my machine.
Confirming the window opening speedup! Previously FF spent a lot of time compiling shaders when opening a new window. Perhaps these are now been cached.
Oh that's very nice. In iOS 7 we had a jailbreak tweak called NoSlowAnimations[0]. It was amazing how much a simple thing increased the perceived speed and responsiveness of the OS.
Don't have an iphone handy but I think so. There are lots of options there to explore, there's also one to increase contrast and another to make buttons more evident. There's some cool stuff there in accessibility options
> lobby for a legally binding do-not-track header.
DNT is pretty dead, and IMHO was never a good idea in the first place. Opt-outing of invasive and unethical tracking is just weird. What about people who don't know about it? Or don't fully understand what it means?
It's almost like the Hitchhiker's Guide: "Well, you should have visited the planning department in the disused lavatory with a sign 'beware of the leopard".
Well, if you're going to lobby for a change to the law then you might as well lobby for a change that's actually a good one, instead of a half-baked one that barely works.
I would say it depends on how good any alternative solution is and how realistic it is to pass.
Edit: Having read your blog post, I want to say that I'm not necessarily asking for the original definition of DNT to be reused. What I want is for the DNT header to have the same effect as if I had said no to all optional categories of cookies.
The status quo is that we are asked a "yes or manage" question where "manage" means something horribly complicated that no one does. If we redefine DNT to mean "no to everything optional" then it becomes a matter of local law to define what is and isn't optional. This allows for the sort of minimum level of consumer protection that you're asking for.
The law could also require that this DNT header is opt-out. A per-site overridable DT header if you will.
I'm not going to claim that I have thought this through properly. It's just something that I wish Mozilla did some work on.
I always thought the point of DNT wasn’t technical, but to make it impossible to argue either (a) that users don’t really care (many opted in), or (b) that Google et al respect user’s wishes without regulation or technical barriers (they are aware of DNT and its specific meaning and they ignore it). In other words, political maneuvering towards getting the entire industry to accept implementing that technical solution.
Regarding the native key store, why not just use full disk encryption? The session-specific encryption never made much sense to me, unless you share the computer with others. How many processes do you really have that are not running under your user or root?
I do use full disk encryption, but when I'm logged in, any program that runs as me can access my files (subject to some partial restrictions that Apple introduced a while ago). And anyone gaining access to my unlocked computer can easily copy that password file.
You could say once malware is executed with my credentials it's game over anyway. But I disagree with that. Having a file full of passwords stolen is far worse than than anything else, including key loggers, because it's maximum damage in a minimum amount of time.
I want to make that as difficult as possible for any attacker.
> when I'm logged in, any program that runs as me can access my files
On Windows, the same is true of the encrypted data; any program that runs as you can decrypt it. Is it different on macOS? How does the system authenticate a specific program?
On the Mac, Safari and Chrome store passwords in a key chain. To reveal any of those passwords you have to re-enter your macOS login password or whatever authentication method was used to log into the device (fingerprint, face ID, ...).
So it's not the program that is authenticated. The system simply makes sure that any program accessing that particularly sensitive data is really controlled by the logged in user.
This is entirely separate from protections for other files that may or may not be encrypted.
That's interesting and I wonder how they do it. Nevertheless this isn't good as it offers a false security sense since the end result is similar to using a master password which is the same as the login password.
You are more likely to lose your data while the computer is on and drives decrypted than when it is off. The threat model for FDE is thieves or government physically stealing your drive. Much more people get pwned in software than physically. I mean you don't just leave root shells laying around do you?
The question was whether the keychain really protects from those software attacks. I wondered mostly because I know that Windows will allow any process running as a certain user to decrypt data stored by any other process (at least if it's stored using the Data Protection API, like Chrome does for its passwords), so it only really protects your data from other users and their malware.
It seems that macOS authenticates each process, so that might provide some extra security.
>- Give me a setting to autoconfirm all cookie consent requests and lobby for a legally binding do-not-track header. Cookie consent was well meaning, but it has turned out to make things worse. Let's move on.
