Extensions in Firefox 58
blog.mozilla.org
blog.mozilla.org
For example writing some text to the clipboard seems to be a basic thing a web extension should be able to do. BUT to do it, you have to create a textbox, set its value to something, then programmatically select the text, and then invoke document.execCommand("Copy") to copy the text. Come on! Why not something like browser.clipboard.copy('some text')? And it does not even always work like expected³.
I hope the next ff versions will improve on this. Extensions are what makes the browser your own.
¹https://addons.mozilla.org/en-US/firefox/addon/pwgen-passwor...
²https://addons.mozilla.org/en-US/firefox/addon/pwgen-reloade...
Also NoScript reliably breaks autoplay videos, as I found out when the webextension version needed a few days until the release.
##.highlightMenu
.highlightMenu {
display:none;
}javascript:(function () { var pw = ''; var array = new Uint8Array(12); window.crypto.getRandomValues(array); for (var i = 0; i < 12; i++) { pw += ('0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ-_')[array[i]&63]; } prompt('Random password:', pw); })()
Web extensions are probably a small part of it but I have switched back to FF after ~5 years of Chrome. I also like the redesign but the killer feature was the noticeable speed up.
I consider myself a fan again!
This is the best Firefox , but I have lost a ton of extensions including my Vim extensions. So right now I use Qutebrowser and know that the new plugins will get added.
The other issues regarding limited control over the browsing experience may be resolved with future API expansion, but the previously mentioned hobbling is something with which we are most likely permanently stuck.
If Firefox provides the correct API's we might have the power of vimperiator back. Off the top of my head it needs:
1. A way to load config from the filesystem so I can keep its config with my other dotfiles
2. To be able to work on about:addons etc
3. To work in reader mode
4. Vim keybindings in textareas/inputs (although for modularity this could be a separate addon).
5. A way to open textareas etc in an editor - again for modularity this could be a separate addon.
6. Be able to open file://
7. A way to focus on the page content (away from e.g. url bar) - this could be a FF level hotkey though.
EDITED - I'm adding things as they come up.
But a few days before upgrading I found vim-vixen. It does 95% of the job. The performance improvements of FF 57 are well worth these 5% missing.
[0]: https://addons.mozilla.org/en-US/firefox/addon/vimium-ff/
I've been using Firefox since way back in the day when I stopped using Opera 12 (and I had been using it even before that, I just had really crap hardware for a long time so I had to use Opera).
The devs obviously care about the product and its users and sometimes have to make tough decisions.
I think Firefox 57 is a good thing for Firefox's future. It's basically what the Mozilla rewrite or Firefox's being spun off from Mozilla did, but this time with less of a big-bang. And 1 year from now 99% of the current extensions will work.
And no, for the obvious question, they couldn't have done it the other way. I've been part of several migrations of this kind and people don't really move voluntarily 100% during migrations. Yes, the nice folks do it in time (say, 40%) but the rest have to be dragged along kicking and screaming.
Every change breaks someones workflow [1]. It sucks, but it happens because we need to acknowledge a point when the current system holds back progress more than the headaches introduced by a new system.
[0]: https://blog.mozilla.org/addons/2015/08/21/the-future-of-dev...
No it hasn't: Firefox still doesn't ship an API capable of supporting many existing extensions. The developers of those extensions haven't had three years to port; they still can't port.
I really like how fast the new Firefox is, but I won't use it until Keysnail works.
JS-based (crippled IMHO) vertical tabs are also in Chrome. Why would someone switch from Chrome to Firefox to use a crippled extension?
I hope I am wrong but I don't think Firefox will be able to increase their market share. They may be able to stop losing their market share, which is not really a huge achievement.
And I doubt they're ever going to.
It's the Design Disease: Once a company has it in its head that it Knows Better, anyone who ditches their way for a different one is Wrong, just... Wrong in some way the company thinks is objective, based on whatever Design it has in its head, and must be brought back into the fold.
How tabs work is Design. Design must not be questioned. If you want your tabs to open such that you get a new tab right beside your current tab, instead of over a dozen tabs away at the end of the tab bar, you're Wrong, and having more than three or four tabs is Wrong, too.
They can say that Firefox is faster. I say it's easy to be faster when you're not solving the whole problem.
NoScript, Session manager and Self destructing cookies are the hurdles to clear for me - the rest I'll miss but can probably get by without. Note that both the session and cookie manager are in fact features that should already be part of Firefox and are not. And when I say 'clear', I mean functionally equivalent in all regards to pre 57 (NoScript fails that test).
Many people are unhappy with the extension situation. Mozilla have to take that into account and provide equivalent functionality fairly sharpish or face just being a chrome-clone.
By doing so, Mozilla shows that it cares about his user. Letting use old stuff that are not maintained anymore (if they are, a compatible version will shows up) lead you to situations like win32
By completely trashing a tool I'd spent years customizing without anything resembling feature parity ready by release? It's not really on the devs to port something when the new system is missing APIs. The "They had 5 years!" argument directed at the addon devs could just as easily be directed back at Mozilla...
