Firefox 79
hacks.mozilla.org
hacks.mozilla.org
Even more importantly, it has overridden data collection preferences after the update. Check "Settings > Data collection". I had to disable "Marketing data" and "Experiments" toggles. Not cool!
Which is more appropriate than the one about Facebook.
For me, there was no indication that this update was going to be any different from a minor bugfix they've rolled out a couple of days ago, until I've launched FF for the first time after the update and was a greeted by a full screen splash with the new version info.
I've been running Nightly and Regular side-by-side for a while. I still use Regular the most, mainly because I find the new UI quite horrible to use. Nightly is clearly much smoother to use though.
Brave sent me a push notification this week too :( what browser should I switch to that doesn't try to exploit my 'engagement'
Uber are pretty bad at it as well
Personally I give that award to Spotify.
In this one case they are right.
it was game-over to your opinion when you purchased their product and use a proprietary client to connect to the their proprietary servers.
If they want to break every single feature, while also denying access to older versions of the client, there's nothing you can do. Besides cancelling the membership, that is. But in this case there is nothing the changelog that would have helped you.
...all that is the complete opposite of Mozilla though.
""" Fix bug where sync could fail in Android 7 devices
Also, for a few users we rolled the following experimental changes:
* X, Y and ZYes, this is a complete and total pain in the neck, but it has saved me in the past. Jailbroken devices can also use tweaks that allow downgrading right from the App Store.
---
[1] https://support.apple.com/en-us/HT208079
P.S. There's actually one more way to downgrade apps—use a version of iTunes from before Apple introduced certificate pinning (iTunes 7 worked for me) and use Charles Proxy to edit the request so Apple gives you an old version. This worked for me as of last fall—but in this case, "pain in the neck" does not even begin to describe it. :)
Quality of changelogs has a big impact on how trustworthy I perceive a project to be. Releasing new versions without a (meaningful) changelog says that you expect blind trust from your users; that you expect them to update anyway regardless of the changelog content; or that you don't consider them able to evaluate whether updating is in their best interest -- in summary, "we know what's best for our users better than they do." Whether or not that's true, the attitude strikes me as vaguely paternalistic. It feels to me that a developer who thinks so little of their users would be more inclined to try and pull a fast one.
As an example, Signal has recently stopped writing changelogs for many of their beta releases, and it has affected my trust in them, and willingness to install said updates.
Let me be very clear: I am NOT saying lack of changelogs makes a developer less trustworthy. I AM talking about how they make me feel about them. I'm also well aware that this is an indirect measurement, so there's some variability and Goodhart's law applies (https://en.wikipedia.org/wiki/Goodhart%27s_law).
WHAT'S NEW
• Bug fixes and performance improvements.
Actual updates: Version number bump. In my experience outside of the Big Tech releases the most egregious offenders seem to be Credit and Banking Apps with auto expiration. I loathe that feature, it's bitten me in low connectivity areas.
Sadly, people have been trained not to because devs always say, "Bug fixes and improvements".
Some people really do want to know what changed.
And then the vendors (and the security people) claim that automatic updates are fundamental, because otherwise people wouldn't install them. But maybe, just maybe, people don't want to install updates because they're being force-fed them, with no information about what to expect, and half of the time the update makes the experience worse.
The road to a better user/vendor interaction starts with providing meaningful changelogs and dropping the paternalistic attitude about updates.
> dropping the paternalistic attitude about updates.
Now that we are trying to treat everyone equally, shouldn't we find a better word instead of blaming it all on the dads? ;-)
At least were I came from I cannot remember this attitude come from my dad.
I have adapted kind of the same attitude: As the one who brings most money to the table I demand a final say on anything major, but in reality my wife has found our last 3 cars, do the majority of the shopping etc, I mostly just rubberstamp and sign and keep telling my kids to respect their mum.
So far it has worked fine : )
Same thing on iOS to be fair.