Maybe there should be standardized interface for accepting privacy policy and cookies, all managed in browser UI so you could set default settings and give exceptions for specific sites without cluttering website with various popups... oh wait https://en.wikipedia.org/wiki/P3P
"Say goodbye to annoying notification requests (unless you want them)"
For a moment I thought they had done this for the cookie and privacy notices. Oh how that would be amazing to move that functionality into the browser.
Switched to Firefox recently on my desktop. I had the strangest problem with Chrome: Somehow it just kept forking more and more processes until the rlimit got filled (100k+ threads/processes).
Too much trouble to debug it so I opted to switch instead.
542 comments
[ 3.0 ms ] story [ 422 ms ] threadEdit: Answered by zamadatix.
https://hacks.mozilla.org/2019/09/moving-firefox-to-a-faster...
i'm using the latest macos and chrome. what exactly isn't supposed to be working?
i also have firefox installed, but on macos firefox is mostly an alpha/unfinished release since years ago.
While I still consider DuckDuckGo to be in the "not bad" category, Firefox is in the "seriously awesome" category now.
If you have been away from Firefox for a while like me, give it another shot. It won't disappoint you for sure.
You won't be worse off, it isn't a big effort to get used to FF devtools, but you must accept a learning curve if you want to switch.
I had to get used to Chrome devtools because javascript debugging is a lot better there. For completeness sake, CSS debugging is much better in Firefox. In time, I think both will converge.
That being said, I still switched back to Firefox and ddg after I close devtools.
> Firefox now supports simulation of meta viewport in Responsive Design Mode.
we had a recent dialog about it here:
https://old.reddit.com/r/javascript/comments/ecyhmr/mozilla_...
there's some cool stuff in the works but it doesnt sound like there's a plan to internalize the work into devtools but to keep it as a web app and rely on a serviceworker for "localness" - not great, imo.
I've found it's extremely valuable to be developing sites in Firefox using the dev tools simply to help catch issues that the rest of the team ignores as they're all stuck on Chrome. Fights monocultures by default!
Maybe make a perf switch in the console, which dumps the text without formatting or other features. Something that can handle one million lines of logs and search in it. Being able to search the logs is critical. Filtering doesn't work as I need to see the surrounding logs, what happened after. I don't need to see all one million lines at the same time though, it can be buffered to only show 100 lines above it and 100 lines below it. Chromium has such an optimization. But search is a bit slow.
I also have to keep clicking the clear button to prevent the page from freezing if the dev tools console is open when loading the page.
I generally like the FF tools better nowadays, but Chrome’s are a bit faster.
Really like the ability to outline flexbox and grid elements.
Rather than having to remember all the small changes you made, you can just copy and paste the script out (or even just view what you changed).
I have Firefox in the dock on iPhone and it's really great but random links, e.g. via messaging, get opened by Safari first.
[0] https://blog.mozilla.org/futurereleases/2019/12/03/firefox-p...
For me its been fantastic.
The only thing I missed from the ordinary Firefox on Android was that Firefox used to have a brilliant hack with the sharing option where they would show the two most frequently used options (+ a sharing button to access the rest) instead of just one or just the sharing button.
This saved me a few seconds again and again and again.
Wonder why no one else does that?
And there really needs to be an easy way to switch profiles. Firefox has had profiles for basically forever, but it's much easier to switch profiles in Chrome.
[1] https://support.mozilla.org/en-US/kb/containers
Of course you can also create launchers/shortcuts for Firefox with "firefox -P profilename" but again that's work you have to do on each machine you use.
If you're into privacy-conscious search engines and want one that's not susceptible to blanket FISA warrants[0], check out either Qwant[1] or SwissCows[2].
[0] - DuckDuckGo's address is in Pennsylvania - 20 Paoli Pike, Paoli, PA 19301.
[1] - https://www.qwant.com
[2] - https://swisscows.ch/
EDIT: I right click the search box and make qw and sw keywords for them.
EDIT2: I'm just looking for a functional google alternative.