Or perhaps by ignoring the loud negative feedback when I and many others said we didn't want paternalistic "thou shalt not install unless we say okay" controls on add-ons like Chrome has?
Or was it instead by ignoring my repeated questions on how to fix a ten-year-old SSL handling bug that causes a great deal of pain for sysadmins? (Vendors have a poor habit of reusing SSL certs - Firefox will flat out tell you to pound sand and not allow you to visit a site with a dupe cert. Chrome, IE, Opera, and I believe Edge will warn you that this is abnormal, but otherwise continue)
If this is how Mozilla cares for its users, I'd hate to see what they consider neglect.
It's not a zero-sum either: those of you who always used Firefox will surely benefit from it not dwindling into obscurity.
FF57 is absolutely the greatest version of Firefox to date. I have yet to hear a single legitimate complaint against it that actually holds up to scrutiny.
It breaks about 2/3 of the extensions I was using, and many of them don't currently have similar replacements available. That's a big loss in some of the functionality that made Firefox attractive as my default browser.
I understand the desire to fix fundamental architectural limitations. In the medium term some of those extensions will probably be updated or replaced. In the long term, the improvements may well pay off in terms of better security and better performance and easier development allowing faster progress. As a software developer I can see that the move was rational even if it is also somewhat controversial.
But in the short term, the loss is still significant for some users. That's a perfectly legitimate concern, and it's apparently sufficient that some people are not upgrading this time.
I tried to think forward, and proactively convert my greasemonkey scripts to new style FF extensions a few months back, to avoid dependency on GM. But abandoned that after it became clear that I'm not allowed to install my own extensions on regular Firefox, because of forced thir-party signing requirement. I have no need for signing. I could create an extension by zipping a directory. Now the workflow is 1000x more complex with all the crap loaded from npm required to sign it.
My web experience is s*it, ATM.
Firefox is great anyway. But it is power user hostile in some aspects too. Personal extensions/userscripts are central to my use of the web. So this is all quite annoying, since signing was enforced. And now even my userscripts broke with 57, as expected.
But Mozilla doesn‘t owe every single user his own build with their pet features.
> It's my computer, not Mozilla's.
Then you'll be overjoyed to hear about the unbranded builds! The exact same code, except it's yours, so it allows all extensions and doesn't say "Firefox" on it.
Sorry to hear this; but I think it's arguably more important for Mozilla to improve the web experience for a hundred million users who may never "write userscripts / extensions" for any website than to hamstring their development in order to avoid inconvenience to a single user who feels the need to customize every point of their web experience.
Also mandatory signing doesn't improve web experience for anyone. It's a security feature. Security is always inconvenient, almost by definition. So your point is invalid.
I don't mind change, I like new Firefox features, what I dislike is imposition of stupid lockdowns, and pointless control. One valid point, I might concede, is that there are innocent third parties affected if someone clueless confirms installation of some malevolent extension. So restricting it is somewhat justified.
Anyway, it's all still a sham. Anyone can still disable mandatory signing with a simple 10 line script patching omni.ja in any Firefox. So it's still no security against people who can be persuaded to enbale something in about:config, or run firefox with a command switch, or add some file to /etc/firefox/, or run a simple 10 line script "to make firfox compatible with our great extension".
To all those claiming that Mozilla doesn't owe this to their users, you are technically correct. But why piss off users when you can easily satisfy them with a simple option. I shouldn't have to use a patched browser for something so basic.
I think that the issue was that they were worried about people who don't understand the security implications turning verification off and getting themselves into trouble.
1. Video performance. Even something as simple as opening a video in reddit spikes my CPU to 100%, and before reluctantly installing Adblock Plus I ran into multiple kernel panics from opening articles on mainstream news sites and blogs.
2. Pinch-to-zoom isn't supported (closest alternative I could find is some about:config settings to make the pinch gesture equivalent to cmd++/-, which isn't useful).
I'm still using Firefox, and there's a lot that I like about it more than Chrome, but I'm seriously considering switching back because of those and some other minor issues.
Is it that easy to panic the macOS kernel?
At one point the whole browser crashed. So right now got Chrome and FF open and am switching depending on the site. Might just end up back in Chrome if this isn't resolved.
But then Mozilla management has been "weird" ever since they decided to chase Google's tail with rapid fire releases.
And frankly i fear that it, like some other big name FOSS projects, have attracted managerial types that are there more to pad their resume with "social" projects than actually caring for what they are dealing with.
And those in turn have introduced a "push to prod" culture of development...
SeaMonkey isn't a good role model here. Even before 57, the SeaMonkey release interval had lenghtened to multiple months. That is, evidently it hasn't been tracking Gecko security updates at Firefox's pace.