I love reading the what's new copy. Some companies put a lot of effort into it, like Signal/Slack (when I still had it installed) Some companies leave it empty which I think is very bad
But the worst are companies that put a constant what's new section like "we're constantly updating our app to make it better for you" bullshit - Uber. This just feels condescending to me.
Overall I think it's a prime channel to let people who are interested in your app to know ... what's new
ALL the majors apps I am familiar with basically launch new features in the same way : first under a feature flag (e.g. the feature code is in the app but deactivated by default, we control its activation server side). We open up the feature to a small percentage of users, look at how it is behaving (bugs, crashes, impact on metrics) and then decide whether we want to open it up to more users or not.
With such a system where your version number does not mean much, at best release notes would be "we added a new feature, you might or might not see it".
I would like swiping on address bar to switch tabs, though.
https://www.androidpolice.com/2020/07/27/firefox-now-support...
Here's the thing... sometimes when you unload Fenix and then open a link from another app, the browser will fail to load... you get a white screen... and the page never loads, but at least you can pull to refresh, which does nothing... (press Back to get out of this state)
So the only time you can "pull to refresh" is when Fenix doesn't work!
Uhhh, what the hell. Is that only a thing on Android or should I be checking all of my devices?
Oh ffs thanks for the heads up. Re-defaulting settings is a big reason I jumped ship from other companies' products.
This Firefox Nightly one uses the Gecko rendering engine, while the standard one uses Android's (or Google's?) Chrome Webview. Has this changed?
What has changed is that Fenix uses Gecko via https://mozilla.github.io/geckoview/
Since I have had issues in my app with Chrome/Android WebView, I will start experimenting with GeckoView. I didn't know it existed as a component. Thank you so very much!
For once my paranoia paid off!
On the other side, Visual Studio forces me to open up a port and log in to MS account - tied to my real name and credit card - so it could upload my usage patterns every month, as a condition of the 'free' license. Perhaps that's the model we're converging towards, even for open source.
It would be nice if Mozilla adopted privacy-by-default rather than opt-out.
Not particularly pleased with that.
I really sometimes grow very frustrated with Mozilla. In Cory Doctorow's _Eastern Standard Tribe_, some people work as saboteurs, planting really unpopular stupid ideas, and sabotaging clean functionality.
It feels like Mozilla is really rife with such characters.
With Fenix, Mozilla decided to just abandon that feature. Issues are closed, it got removed from the feature list [1] and further questions are ignored. I fully understand that you can't keep every feature everywhere, but this was THE main benefit of Firefox (besides ublock) for me and if you look at GitHub/Reddit/Twitter I am not the only one.
Now I have to stick to an outdated browser because of an (for me) completely unnecessary, degrading update :/.
I hope some other browser picks it up. Would probably be a good fit for Vivaldi, which is meant for power-users who might go through link aggregators a lot.
- Recently closed tabs (AFAIK "Undo close tab" currently fakes it by not actually closing the tab until the "Undo close tab" popup has disappeared)
- The Firefox share target that actually gave you a choice whether you'd like to open the page in Firefox directly, merely bookmark it or use Sync to send it to some other Firefox instance without having to actually open the page in Firefox first
- Add-on support that isn't limited to a few blessed "Recommended extensions"
- viewing local HTML files is not possible (although admittedly Google hasn't helped there, either, by vastly complicating file system access in recent Android versions, and their purported replacement method is absolutely unsuitable for HTML files that depend on additional resources such as images, styles, scripts, other HTML files etc., but in the end it was still Mozilla's decision to disallow it completely right now)
- about:config
- View source
- bfcache is broken
- cannot force-refresh a page
- the tab import from the previous versions drops all the session history of those tabs, i.e. it only imports the currently viewed page, but you can no longer go back or forward
I personally saw a number of regressions[0] on Debian testing for video playback on the beta releases for 79, but it largely seems to have settled down now.