In the console I get a warning and an error:
> Content Security Policy: Couldn’t process unknown directive ‘script-src-elem’
> TypeError: browserLanguage is undefined
France has a more permissive surveillance regime than the US[0], and as a foreign (US) user I have no recourse if they spy on me.
Furthermore, here are some quotes from qwant:
>We do not use any cookies or other advertising tracers to create your profile for commercial purposes. This means that you can use Qwant with confidence, we will never try to establish your psychological or commercial profile to sell it to third parties, here or elsewhere. [1]
Note the "for commercial purposes" in the first sentence, and the "to sell it to third parties" in the second. These are weasel clauses.
[0] https://www.theguardian.com/world/2015/jul/24/france-big-bro...
[1] https://betterweb.qwant.com/5-reasons-to-be-brave-and-choose...
Hahahahaha, enough with the jokes. Comparing France to the US... Thinking you actually had some kind or chance of recourse against any of the three letter agencies. You are funny!
> Note the "for commercial purposes" in the first sentence, and the "to sell it to third parties" in the second. These are weasel clauses.
Could be, could not be. With Google you can be sure that they sell your data to _anybody_ that is willing to pay something for that.
This is not true. There is no documented case of Google selling any user's personal data to anyone.
That should probably read 'more pervasive'.
Also >France
https://news.ycombinator.com/item?id=21371577
I don’t necessarily mean that good or bad. Just interesting how... different it is. Seems good if you know your thug exists already and someone approved it for their results.
Me too! I'm not as bothered by software-politics as many of the commentators on here (I don't hate Facebook, and worked there for four years; I don't hate ads; I don't worry too much about trafficking and privacy...) but Chrome fucking with the URL was a bridge too far. I switched to Firefox and when Google walked back their hiding-the-URL business, I never bothered switching back, because I didn't trust them anymore.
Companies take note: users' trust that you won't arbitrarily change their workflow in annoying ways is hard to win back.
- Do they really need to see* http:// or https:// if immediately to the left is the icon which indicates secure/insecure (which is more than just "protocol is https or not")?
- Do they really care to see* if they are at www.site.com or m.site.com or site.com or do they just want to know they are at site.com which they expect to be as they make their purchase or log in?
* noting that bookmarking, copying, or editing the URL operate on the full URL and this is just a visual change otherwise
And voila now the leftmost (first) text they read is not a bunch of jargon it's the thing people look at the URL for the most.
That you couldn't think of any reason tells me you didn't try not that you disagreed with the reasons and had an actual lesson to share about why. Again: "it changed, change bad" people impossible to distinguish from "it changed, actually bad".
As you said www.example.com being functionally different from example.com was already a broken workflow, users weren't differentiating it. Continuing to display something users haven't understood for 20 years was not considered a strong enough reason to simplify the URL display now in a way that causes the same errors users were commonly making anyways.
Maybe dropping the protocol definition in the URL is debatable, as probably 99.99% of the time in Chrome a user is going to be either http or https, but dropping a part of the hostname is unacceptable in my book.
Would dropping .com be acceptable? All a user wants to see is that they went to Google, to an average user the .com could be seen as redundant clutter as well. Might as well have the address bar show "google" or "cnn" or "facebook" in it, as clearly that's what a user cares about.
You can't really get rid of the TLD without breaking the security model of the web (e.g. a lock icon and google.com is different than a lock icon and google.gtld). Would it be nice to refactor that? Probably. Is it reasonably possible at this point? No.
You mean the strawman repeatedly trotted out by people defending shitty changes as justifications for why their shitty change is necessary? Ever wonder why so many things suck nowadays?
Can "the larger group" be used as an excuse for a bad change? Sure! But so far the only reason mentioned as to WHY this change is bad is "www.site.com and site.com could technically be different sites" and I'm not sure how that is supposed to be clear regardless if www shows post load or not or how that's supposed to outweigh the advantages.
HTTP currently shows "Not secure" to the left of the URL. It's still easy to differentiate.
(I do think hiding www is stupid, but hiding http vs https is reasonable so long as they're still marked differently. Ideally http should be a more noticeable error, like it is if there's any user input box, or better yet a massive error like it is for self-signed certs and HTTPS.)