I understand getting salty about losing extensions, but sometimes a company has to make decisions that will get them back on track and recording some growth again. Sometimes a company has to overhaul their product to make it work for _more_ people and not their "minority but vocal core."
You run arguably the most "free" and "open" internet browser in the business. You are constantly pushing for standardization in the face of competing browsers like Safari, Chrome, and Edge; all of which are trying to establish a walled garden powered by browser specific features. Your browser is consistently losing market share to Chrome because you don't have the cash to spend on marketing partnerships and pre-installs like Google does. Your browser is a means to an end, not the end itself. Your browser exists purely to push forward the tenets of the open internet and open source software. You need market share to push further standardization and improvements to the web experience. To get this market share you need to ensure that your browser offers at least the same performance and security as other browsers. In order to do this, you need to make some backwards incompatible changes or risk falling into obscurity. It's a sacrific you have no choice but to make.
But they're clearly working on improving what you can do with extensions, only now they're doing it in a coherent manner rather than just letting people have access to the browser chrome to hack about whatever they fancied. And that feels much more stable and architecturally stable as something to build on for the future. I'm prepared to give them a few releases to build on that and make it both fast and extensible.
This should also enable them to keep the extensions that use supported apis working between releases much easier. I remember the early days of firefox where any update meant that some random subset of your extensions would break because they were doing something to the UI that the devs didn't expect.
Anyway, Mozilla's history is such that I am not convinced they will do this.
https://hackademix.net/2017/11/21/noscript-1011-quantum-powe...
edit: I see it's there. Released yesterday. Yay! Now if only Firefox could add a "Add to desktop" option, I could really switch!
No, it isn't. uMatrix is a replacement only for the shallowest feature of NoScript. There's a ton of stuff that classic NoScript does that nothing else provides (including the new WebExtensions NoScript, so far).
Agreed, I plan to remove this. See: https://github.com/gorhill/uMatrix/issues/771#issuecomment-2...
NoScript classic has a lot of functionality that helps you keep sites usable when scripts are blocked, or to help you stay secure when you have to allow some scripts. Some of the latter features are protections that will occasionally get in the way of sites that are fucked up in non-malicious ways, but that doesn't make them misfeatures. In my experience, the only XSS filter false positive has been Wolfram Alpha, and I've only encountered the clickjacking protection with advertisements that are getting in the way of the content I'm trying to interact with.
Sounds good, but have you looked at the code behind this? The XSS filter is the same way – a giant mess that only protects against the most trivial of XSS anyway.
I forgot about its clickjacking feature, though. You’re right, that one is valuable.
Can't do that with UMatrix.
FYI you can actually use both NoScript and UMatrix at the same time.
NoScript ("allow first-party scripts" and block everything in "Embeddings" tab) + FlashBlock seem to be the only way to block all autoplaying crap in Firefox for me.
You can also add uMatrix, uBlock & RequestPolicy and surprisingly they don't conflict with each other!
>I also like the redesign but the killer feature was the noticeable speed up.
FYI, it was faster when NoScript worked. Tabs didn't crash constantly either.
I do have some empathy for your parent comment: I'm not particularly fond of NoScript v10's UI either, if that's their chief complaint. It's different and somewhat cartoonish, but I don't know if my reaction is because of the stylistic changes or the fact that it just changed.
Regardless, Raymond Hill released a version of uMatrix for Firefox 57 which I feel has somewhat more power than NoScript in certain areas (selectively blocking cookies is a nice addition), and I like the UI better. It's not as intuitive, arguably, as the original NoScript, but it's information-dense and provides a fantastic picture of what's going on--better than NoScript ever did.
NoScript 10 caught me on a good week, if I had been busier this week I wouldn't have had the patience either because it's definitely not doing things in a way I would expect.
Can I ask do you run it with ublock or other ad blockers?
So here are my notes for the previous version, I literally went to type this up only to find NoScript updated itself in the past day or two.
You have Default, Trusted, Untrusted and Custom.
Double-Clicking on either of those will give you more granular control ranging from scripts to media to fonts and webgl, which is nice. Untrusted has everything unchecked by default, Default seems to have scripting and most other things turned off by default, and Trusted and Custom you have to set yourself, with Trusted I think having Scripting on by default.
Now if you change the settings for any of these, at least on the previous version, your changes will apply to all domains under that category, so if you have scripting on for Custom, you'll have Scripting on for all domains you've set to Custom. Now whether this is still the case for the latest version, I couldn't tell you at this exact moment, but hopefully that's enough to get you started.
"Temporarily allow" also seems to have been replaced with a clock icon which you can find next to the Trusted and Custom category markers, which I only knew about from reading the blog updates. The latest update added three icons to the top right which are fairly self explanatory, one is options, one is for temporarily allowing all of the page and one is for revoking temporary permissions.
