I used VimFX before and I'm using Vimium now. It's okay, but it works nowhere near as well as VimFX did. The biggest problem is that it doesn't work until a page has loaded. So if I open a page and while it's loading I want to open a new tab, I can't use Vimium shortcuts to do it.
Want a bind to open something in mpv/youtube-dl
`";v": "hint -W exclaim_quiet mpv"`
Want to open a text field in gvim? Hit `ctl+i`
Want to remove the extra chrome: run `:guiset gui none`
I stuck on the ESR of firefox for a long time because I wasn't willing to give up vimperator. Its not the same, but its pretty close. There are 2 main pain points, which I don't know it will be possible to address. First the page has to before its functional. Second currently domain keybinds take priority (looking at you github).
WOW, thanks for this! This extension is awesome. This provides all the functionalities I use vimium (in chrome) for, but what does it for me is this:
> Want to remove the extra chrome: run `:guiset gui none`
Top bar with so much padding is one of the main reasons I am not using Firefox.
Also the awesome command bar at the bottom is something that I've always found missing in vimium, (which has an omnibar but it's placing feels awkward.)
You know how mathematicians get excited by exp(-i*pi)+1=0? This is like that but for all of the things I'm embarrassed about in Tridactyl:
- our documentation is not great, particularly of current configuration/possible configurations, hence why the GP didn't know about it
- hints don't support pipes - obviously this should be `hint | ! mpv`
- hints don't support pipes because of a bad architectural decision: we have most stuff living in a background script, when really we should probably have most stuff run on the web page where all the objects live, so that passing them around is easier
- we don't have an argument parser, so silent versions of commands have to be separate functions
...but I'm very glad you like it!
Regarding your pain points:
1. Tridactyl actually works really soon after the page starts loading, but only for commands that don't add UI. Personally I use quickmarks to escape from slow loading pages. Otherwise, this will probably never be fixed.
2. This isn't actually true. Tridactyl does override page binds. The reason you think GitHub is stealing a Tridactyl bind is because GitHub steals `/`, but Tridactyl doesn't bind anything to `/` - that's a default Firefox bind! (And whoever decided it would be a great idea to use that in GitHub should probably test their code in more browsers).
Yeah Tridactyl is great. While I'm not a great JS/TS developer I might be able to offer a couple of pull request for documentation. Is there anything I can help with currently?
> 2. This isn't actually true. Tridactyl does override page binds. The reason you think GitHub is stealing a Tridactyl bind is because GitHub steals `/`, but Tridactyl doesn't bind anything to `/` - that's a default Firefox bind! (And whoever decided it would be a great idea to use that in GitHub should probably test their code in more browsers).
The next big win in documentation is configuration docs, but there's a type doc bug I'm waiting on first before I can merge it.
However, the `tutor`, new tab page, help page introduction and addons.mozilla.org page (stored on repo as amo.md) always fall out of date, so time spent reading through those and adding things that are missing or out of date is always time well spent. The changelog is probably the easiest source of truth to work from.
Can native message fix the fundamental problems in keyhandling? Meaning it can react to keys without the site fully loaded? Work on all sites and dialogs? Have a predictable behaviour instead of timing-problems which can happen now if you type to fast? For example, opening the commandline and typing something to fast can have the effect that the typed text does not end in the commandline but instead is handled outside of it.
With all due respect, less than 0.1% of browser users likely feel that way.
Don't mean to discredit or invalidate your feelings, I just feel the need to disclaim that it's impossible to meet people's need with that granularity.
The alternative is that they keep XUL, make 0.1% of the install base happy, and ditch the performance increases that Quantum/WebExtension have given us... at which point there wouldn't be a NYT article and we'd all just be stuck in a browser monoculture.
I do not believe that was the only alternative. I have spent a lot of time reading bug reports / feature requests relating to this functionality, and it certainly seems that much of it could be enabled, but apparently never will. The developers seem to think that giving extensions "too much power" would be dangerous, and thus it is best not to let anybody choose power over safety guarantees.
It seems a little too harsh to bring up Linus Torvald's "If you think your users are idiots, only idiots will use it" quote, but it tends to come to my mind whenever I read the abovementioned discussions.
Ironically, the thing that drove me to Firefox was the excellent Account Containers system that I found necessary to manage being logged into multiple Google Apps accounts simultaneously (I have five). Each tab owns one account — that way, Google Drive files from my work don't randomly refuse to open because I'm also logged into my personal account.
(Chrome has no solution for containing accounts across tabs, last I checked. You have to spawn entire new Chrome instances.)