I care what I need.
A standard user was probably fine with something like AOL keywords.
If other people have different preferences, that’s fine! They can keep using Chrome.
Also, even if that weren’t true, change is a cost. It’s perfectly reasonable to complain about change as being intrinsically bad without giving any other reason.
https://www.bleepingcomputer.com/news/software/chrome-and-fi...
E.g. "<lock icon> https://www.google.com/example/" would display as "<lock icon> google.com/example" until you interact with the URL.
It's still not as quick as Chrome (my laptop is not recent, not helping) but I'd rather support Mozilla.
https://news.ycombinator.com/item?id=21945206
It is not 100% bug free yet, though
But is it good enough? Hell yes.
HN discussions:
https://news.ycombinator.com/item?id=20050173 (281 comments)
https://news.ycombinator.com/item?id=20044430 (893 comments)
https://news.ycombinator.com/item?id=21371577
I often had a window open only playing a youtube video, having it be native to the browser and always on top is such a great feature.
App developers spend a lot of their time making up for the poor decisions of OS developers. A good window manager would give you tabs for free, PiP (aka an always-on-top window) for free, dark mode for free. Instead OS devs go overboard in simplifying things to the point that app devs have to pay the cost of delivering what users want.
I'm stuck on a Mac, where I cannot always on top, hide the title bar, or alt-drag windows to move them.
With a proper PiP mode, that's no longer an issue.
On the other hand, Firefox keeps the close button for tabs on the right, even in macOS where every close button is on the left.
Thankfully, I've found a bit of userChrome that corrects this, as a workaround.
It can create, for example, a pop-out video overlay, which can be moved around and resized. It then sits over top of the other content on the page (and remains over other tabs) and will remain fixed as you scroll down the page and similar. On a site like YouTube, you could pop the video out and proceed to scroll down and interact with comments while the video plays to the side or above.
https://support.mozilla.org/en-US/kb/about-picture-picture-f...
[1] https://www.mozilla.org/en-US/firefox/73.0beta/releasenotes/
They should have thought about this - it is possible to force resolve the individual domains in the NextDNS UI but that gets problematic fast.
If you just want to know how to toggle it normally in case something goes weird with one of the changes it's in the network settings part of the settings page.
edit: there is also the canary domain https://news.ycombinator.com/item?id=21825443
HDR is very much in the list of things we want to do (especially for videos), but as you can imagine, it's not particularly trivial, and will require platform specific work, and deep changes into Firefox. On the other hand, hardware is shipping (we got some devices so that developers can start assessing the situation), and HDR content is becoming increasingly available, so we're trying to figure out what it will take to do it properly.
Update: Just got to try it out, seems like it indeed works now. Yay Firefox 72!
This is fantastic! I had finally figured out I could turn this off in settings a while ago, glad it's now a default. I get so annoyed by this, annoying indeed!
In a webapp where you present user a button to activate notifiactions, when the user clicks the button seemingly nothing happens in FF72 (user is focussed on a big enable notifiactions button in the web app and may not notice that some tiny gray icon wiggled a little in the address bar).
On a big screen a button in the middle of the screen is so far away from the address bar, that you don't see any change in the address bar at all in the peripheral vision.
So yeah, web apps that don't try to force the user to enable notifications are now punished for good behavior again.
EDIT: So it's not so stupid, see below.
This will just not work anymore. I'll have to add some long winded FF specific explanation on what to do after clicking the button.
Anyway the FF UI is broken if user physically can't see any reaction from the browser after he requests notifications by clicking anywhere on the page.
https://hacks.mozilla.org/2019/11/upcoming-notification-perm...
This is a frequent "mistake"/issue however, and we're working on a mitigation for it.
https://megous.com/dl/tmp/sub.txt
I guess this is then happening because I'm first checking if there's a subscription via `pushManager.getSubscription` before creating a new one in the event handler.
Yep, dropping getSubscription makes the popup appear again.