[0] particularly this one: https://bugzilla.mozilla.org/show_bug.cgi?id=1643855
(Copied from my comment on the submission for the official release notes)
the return of shared memory
> At the start of 2018, Shared Memory and high-resolution timers were effectively disabled in light of Spectre. In 2020, a new, more secure approach has been standardized to re-enable shared memory.
mondo details here https://hacks.mozilla.org/2020/07/safely-reviving-shared-mem...
more secure target=_blank links
> To prevent the DOM property window.opener from being abused by untrusted third-party sites, Firefox 79 now automatically sets rel=noopener for all links that contain target=_blank.
Only took what, three years since that vulnerability was found?
Unthinkingly following Chrome, Firefox started doing this too. To my knowledge, Firefox still defaults to doing this, but I could be mistaken.
"Preferences > Search > Search Bar" allows you to add back the separate bar. And you can toggle search suggestions in the address bar as well. I don't see how that breaks anything or that it is a "serious UI change".
And finally, the security implications of search suggestions in the address bar are debatable for the end user.
And the security patches for FF are bundles with the release, so if you want security fixes you also HAVE to take whatever UI or feature changes FF put into the browser. You shouldnt force users to relearn functionality for a piece of utility software in order to get fixes for CVEs. I understand its easier for them, but having to retrain people every couple of weeks because there is a new FF release isn't ideal.
Yes, exactly. The amount of UI/UX churn in Firefox is pretty crazy and I'm sure discourages some number of people from updating regularly. Mozilla "fixed" that issue on the desktop versions by completely removing the easy options for letting users decide for themselves when to update, rather than just making auto-updates the default.
I just don't get it. Does Mozilla have a problem with too many developers and needing to invent things to keep them occupied with?
> But its crazy that you have to accept serious UI/UX changes to get security fixes too.
It is sad, but it basically seems par for the course for big, semi-commercial software. I don't like Mozilla, but I appreciate them making an alternative to the massive Chrome near-monopoly that's not only just as fast and lighter but competitively easy to use for normal people. I personally switched to Waterfox years ago, and Pale Moon not long after that. It receives some security fixes slightly after FF (they're fixed after Mozilla publishes the issue), but some of them are not applicable (http://www.palemoon.org/releasenotes.shtml) . Overall, to me then Pentadactyl is worth worse than that, and I think you might love the UI.
I totally get why they do it too, Im a big Mozilla fan and have been for a long time. Im glad there is browser competition, I love their take on Add-ons and allowing the user to make decisions for themselves. They are fighting the good fight, Im just expressing my opinions on things I don't like, but Im still going to use FF. I did stop upgrading with version 76 because it was just too much change and disruptive enough for me to downgrade and turn off the installer.
Ive never seen Pentadactyl, Ill check it out.
This is interesting, I always figured it was the opposite. Many casual users don't know or care much about the difference between a search bar and a url bar and just want to type something in and get a result. Personally though I prefer having separate search and url fields because the actions are more precise.
On more than a couple occasions I've tried to make a web search in Chrome's url bar that the browser insisted was a nonsensical url. I know there's a flag to enforce one behavior or the other, but I can't ever remember it and honestly don't feel like having to memorize magic words to get basic functionality working.
It's easy to remember because the shortcut for the URL/Omnibar is more often cited as Ctrl-L.
- Return of shared memory between parts of the same page (including web workers). Parallel processing becomes more efficient, good for complex apps and games.
- Time-traveling debugger of sorts: search for "restart frame".
`dom.workers.serialized-sab-access` is the flag they've put in so that way they can disable concurrent execution of JS threads that share memory in case a novel cross-process attack shows up. Spectre is purely an in-process attack and the whole article that comment is attached to is about the work they did to enable shared memory while defending against Spectre. `dom.workers.serialized-sab-access` does not affect Spectre. It appears to be intended for preventing a novel cross-process attack from leveraging shared memory in other processes into becoming a high-resolution timer.
Also note that threads that have access to shared memory in Firefox 79 also have access to the full high-resolution performance.now(), and flipping dom.workers.serialized-sab-access doesn't affect that.