All in all I wouldn't say it is the most intuitive UI, particularly after having used NoScript for about 10 or 11 years. That said, if you're willing to allow yourself a bit of time to get used to it, most of the core functionality is retained, and is in fact a little bit more granular than before. When I have a little bit more free time, I intend to test out Custom a little bit and see if I can use it to allow some websites to render custom typography without having to run JavaScript. Theoretically that should be done completely in CSS, but I haven't kept on top of web standards since deciding not to be a web developer about 10 years ago so I'm not sure to what extent websites still depend on JavaScript for typography.
There's talk that some API's will be built back, and hopefully with the increase in new users it should be worth it.
For example, on Xubuntu, a low resource friendly Linux distribution run by many on devices with small screens, Firefox 57 now is unable to hide the title window bar and unable to customise the size of tabs, reducing available screen estate by about an inch. In development versions of Firefox, they are enabling CSD for Firefox windows, so in a few months this might be better for the title bar, and in the article it seems as if theming tabs are being addressed.
A particular bug for me is the second time they removed being able to change to the next tab by mouse wheeling over the tab bar. They removed it in Firefox 34 or so (it was default, in built function) and pointed to using Extensions as the workaround. Now they have removed these very extensions that enabled that.
If only an extension could change userChrome.css! That would solve many people's problems.
This! Even if extensions were very restricted to only toggle pre-defined css rules or something, this could go a long way.
There are hundreds of people tweeting about how Firefox 57 has ruined their browser experience, and they are representative Firefox users -- the kind who have been using it all along, and relying on all the software Mozilla just broke. I don't think Firefox will gain market share by alienating all these people, while attempting to beat Chrome at its own game.
I wonder if there is any set of circumstances that would cause Mozilla to realize that making a new browser and calling it Firefox was a bad idea. If market share still continues to tank -- like were they realize for every person like the parent commenters, there may be many pre-existing users who have no reason to use Firefox anymore -- would they then admit it was a mistake? My guess is there is no such set of circumstances... witness the fact that Mozilla originally cited declining market share as a motivating factor for this big change, but then recently we have stories like this https://www.theregister.co.uk/2017/07/25/new_war_for_mozilla...
While they could have split the product into two (like they did with Firefox & Seamonkey) and let the former product languish, I'm not that would have made many users much happier. For what it's worth, you can still use Firefox 56 (or even Seamonkey).
"If true, the resistFingerprinting preference makes the browser report generic spoofed information for data that's commonly used for fingerprinting. Such data includes the number of CPU cores, precision of JavaScript timers, and the local timezone. It will also disable features that are used in fingerprinting, such as GamePad support, and the WebSpeech and Navigator APIs."
And well, apparently it has a pretty bad bug. It spoofs my screen size from previously 1366x768, which 1 in 9.43 browsers have, to 1318x686, which by itself is already unique, because that really is just not a screen size that anyone sells.
So, that's a thing. Maybe it's just isolated to Nightly, but if in doubt, maybe also just test yourself. Now where to report this...
Is it sticky though? I don't know how the spoofing works, but if the next time, the same site gets a different size, wouldn't your profile become less identifiable across visits and other web sites.
No, it changes every time...
For security reasons, yes.
For people who don't want to say bye to old extensions which are not (or cannot) be ported to 57+ there is the option to stick with 52 ESR version (Extended Support Release). According to Mozilla FAQ [1] this version gets obselete in march 8 by Firefox 59 ESR.
[1] https://www.mozilla.org/en-US/firefox/organizations/faq/
https://addons.mozilla.org/en-GB/firefox/addon/tree-style-ta... and look through teh list til you find something that supports FF52+
I'm a little concerned that any bugs/holes in extensions will be left open though, since it's very unlikely the extension devs are going to be backporting features/fixes.
As for me, I'm just happy to be running Firefox as a daily driver again.
Personally I'd be happy to see Firefox have tab feature chrome natively has:
- Scroll wheel to move between tabs (https://bugzilla.mozilla.org/show_bug.cgi?id=1285812)
- Ctrl+1 (2, 3, ...) to jump to a tab. Ctrl+0 to jump to the last tab. [Edit: This works with Alt, nvm!]
- Multi-select tab management (Ctrl+Click, Shift+Click to select, drag, snap and move multiple tabs)
Back when Firefox initially got a lot of market share, its killer feature was tabs. It almost feels like they forgot that :(
- Multiple-row tab bar.
Such an obvious improvement over tabs scrolling off the screen sides or squashing them like Chrome does, I don't know why it's not being considered.
It's already been done by Tab Mix Plus (or was it Classic Theme Restorer?). They only need to copy it and polish it a little.
Personally I don't find tree-style tabs appealing at all because (1) I usually have few pages open on the same domain, but I have many pages open on different domains. And (2) I don't want a vertical column of content width lost to a sidebar.
Limiting tabs to a single row per window is awful for usability.
You have to understand that this is a pretty unusual concern. Most of the world struggles with screens with aspect ratios optimized for watching TV, making vertical space very precious but leaving plenty of excess horizontal space that cannot be effectively used by the majority of web pages that only use a single column of text.