Do you mind me asking what your notifications are for?
https://addons.mozilla.org/en-US/firefox/addon/multi-account...
[1] https://addons.mozilla.org/en-US/firefox/addon/multi-account...
[0] - https://support.mozilla.org/en-US/kb/containers
I used to have this script built as an app bundle via Platypus, but that stopped working somewhere between Yosemite and Mojave (likely due to increasingly strict OS file permissions or something). The Automator app doesn't seem to suffer from this issue...I think I just had to grant it a couple permissions the first time I ran it.
I wasn't able to figure out how to set a custom icon from within Automator, but it's not too hard to do if you're comfortable mucking around inside app bundles. The icon file is stored under <app-bundle-root>/Contents/Resources/, and its name (sans extension) is referenced in <app-bundle-root>/Info.plist.
[0]: https://gist.github.com/ilikepi/9d2e17e0d3b3efd6fc0584f46f09...
That should pretty much kill off a lot of the notification request crap, especially if Chrome follows suit. I can envision the conversion rate massively falling off when it's no longer something right in your face.
FF needs a “disable notifications, but lie to the website and say they’re enabled” button.
This does not seem like a reasonable complaint to me.
In the specific example, it makes no sense for any single website to forgo, say, 5% of revenue to “teach browsers a lesson”, especially since any change would benefit them accrue mostly with their competitors.
See also: “Google/FB/... should leave the EU to protest these privacy laws...”
In preferences type 'notifications', click the Settings button and check the box to "Block new requests asking to allow notifications"
That was the era when majority of public opinion on Google was Do no Evil. And Google at the time can do no wrong.
"OneSignal Prompts are a 'soft request', meaning that they are not invoking the 'hard request' of the browser's Native Permission Prompt. This is important because if a user denies the native prompt, the developer is unable to prompt the user again, unless the user goes through a multi-step process to re-enable these permissions. On the other hand, if a user dismisses a Soft Prompt, the app or website can still present them the option later on."
Speaking for myself, showing such prompts repeatedly is a very good way to ensure that I'll never visit your website ever again. No, I don't care about your mobile app, anymore so than I did two days ago.
To provide some context on this, we provide clients with 4 built-in prompting options and we do our best to encourage them to use the one that is best for their user experience.
We definitely want to discourage people from asking for notification permission in an obtrusive way. It's not good for anyone when that happens. We wrote a blog post with some recommendations here: https://onesignal.com/blog/web-push-permission-prompting-cha...
> OneSignal encourages all websites to adopt a two-step prompting system if they don't already do so. This will help prevent the website from being penalized and forced to show a quieter permission prompt if too many users have denied the traditional native prompt in Chrome.
Websites are being penalized for a reason, and you know this. Simply adding another modal is just making the problem worse.
> The easiest prompt to transition to is OneSignal's Slide Prompt. Ideally, however, we encourage users to use the Custom Link prompt.
The Slide Prompt should not even be offered as a transition. It's completely at odds with the intent of this browser change, which is all about user intent. Providing facilities to present annoyances that bypass user interaction is the complete opposite of this.
> Offering a coupon in exchange for users opting-in to notifications.
…no?
Please, please, please reconsider how you are implementing these. If you are genuinely unaware of the widespread misuse of your tools, I'd be more than happy to direct you to numerous examples of zero-click, near-immediate, largely irrelevant faux notification requests coming from websites using your product.
Think about it: it's a web browser literally built by the biggest search company in the world. The latest version of Chrome will even add Google Drive files to the omnibar search for G Suite users.
There is no excuse that they have a worse omnnibox search than Firefox.
Edit - default for this config variable could be platform specific; the above is based on Windows 10. I'll check on my MacBook tonight.
Does anyone else get bothered by this? Is there any way around this? I'm not an aesthetics person by any means but this is quite annoying. It's a fantastic browser other than this, I've been a user for as long as I can remember.
I'm sure you can probably modify userChrome and all this and I've not really tried because I'm not THAT bothered by it, but surely there should be a way to set your preferred themes for light & dark?
