What's missing ?
What's missing ?
1. Since Webextensions are new, some extensions that I use in chrome, need to be ported over. Most are not must haves to be fair.
2. Hangouts for work. :(
3. A solution to manage different personas - Chrome has the "user profile" thing linked to google accounts, firefox as container tabs. I think user profile is better than container tabs, although both are kinda bad.
Except hangouts, none are deal breakers - I use firefox as much as I can, and in fact overall it feels more faster and more repsonsive than Chrome starting with Quantum. It also looks and feels better than chrome.
1. Are there any specific extensions you need? Happy to add them to our outreach list.
2. Sorry about Hangouts; follows a trend of Chrome-only product launches/re-launches from Google (five this year so far). Wish we could do more, but if Google's OK shipping things that only work on one browser, that's their prerogative.
3. Firefox also has completely separate profiles, in addition to container tabs. The UI isn't as polished as in Chrome, but it works: https://support.mozilla.org/en-US/kb/profile-manager-create-...
Keysnail. It requires the ability to override C-n, though.
hey callahad, thanks for shiming in. My biggest pain with profiles is that you cannot use separate profiles at the same time. A Firefox instance can only run one at the same time, making it useless to run e.g. a work and a private profile together. Container tabs also do not help here because they share the same history which I do not want for privacy reasons.
Nonetheless, thanks for a great release!
I may be misunderstanding what you mean, but I believe you actually can do that. If you navigate to about:profiles, there should be a "launch profile in new browser" button that runs multiple profiles at the same time, each in their own, isolated Firefox process.
To make this a bit easier, you can create desktop shortcuts which launch a specific profile like `firefox --new-instance --profile Work` (on Windows, use --no-remote instead of --new-instance).
LastPass please. I'm sure it's on your list already, and I'm sure they know about it too but every little helps.
I am happy to use pass/keepass (whatever version) but the family says only LastPass is acceptable.
for me, vimperator/pentadactyl and tab groups. i can't use a browser without them so i won't be upgrading :(
Chrome allows users to simultaneously run multiple profile instances, each with it own profile. Firefox didn't seem to support that.
https://addons.mozilla.org/en-US/firefox/addon/multi-account... (or a similar addon) might bring me back to using Firefox.
With Chrome, one set up a profile shortcut to a given profile, and then all tabs/windows within the instance corresponded to the same profile. This meant hitting Ctrl+T to open new tabs, rather than having to click and dig through a menu to pick the right profile.
I'd like the containers addon to support easy shortcuts. Maybe on hitting ctrl+t, one can then press 1-5 (or some other selector) to determine which profile the new tab opens in. Alternatively, more safely and less confusingly, just having windows sticky to a specific profile, so all new tabs predictably open into the same profile.
The UX around multi-account containers is still being worked on, but the contextualIdentities WebExtension API (https://developer.mozilla.org/en-US/Add-ons/WebExtensions/AP...) allows add-ons to interact with containers, so a third party developer could create exactly the experience you've requested.
about:profiles is not discoverable (obviously), easily usable (you have to restart), or easily switchable (you have to open about:profiles or the terminal).
By the technical definition of "UI", the raw filesystem and terminal counts, sure. Opaque and unknown about: pages count, sure. But that doesn't make it good.
For features like this the UI is the feature. about:profiles is not a good UI, so the feature is 80% useless.
I'm generally optimistic about Rust code having fewer exploitable defects, although it's a reasonable argument to suggest that the previous C++ parts had a lot more public scrutiny.
Also, for me personally: I use a lot of different profiles and using multiple Firefox profiles simultaneously is a mess. The only UX that still works is about:profiles (all of the profile switcher extensions broke since the move to WebExtensions) and I've had Firefox beachball (hang -- I'm on macOS) on me twice while doing that.
Don't get me wrong! I love Firefox and I'm really excited about this change.
Not sure if it meets your use case, but checkout Firefox containers.
Containers only separate Cookies and login data, to make you look towards the internet like multiple people.
Profiles on the other hand separate everything, as if you had two completely independent Firefox installations. Different browsing history, bookmarks, themes, extensions.
They work very well for allowing multiple people to all have their own Firefox, even though it's the same installation. But also really well for single-person use, for example I have a normal Profile for everyday browsing and then a webdev Profile, where I don't put in content/tracking blockers.
You can manage Profiles by typing "about:profiles" into the URL bar or with this: https://support.mozilla.org/en-US/kb/profile-manager-create-...
I'd love to talk to you more about the ergonomics of containers / profiles for your use case. You know how to reach me. :)
Thanks for that, I can now replace `chromium --temp-profile`.
On my 2017 iMac (3.4 GHz Core i5, Radeon Pro 570) I get 60 on Mozilla's Speedometer 2.0 [1] test with Firefox 57.0. I get 90 on Chrome and 93 on Safari.
On my Surface Pro 4 that gap is smaller, but Chrome is still about 15% faster on that test.