You are right about that. I use two external monitors, one of which is rotated 90°. Guess which one I keep my browser window on? :)
Ctrl+1 (2, 3, ...) to jump to a tab.
works for me, Ctrl+0 doesn't
You could try the Multiple Tab Handler[1] extension. It seems a bit clunky and limited if you use it alone, but it integrates well with Tree Style Tabs [2]. Note that you could open the vertical tabs sidebar when you need to manage tabs and close it right after. You can also disable the "tree" feature of tree style tabs.
[1] https://addons.mozilla.org/fr/firefox/addon/multiple-tab-han...
[2] https://addons.mozilla.org/fr/firefox/addon/tree-style-tab/
For users that are okay with the common way browsers work, all this new stuff is great though. Firefox Quantum is so fast and it still delivers all of the mainstream add-on functionality.
I'm firmly in the second camp. Give me any browser that I can install uBlock in (and hide most buttons except back and forward), and I'm good to go. With Quantum, I switched over all my browsing to Firefox. Including on Android, which hasn't even had most the Quantum improvements yet (its UI has just gotten a much needed update).
It takes 12s after launch for my newtab to populate with icons -- which incidentally I dislike compared to the previous page thumbnails -- there's regular stalling when it fails to respond.
I'm prepared to give some space to get acclimated to new versions, there were font problems on lots of pages (including newtab) for me which I've worked through. But I'm failing to see any benefit yet - even then turning the icon super bright has annoyed me ...
Obviously there's something slowing down your Firefox, but it's definitely not to do with how modern your hardware is because 57 is noticeably faster than the previous version on my 2011 MacBook.
browser.newtabpage.activity-stream.enabled = false
browser.newtabpage.activity-stream.feeds.section.topstories = false
browser.newtabpage.enhanced = false
browser.newtabpage.enabled = true
seems to respect my 5 columns and 3 rows with large preview icons.
And on the "improvements" side -- I have to say I've never felt the browser was too slow. Sometimes the network is lousy, sometimes stupid websites are terrible, but the browser itself has been just fine for me for years. So I'd rather have had a less painful and abrupt transition, which maybe took a bit longer.
With this change, Firefox targets the lowest common denominator just like every other major browser (tracing the footsteps of Chrome). In the progress, it too, is becoming just like Chrome.
This whole saga begun back when Opera dropped support for their 12.x browser, and became just another Blink-based browser. At that point, the only way to get a usable browser (with vertical tabs, MRU tab switching, gestures) was Firefox with a number of addons (Tree Style Tab, Tab Mix Plus, Firegestures).
One may argue that users like this should just use a fringe browser with such features built-in, but that's not viable because they're either unstable or lack timely security updates.
You may also argue that you can get all those features, even in Firefox 57, but that's only true with caveats:
Tree Style Tab still exists, but if you use either the history or bookmark sidebars, you better get ready to switch back and forth all the time (since only one sidebar can be active at a time).
The about:config setting browser.ctrlTab.previews (now also exposed in preferences as "Ctrl+Tab cycles through tabs in recently used order") is mostly useless, since it is limited to the 6 most recent tabs (probably even fewer on lower resolution screens).
There are several mouse gesture addons for Firefox 57, but none of them currently work on Linux/Mac. Even on Windows (and the others when they're fixed), gestures will work neither on internal pages nor on addons.mozilla.org.
It's a shame tradeoffs are necessary, but I think it's worth highlighting that it's a tradeoff that also brings substantial advantages. For all users, including the power users.
It's likely available from your package manager.
Things that bother me on desktop developer edition:
* no usable mouse gestures extension
* constantly changing UI
* hid really useful things like file/edit/history/bookmarks/tools menu behind
holding alt button on Ubuntu.
Pretty sure that whole menu will be gone soon.
* it's impossible to clear for example last week, last month etc of browsing data (not sure if you could do it before).
In fact, I'm not sure if I can clear cookies but not site history anymore after spending 5 minutes in the settings.
* tree-style tabs turned into something unusable
* getting rid of all the rare extensions that worked previously and are no longer supported by the original author
* tons of greasemonkey userscripts broken
* greasemonkey a lot less usable now (still can't figure out how to write a custom script from scratch)
Unfortunately, on desktop I'm stuck between this new FFX and Vivaldi. On mobile, I already switched to Brave even though it's terrible because you can't easily google for selected text or select a part of a paragraph easily.
But couple of your annoyances can be fixed rather easily:
1. Get back the toolbar back: View (or Alt-V) - Toolbars - Menu Bar
2. To clean recent history: Ctrl-Shift-Del will give you the clean recent history dialog box. It can be accessed from toolbar too: History - Clear recent history.
Weird that the only way to get to that menu is through a toolbar that's not visible by default.
What's unusable about the new tree style tabs? I use it constantly.
If you want old greasemonkey scripts to work without changes, use violentmonkey.