[1] https://imgur.com/a/t9tvbQr [2] https://i.imgur.com/GXrTw7C.png [3] https://old.reddit.com/r/unixporn/comments/9f29vr/i3_doom_co...
https://reddit.com/r/firefox/comments/dgth9i/automatic_dark_...
Looks like something that was cloned from some other product and rather crudely shoehorned in. At the very least it should have been introduced after an update and given an option to opt-in to using it, rather than automatically enabling it without any notice.
I really dislike Twitter.
I was kind of skeptical about it but given that in practice all it does is display a small popout icon on video mouse-over, it's not really that annoying.
Using an embedded youtube or other video player that doesn't let you resize? Pop it out!
Want to put a youtube vid in a corner while you do other stuff? Pop it out, switch tabs, and go about your business!
For example, I love it for Netflix so that I can watch a show while doing other things on my laptop.
A great example is Twitch (which I would have assumed was one of the prime use/test cases). Twitch has these things they call "extensions" that streamers can add to the player of anyone watching the stream... and they pop out on hover from the right when you want to access their main settings. So instead of clicking the button to modify the extension suddenly the video is in the corner of my monitor.
That's when I decided to disable it for now :).
Edit: clarification
It is in the right click menu, under "Picture-in-Picture"
A native media player with youtube-dl is still king for power users.
It's actually quite convenient to watch a video while you're browsing elsewhere (e.g., a passive video you aren't intentionally watching all of but also trying to do other things). I didn't think I'd use it but every once in a while it's nice.
Most content providers probably don't like this feature.
I'd love for them to add a progress bar and subtitle (think youtube, netflix or prime video) in the PIP window.
If we went by this argument then browsers would have every single feature ever. Which they kinda do...
I don't know whether it's right for a browser to add that feature rather than the OS, but Ubuntu doesn't have any plans to add similar functionality so I don't really care.
As far as making it opt-in I dont really understand your reasoning. If people want to use the feature then it's there, if not then they probably won't notice anything is even different. Do you feel similarly about things like reader mode?
One downside is that subtitles (at least with Netflix, don't remember for YouTube) are still displayed in the original tab.
You are probably better of using Tor Browser.
https://security.stackexchange.com/questions/13799/is-webgl-...
TLDR: if js gets direct control of GPU, it can read/modify present and past screens.
Maybe after Wayland implementations get good enough, distributions will make hw acceleration work securely with wayland compositors, probably years? Just guessing.
Playing video on youtube is easy, but performance sucks due to bloated youtube website. Performance of the same video URL in empty page with video player is much better, close to that in video player with CPU only.
If you need GPU acceleration, you can use something like youtube-dl or mpv to play videos on a URL in a video player.
The question is about whether FF is finally using va-api to accelerate video codec decoding.
VA-API is not mentioned in the question at all. It is the most common api for video acceleration, but there are others - direct vdpau, or nvdec.
Once rendering is hardware accelerated, it makes sense to look into hardware video decoding. Doing hardware video decoding without hardware rendering roughly means we need to read back from GPU memory into main memory, composite the image in software, and then upload back to the GPU to display, which is super super inefficient.
[0]: https://wiki.mozilla.org/Platform/GFX/WebRender_Where [1]: https://mozillagfx.wordpress.com/
- Open new windows more quickly. Firefox feels sluggish (on Mac) even though it isn't, simply because it opens new windows far more slowly than Safari or Chrome.
- Use the platform native key store. I don't want my passwords stored unencrypted on disk. But I don't want to enter a separate master password either. I do want to use fingerprint/face unlock on mobile to reveal passwords.
- Give me a setting to autoconfirm all cookie consent requests and lobby for a legally binding do-not-track header. Cookie consent was well meaning, but it has turned out to make things worse. Let's move on.
Most of those are implemented in JS, so NoScript goes a pretty long way to fixing this. It would be nice if Firefox somehow auto-detected them, and other useless pop-overs, and nuked them.
Given that I prefer brave or bromite custom tabs (via Lynket) for short sessions on mobile (and prefer not to have long sessions on mobile at all), I generally just deal with it.