In the case of Speedometer, it's measuring one specific type of webapp interaction in a single tab; it's not capturing things like perceived speed of opening or closing the browser, switching tabs, loading pages, etc. It doesn't test how the browser performs when many JS-heavy tabs are loaded in the background. It also has lots of corner cases: installing a common password manager halves the score, and a common adblocker nearly does the same again, yet neither show that kind of noticeable degradation in real-world browsing.
Benchmarks are useful for saying "we've improved our browser this much in this area compared to where we were last year," but they're much less useful for making broad generalizations across heterogeneous browsers; there are too many other variables at play.
Multiple profiles within a single window, visually distinguished by tab highlight color.
Create a new profile by going to about:profiles, and thereafter when you start Firefox it should present you with an option to pick which profile to use.
More options: https://developer.mozilla.org/en-US/docs/Mozilla/Command_Lin...
c:\apps\firefox\firefox.exe -no-remote -profile "C:\profiles\some-clean-profile"
was my was-to-go method for testing various configurations.Knowing about
%USERPROFILE%\AppData\Roaming\Mozilla\Firefox\profiles.ini
is also often helpful, eg. when you want to move certain profile folder.While pieces of Firefox are moving to Rust the majority of attack surface is still in C++, so I want a strong sandbox if I'm going to run that browser.
It's described in detail here:
* https://wiki.mozilla.org/Security/Sandbox
* http://www.morbo.org/2017/11/linux-sandboxing-improvements-i...
Smart Zoom and Handoff is completely missing and Applescript scripting support is extremely poor.
To my knowledge, the only bigger architectural difference now is that Chrome sandboxes each tab individually, unless they're from the same domain as another tab, whereas Firefox by default only uses as many processes as you have CPU cores and then round-robins your tabs among those.
This is for performance reasons: You still get almost the same parallelism, but much less RAM usage especially with lots of tabs, and then that free RAM can again be used to speed things up elsewhere.
You can however configure this. Set dom.ipc.processCount in about:config to as many processes as you want to maximally use. If you set it to a really high number, like 1000, then it will create a new process for each tab (until you open your 1001st tab) and those will then all be individually sandboxed.
Another big security difference were extensions. Firefox previously with the classic Linux approach of "If it's installed, it's trusted".
Now they have the same extension model as Chrome with a number of additional APIs. Those APIs are potential security holes, but also allow the implementation of better security, for example NoScript is being ported to Firefox 57, still can't be ported to Chrome.
Well, and then going away from the model itself, Mozilla enables much better security on the extension side, again with the Linux approach, by reviewing extensions and extension updates before they allow the publication. I'd have no concern with my mum installing whatever extension she wants in her Firefox. With Chrome, I would be concerned.
Lastly, only some of us are security super humans. Chrome always gives an option for security to calm security people, but if it benefits Google, they'll leave this option off by default, even if it should clearly be default on.
For example Sync in Firefox is end-to-end-encrypted by default. Only one password needed. In Chrome, it's not. Because Google wants to evaluate your browsing history, bookmarks etc. You can have E2EE, but you need to separately enable it and use a second password for it.
Again, fine for security super humans, but the fewest of us understand browsers enough to know about all potential security holes and we simply do need to rely on browser vendors for our own security.
Don't think Google has actually fixed that Auto-Fill phishing attack either yet, which comes up around here every half year or so.
There are also many features in Chrome that I’ve come to appreciate, such as right clicking a tab and selecting “duplicate”, which opens a copy of the tab and retains the browser history of the original one.
These kind of hidden features don't require a whole lot of designing and forethought, so it's something that usually only really needs one dev thinking that it'd be nifty to have them implement it.
FF 57 also has "duplicate tab" with duplicate history.
Faster than old fox? Yes. Faster than Chrome? Nope. Still long way to go.
They should probably ditch all in-memory stuff, and just use the drive, so they can support those cases when you really want to have a "low millions of tabs open."
>Chrome's tab UI makes more tabs completely unusable
True, thankfully we have extensions that deal with this (I've been using Keepin' Tabs for this, until I started using cVim)
When I have 20 open I already start to get confused about where stuff is...
* Firefox's Awesomebar (the address bar) is much more powerful. Whereas Chromes sends you to google mainly: https://support.mozilla.org/en-US/kb/awesome-bar-search-fire...
* Tree Style Tab: https://addons.mozilla.org/de/firefox/addon/tree-style-tab/
Thanks to Firefox's userChrome.css, you can pull pretty nice customized UIs even without any additional extensions :) https://github.com/eoger/tabcenter-redux/wiki/Custom-CSS-Twe...
I have 9 virtual desktops; I typically have browsers in 6-7 of them for various in-progress tasks. That's somewhere between 6 and 14 browser windows (I can fit two side by side on my screen), with multiple tabs in each.
Finding things is easy. Even ignoring the per-task categorization by desktop, Firefox lets you search all your tab titles/urls: just type "% " in the URL bar followed by your search string. If I'm looking for something specific, this is the simplest way to find it.