This is exciting! It opens up faster possibilities for wasm apps
And... Firefox 79 with one active tab is taking 6x more energy than Safari with 20+ tabs.
Maybe it is better on other OSes, but on macOS nothing beats Webkit in terms of performance.
Not to mention the home page bloat - Firefox is starting to look like cnn of browsers.
I just compared clean Firefox install with Safari on macOS and they both show energy impact 0.1
Care to share more details?
Having to see mentions of Nazi camps on my home page or seeing a notification for Facebook containers although I don't use Facebook. All that bloat being enabled by default is troublesome and goes against very principles Mozilla advocates.
Between Chrome, Firefox, Safari and their plugins it's a dance to get the desired behavior and efficiency. Lazy loading of pages (when you have multiple tabs and close/reopen) can either be really efficient (because they don't even load the page) or annoying (if you don't have Internet or they change/remove the page when you go back to read it).
It could also be mention that Firefox now use macos coreanimation that drastically improved Firefox battery performance [1]. So give it a real test again, Firefox is really awesome and needs all support :) We need bigger diversity among webbrowsers, not just chromium based.
[1] https://mozillagfx.wordpress.com/2019/10/22/dramatically-red...
You're complaining that FF is using 1.3% CPU while in the foreground?
And worst of it, no about config. I dont like the direction mozilla is taking. Do they have any reasoning for no abour config.
This is quite a downgrade, I think I am switching to another browser on mobile.
Feels very snappy. Upgrade was a breeze. Only one addon that didn't work, hope full addon support is back soon.
The release notes say "The wasm-bindgen documentation includes guidance for taking advantage of externref from Rust", but I didn't yet find anything about it there either.
- Ctrl-L to go to the address bar
- release Ctrl (otherwise, the next keypress fail)
- Shift-6 to type "%" in the address bar
- space
- [your query and hit tab/enter to navigate results]
The last bullet is a close to ideal, native search among open tabs and make it so smooth to find an opened tab among dozens. But the key presses necessary to get there? Who can use that without weekly hospital stays for finger RSIs?
I love firefox. If someone, somewhere reads this, please please please think of simpler key presses to use this nice, already built functionality.
(I know non-native extensions provide similar feature. But native would be so cool and stable, especially that it's already built).
; Make alt-q in Firefox type ctrl-l followed by "% "
#IfWinActive ,ahk_class MozillaWindowClass
!q::Send, ^l`%{Space}
#IfWinActiveDoes anyone by chance knows how to remove it?
aka Menu > Options | Search | Search Suggestions > Provide Search Suggestions
a ??= 3;
(It would be even nicer if it could mean the same thing in PHP.)
The rest of the coins... Not really. Maybe privacy coins might be useful.
Every decentralized or censorship-resistant network gets dragged down to the worst of its adopters in the public consciousness (not entirely by accident IMO). It's sad, but should be acknowledged as par for the course at this point. If you're not going to fall into line, expect the power brokers to mix you up with the worst of 'em.
Jeez...
Im on Win10, this happenened on win8.1 aswell.
>Polymer.version
3.4.1
I also didn't know Chrome could emulate different levels of slow network.
Edit: trying again, the macOS track pad seems ok but scroll wheel behavior is different. Firefox requires 2-3 times the scrolling distance and transitions slowly to the final scroll destination. Chrome does not.
Edit 2: I’m almost positive it’s smooth scrolling. Some people hate it and some like it. This reddit thread sums up: https://www.reddit.com/r/firefox/comments/5zoa1t/do_you_use_...
Expected behavior is move the wheel at all and the page moves.
Also on a track pad the page should move in sync with your fingers. It doesn’t in Firefox.
It does just fine for me on GNU/Linux.
Considering that scrolling clearly isn't broken for Firefox users (I think Mozilla would notice), doesn't that point to the fact that it's probably something on your end and not Firefox?
options > browsing >
then uncheck smooth scrolling option
From what I can tell scrolling works fine and speed can be adjusted for personal needs in macOS.