I use I on Firefox mobile and desktop - don't know if there's a chromium version.
Almost forgot I use it which is the most positive review an add-on like this could receive.
Looks available on a gamut of different browsers.
Firefox addon: https://addons.mozilla.org/en-US/firefox/addon/i-dont-care-a...
[1] https://news.ycombinator.com/item?id=21980909
I assume you mean new tabs?
That's very quick on Windows using FF 72. Interestingly that's been a long-time annoyance of mine when using Chrome on Mac - when compared to Safari which shines here.
No, I mean windows. Opening tabs is very fast.
Many of my workflows begin by opening an new empty browser window. I also tend to open links in new windows instead of tabs. I'm not a huge fan of tabs generally.
Whilst there are the "every tab in o e window open forever" people, there are plenty of "window for task" people.
Personally, I have been a window for task person 'since forever' (tabs are context depth in said task, hence "close tabs to right" being one of the most useful features).
This can be done with uBlock Origin and an "annoyances" filter list such as EasyList Cookie. It doesn't actually give websites consent to use of cookies, only hides the consent form.
Windows open more quickly now than before. A few versions ago, it had become unbearably slow (I'm an FF user since Camino died), and I stuck to using tabs only, but 72 seems to have caught up again. I estimate that it's around 0.3s slower than Chrome on my machine.
> Use the platform native key store.
Amen to that.
You can disable it for all windows in all applications using a terminal command:
You might need to restart apps to see this change.[0] http://cydia.saurik.com/package/com.marcosinghof.noslowanima...
DNT is pretty dead, and IMHO was never a good idea in the first place. Opt-outing of invasive and unethical tracking is just weird. What about people who don't know about it? Or don't fully understand what it means?
It's almost like the Hitchhiker's Guide: "Well, you should have visited the planning department in the disused lavatory with a sign 'beware of the leopard".
I wrote some more about it over here: https://www.arp242.net/dnt.html
Edit: Having read your blog post, I want to say that I'm not necessarily asking for the original definition of DNT to be reused. What I want is for the DNT header to have the same effect as if I had said no to all optional categories of cookies.
The status quo is that we are asked a "yes or manage" question where "manage" means something horribly complicated that no one does. If we redefine DNT to mean "no to everything optional" then it becomes a matter of local law to define what is and isn't optional. This allows for the sort of minimum level of consumer protection that you're asking for.
The law could also require that this DNT header is opt-out. A per-site overridable DT header if you will.
I'm not going to claim that I have thought this through properly. It's just something that I wish Mozilla did some work on.
You could say once malware is executed with my credentials it's game over anyway. But I disagree with that. Having a file full of passwords stolen is far worse than than anything else, including key loggers, because it's maximum damage in a minimum amount of time.
I want to make that as difficult as possible for any attacker.
On Windows, the same is true of the encrypted data; any program that runs as you can decrypt it. Is it different on macOS? How does the system authenticate a specific program?
So it's not the program that is authenticated. The system simply makes sure that any program accessing that particularly sensitive data is really controlled by the logged in user.
This is entirely separate from protections for other files that may or may not be encrypted.
Really? I thought Chrome, same as Firefox, stored credentials in an SQL file, the reason behind it being profiles and its built-in sync service.
Autofill doesn't require password entry, but before revealing any stored passwords you have to re-authenticate using the regular system login method.
I wish Firefox did the same.
It seems that macOS authenticates each process, so that might provide some extra security.
Maybe there should be standardized interface for accepting privacy policy and cookies, all managed in browser UI so you could set default settings and give exceptions for specific sites without cluttering website with various popups... oh wait https://en.wikipedia.org/wiki/P3P
Passwords are always stored encrypted when you save them in Firefox. You can read more about it here: https://support.mozilla.org/en-US/kb/how-firefox-securely-sa...
For a moment I thought they had done this for the cookie and privacy notices. Oh how that would be amazing to move that functionality into the browser.
https://addons.mozilla.org/en-US/firefox/addon/i-dont-care-a...
Too much trouble to debug it so I opted to switch instead.