There must be a bug with the new greasemonkey that I can't find the editor either. Use violentmonkey for that too for now.
I'll grant that the new responsiveness of FF makes TST feel much better. But there's one thing that's constantly giving me headaches. I don't feel like the highlighting given to the currently-selected tab is in-your-face enough, so that I'm always scanning up and down the tree trying to see which one it is. The other formatting options it used have were nice, but being able to at least make the selected tab's title bold was a big deal for me.
I realize that this is technically possible in the new regime via userChrome.css. However, the element classes are completely undocumented so I have no idea how to accomplish it.
Before FF57, I was able to set the font and background color for tabs depending on their status (current, inactive, unread, etc.), and that's what I want now. IIRC, this was actually a feature of Tab Mix Plus.
It's clear that this is still possible even without TMP, but requires putting code in userChrome.css. That would be fine with me, if I could just find any documentation on what element classes (or whatever would be the most appropriate selectors) I need to reference.
document.querySelector('#sidebar')._contentWindow[0].document.querySelectorAll('.tab.active')[1] https://github.com/piroor/treestyletab/wiki/Code-snippets-fo...
With tree style tabs I didn't realize there is a CSS hack to remove the top tabs.
Ditto re: weird css hacks to hide the top bar.
Why doesn't it get prime support?
Iirc, they decided to move it out of the browser and in to an add-on, which is frustrating in light of the recent changes. Don't get me wrong, I'm for the migration to WebExtensions, but it's annoying that there wasn't a better path for a feature that used to be a part of the browser.
Here's hoping most of this will be resolved in the upcoming months.
Firefox plans to add an API for dynamically hiding the tabbar. You can follow progress here: https://bugzilla.mozilla.org/show_bug.cgi?id=1332447
Blindly copying Chrome is the wrong path to head down. You end up playing into Google's game of building systems with user hostile misfeatures meant to protect ad revenue.
Edit: better link https://wiki.mozilla.org/WebExtensions/TabHiding
> Some of the key goals of WebExtensions are that it’s not about changing the Firefox UI in the way we used to in legacy add-ons. There’s clear limitations we need to set. However there’s so much interest around tabs and tab groups.
So tab mix plus gets its APIs because people yelled hard enough, but vimperator won't?
https://github.com/vimperator/vimperator-labs/issues/705#iss... "The WebExtension APIs offered by Firefox already contain many features that chrome lacks and I am working constructively with them to add the APIs we need to make something a lot like vimperator in Firefox."
Unfortunately, that's the direction Mozilla is going with web extensions. It must because Mozilla's complicity to bake DRM into the browser requires giving control of the user's computer over to media interests. It requires favoring corporate interests over user freedom.
The horrible thing about Quantum is it replaces the last browser that put power in the hands of users. It removes the last browser that was not constrained explicitly for media consumption.
The sandbox model necessary for DRM is what limits the power of web extensions. DRM priority is why the sandbox model cannot be turned off by the user. Making Netflix in Chrome low risk for Netflix is why the user is no longer allowed to control their browser.
The sales pitch for Quantum is that it is faster for media consumption. The developer pitch is that it is easier to write extensions for consuming content.
Quantum reflects a decision to remove user controlled functionality in support of general purpose computing and turn Firefox into a media consumption appliance controlled by Mozilla and fungible with Chrome.
Virtually nobody who works in DRM would agree that the add-ons model has anything to do with the effectiveness of DRM. If you have access to the user account running Firefox, running arbitrary unsigned native code in the Firefox process is trivial.
> Another is that Firefox no longer supports general purpose computing.
XUL add-ons were JavaScript. Web Extensions are JavaScript. You can write whatever Web Extension you want for whatever purpose you want.
content scripts have some limitations. They cannot:
+ Use chrome.* APIs, with the exception of:
extension ( getURL , inIncognitoContext , lastError , onRequest , sendRequest )
i18n
runtime ( connect , getManifest , getURL , id , onConnect , onMessage , sendMessage )
storage
+ Use variables or functions defined by their extension's pages
+ Use variables or functions defined by web pages or by other content scripts
>+ Use variables or functions defined by their extension's pages
>+ Use variables or functions defined by web pages or by other content scripts
Those are just consequences of the content script not running in the same javascript context. If it ran in the same context, then the web page javascript could access private extension APIs or otherwise break content script code. If you actually want to interact with the page's variables, there's ways to do that.
And the Firefox extension model does not implement Chrome's restrictions. It has APIs that Chrome does not support: https://developer.mozilla.org/en-US/Add-ons/WebExtensions/Br...
Now I'm migrating to using Qutebrowser, as it's the closest to providing a similar experience.
[1] - https://intoli.com/blog/firefox-extensions-with-selenium/
For those interested, the whole discussion about adding it is here [2].
[0] https://github.com/cmcaine/keyboard-api [1] https://github.com/Koushien/keyboard-shortcut-api [2] https://bugzilla.mozilla.org/show_bug.cgi?id=1215061
[0]: https://addons.mozilla.org/en-US/firefox/addon/vimium-ff/
.. works just fine ...
I would have preferred to see this land 6 to 12 weeks after CSP handling originally landed in Firefox, but landing it now is a clear case of "better late than never".
He was the only bigger extension dev that threw a complete tantrum over this decision, refusing to even try to work with Mozilla to get some APIs into place, so that he could maybe actually port his extension in full. And now he's doing a port anyways.
Well, good on him. I do not think that his behaviour back then was entirely justified, so this remedies that.
Another use case coming to mind is to compensate for the horrible bad ability to define your own searchengine. Leting a extension handle this would be an improvement.
And while it's becoming easier to manipulate any website you visit, with basically nothing stopping a malicious extension from stealing user data or just taking over any online account, actually extending the browser UI, the original goal of many former extensions, is extremely limited.
So it's like before, adding a new extension is always a security risk.
But It would be nice to have an extension to turn off resistfingerprinting on some sites (like firefox's extension site) because it fakes the browser's user agent which breaks things that check your user agent. I'm not sure this will allow that. It seems like it'll allow a privacy extension to easily enable / disable it for all sites with a check box.
Giving full read/write access to an extension seems VERY dangerous to me, especially if the extension was compromised unknowingly by someone who hacked into the developers machine prior to a release.
We've started seeing a lot of alternative Blink UIs (Brave, Vivaldi, new Opera, etc) alongside apps built on it via Electron. Those alternative UIs do things Chrome extensions can't, even including things Firefox 57 lost the ability to do.
I don't know how close to that state Gecko is right now, so maybe it would be less work just to recreate the necessary APIs like they're already doing. But if not it could help bring back some of Gecko's market share, which is at least as important as that of Firefox as a whole.
The alternative "Blink" UIs, and Electron, are really just forks of the entire Chromium browser, soup to nuts.
They've become such a fundamental element of modern web apps- that so many of them just fall over and break on Firefox is alarming.
This is nice, but I'd prefer if support for FIDO U2F would be here already. But it looks like something is happening with that too:
https://www.yubico.com/2017/09/firefox-nightly-enables-suppo...
> "This preference makes the browser associate all data (including cookies, HSTS data, cached images, and more) for any third party domains with the domain in the address bar."
They even included a link you can click on to get more info on it.
This means that if you visit example.com and then initech.com and both load a tracker from evilshit.org, then the tracker will see too completely fresh browser installs, not even HSTS will be setup if it was used.
Basically, every domain you type into the URL address bar behaves as if it was a unique container.
> firstPartyIsolate – This preference makes the browser associate all data (including cookies, HSTS data, cached images, and more) for any third party domains with the domain in the address bar.
I wonder if it will be possible to load scripts into the reader view document. Currently, it isn’t, so keybinding extensions, which rely on inserting Keyboard event listeners, don’t work in reader view. I had been prototyping a custom reader view in Saka Key using Mozilla’s Readability library to work around this issue.
The Content Security Policy fix is a lifesaver. Saka Key attaches hint elements to the DOM and attaches a script tag to load common styles/fonts. Sites with the policy style-src: self (like crates.io) block script tags, which breaks hint styling and positioning.
Is this an extensions API limitation or is not using the keyboard part of FF norms/culture? Even much built-in functionality (like save to Pocket) is mouse-only.
Glad they got rid of the old extension system now that the new one is so feature complete. I mean who needs spellchecker API anyway, if you miss spellchecking in your language, you can always start writing in another language, no? Perhaps they could introduce a Pocket API to compensate?
- Workspace support in DevTools (i.e. unify localhost-loaded resources with an on-disk workspace to enable updating JS & CSS without reloading, editing and saving files from DevTools, etc.)
- queryObjects console API method (from Chrome)
- alternate DevTools themes
If/when those land I’ll likely go back to using FireFox as my primary development browser (been developing on Chrome since 2009).
I dislike all the window-based bookmarks, downloads, etc UI ("Library"). You have a world-class extensible customizable graphical user interface system right here in the middle of the browser, use it.
If I care about speed, I use Chrome. If I want custombility, I used to pick Firefox. Right now, it's basically a Chrome clone, so why bother.
And the privacy argument is not strong enough.
I suddenly lost so much with Quantum, I wish they had held onto killing the old add-ons until the WebExtensions API was complete.
It landed in Nightly as a WIP feature, but it's getting there: https://bugzilla.mozilla.org/show_bug.cgi?id=1283299
(Btw, with title bars hidden Firefox and Chrome actually have pixel-for-pixel same size nagivation elements, at least with Ubuntu's defaults – of course only when you toggle on system title bars in Chrome.)
Mozilla deliberately removed a lot of customizability, but instead of being open about it they made addon developers the scapegoat, sending unhappy users their way while blocking requests for replacement APIs with "you are not supposed to do that kind of UI tweaks anymore". After a decade or so of encouraging UI extensions whenever someone suggested advanced UI features, this is a bit lame.
That said, there's also a lot of customisability that simply hasn't been implemented yet. I'd name e.g. hiding of the tab bar where Mozilla is clearly seeking a compromise between the potential downsides of removal of UI elements by add-ons and how useful it could be. It just takes time to make it happen, but luckily, Firefox 52 is still an Extended Support Release for quite a while.
Anyone have info about if/when it will support Quantum?
Makes it easier for me to manage my extension https://weatherextension.com
1) Someone here might know more (updated) about when's the support going to be added for the same,
2) The Extension Author might comment on when he's going to add the support for it if possible,
3) Some of the geeks here might know a good replacement of it (because Firefox's "Find a Replacement" is not as good as it's supposed to be).
4) When people say "me too", some other Hackers around here might be inspired to write one from scratch, or Firefox people might see how important it is (I know about Bugzilla's voting system, trust me, that doesn't work)
Firefox has just broken people's workflows that were set in place for years, and they are naturally concerned. Let it be.
I don't think that is highlighted enough - the majority of the discussion is focussed on cosmetics, but it takes a lot of effort to break habits and muscle memory that have developed with daily use over many, many years (since Phoenix [1] in my case.)
It's also going to take a huge amount of time and effort to research, learn and configure replacement extensions. I really don't need that stress for something as important as a browser, so I'll stick with ESR as my daily tools as long as possible and build a new profile e.g. using Nightly as time allows. It's far from a trivial exercise.
[1] https://blog.mozilla.org/community/2013/05/13/milestone-phoe...
The XUL deprecation in FF57 has been public knowledge for years. The extension authors are to blame here.
How do you achieve 1:1 compatibility between extension APIs when the architecture that enabled XUL to expose the entire codebase for extensions' use no longer exists, and no longer exists because replacing it was the only way to resolve showstopping performance issues?
How do you keep using a browser that no longer supports extensions without which you can't satisfy the use cases for which you chose and adapted that browser in the first place?
No one is to blame here.
In the end, XUL was a major performance and security issue. Good riddance.
Really? What opportunities have you observed in the recent close collaboration between Mozilla and the developers or maintainers of many high-profile, and not a few niche, legacy extensions? I'm sure I am not the only one here who would be interested in gaining the benefit of knowledgeable and thoughtful feedback on how that process has played out.
Look, I get it, okay? It's a pisser. But throwing this kind of shit around helps nothing and nobody. It just makes life a little worse for everyone in the blast radius. Maybe you're fine with that. More likely you're just mad. I get that. But it's not an excuse.
I get that you're mad about some of your favorite niche extensions being unavailable due to relying on XUL functionality that shouldn't have been there in the first place, but you'll just have to find alternatives.
The alternative a lot of people will find is a different browser.
Deleted comment
One plugin that I really miss in firefox is holmes[0]. I bookmark alot of useful sites and it can be a pain to find them in the bookmark list. It's a plugin that makes me really miss chrome.
[0] https://chrome.google.com/webstore/detail/holmes/gokficnebmo...
Otherwise, I'm using Safari and weaning myself off of all the power of VIM like browsing. Safari also happens to be so much smoother and does not kill the battery.
Never had any use for extensions. For me Web browsing is all about HTML/CSS and some occasional JavaScript, everything else is better achieved via native apps.
I never understood why people have such a hard time with choice. And why saying something no longer meets your needs evokes such reactions.
It was necessary the Internet and explosion of security exploits to realize that although plugins in-process are easier to implement, the older out-of-process architecture was safer and harder to exploit.
And yet they broke things like NoScript, which provided security/privacy features missing from plain Firefox and unavailable on any other browser no matter what extensions you installed. If safety was really such a high priority, they would have waited until after a full-featured NoScript WebExtension was possible to implement before deliberately breaking XUL extensions.
They're mostly a way to use the regular browser functions, only with the keyboard rather than the mouse. But this works best for regular sites, and poorly for webapps that use JS-activated links and such.
The solution is for developers to spend a day learning about accessibility. A good starting point is https://webaim.org/intro/.
[0]: https://addons.mozilla.org/en-US/firefox/addon/vimium-ff/
The biggest problem I have with them right now is that FF entirely prevents addons from working in "system" pages like the settings pages or the new tab page (NB this includes some actual web pages that Mozilla has "blessed"). Another problem is that there is no way to focus the address bar programmatically, forcing addons to use clunky workarounds like the Vimium omnibar, which has an inconsistent UX compared to the rest of the system.
Hotkeys, especially in a VIM-oriented model, must work consistently across the entire experience, and Mozilla has decided to explicitly disallow this for addons.
Vimium is better than nothing, but due to limitations of the WebExtension system, is strictly incapable of providing the browsing experience that I'm used to. In its current state, it is not a suitable replacement for me.
The same is true for the various mouse-gestures addons.