ADDED: In general a lot of services don't handle multiple accounts terribly well. I've had issues with Wordpress as well. I actually have two different soft-key authenticator programs related to making personal and work accounts play nice.
Sounds like you're using the Google Accounts feature, and using the [separate] browser profiles feature would exactly solve your problem.
In particular: you can log into different profiles in the browser (each window can be associated with up to one profile, and they can be different), by clicking the little "Sign In" tab in the upper right corner. See also https://support.google.com/chrome/answer/2364824?hl=en&co=GE....
You can then open tabs under the different profiles with right-click, "Open link as", and selecting the appropriate profile.
(Personally, once I tried profiles, I never touched multiple-accounts-logged-in again.)
Yes, it does work now in Chrome, I think it just changed is most recent update so not sure why guy above is acting like its alwasy been that way, I'd say its time to try Chromium (Brave) again.
stackdriver.com has to be one of the worst! selecting the right account just causes an infinite loop if the correct account isn't index[0] in your logged in accounts!
Support for multiple accounts in drive is pretty great (minor papercuts here and there) as long as all but one of your accounts is an Apps account. This feels like an issue that they've fixed but not rolled out.
Except this has been a problem for years, and according to a friend of mine who's had a "personal" Apps account from back when it was free, never has been a problem for Apps users, so I'm not sure what's going on on their end.
That's probably because your work account is a Google Apps account, which works much better especially for Drive (the problem is that you can't have more than one non-Apps account using drive at once in the same session. I have no idea why, but it seems to be deliberate). I'm unsure if the brokenness is fixed with Apps accounts on other Google services that are bad at handling multiple accounts.
What does that mean in terms of resources? Is 2 FF tabs that much less taxing than 2 Chrome instances?
fwiw, I use multiple Chromes, each with its own project / client, own password manager, etc. I don't think I could maintain the silos I do with FF. I wish I could sack Google (Chrome) but I can't, afaik.
Firefox supports both, completely separated browser profiles and these more lightweight Container Tabs.
You can type "about:profiles" into the URL-bar to manage your profiles (which you can also bookmark for easier access).
You can also run "firefox -P" in a terminal to access a similar GUI. With "firefox -P profile_name", you can start a specific profile. To run multiple profiles in parallel from the CLI, you have to start further profiles with "firefox -P profile_name --new-instance" (Linux) or "firefox.exe -P profile_name --no-remote" (Windows). This is because when you click on a link in an application, it needs to know in which browser instance to open that link.
As for resource usage, the profiles separate everything. Browsing history, bookmarks, logins, add-ons etc. It's essentially like a second Firefox installation. This also means that it actually runs two instances, so twice the RAM usage and somewhat more CPU usage to move that data around.
The Container Tabs on the other hand only separate what needs to be separated in order to look towards webpages like two different browsers. So, there's a different cookie jar, different localStorage, different HTML5 Web Storage etc., but other than that it's the same as just opening two tabs. So, probably not even 1% more RAM usage than using Firefox without any separation.
Yeah, in about:profiles, there's a "Launch profile in new browser"-button.
For launching multiple profiles in parallel from the command-line, I already tried to explain it in my previous comment, though it is more finicky that way.
Supposedly you can be logged-in to multiple Google accounts but… it's very broken most of the time and works only for basic gmail browsing sadly.
And containers in firefox are simply amazing! I use it very often to access "admin accounts" on the occasion without the hurdle to use private browsing.
Not even tabs usually live in the same process, so you must be misinterpreting something.
I think, the Windows Task Manager groups processes in the most stupid of ways, where when a (parent-)process has spawned many child-processes, then only the parent-process and its resource usage is shown. You have to go into the "Processes"-tab to see all Chrome-processes and manually add up their resource usage.
I spent about three hours trying to figure out why apps scripts were breaking and refusing to open the editor until finally finding a helpful blog post: logged into multiple accounts? You may have trouble with apps scripting-something I had come to rely on with Sheets.
I wouldn't call it 'infuriating', but it's something that I can't figure out "how does this of all things break this of all things?"
First, it almost never remembers what account you used with what service. If I do some Google Cloud stuff under my work account, I don't want Google Maps using my work account, or vice versa. The list of accounts in the upper right corner, and what's signed in/out, seems completely random and broken.
The second problem is that Google's various apps are wildly inconsistent in how they deal with multiple accounts. Some (like Maps) have apparently been "modernized". Others (like BigQuery; I've had issues with Google Drive, too) don't understand multiple accounts at all. It doesn't matter that you sign into a work account. Open that app and it will use some other account.
A sign they've not mastered this at all is that it doesn't work at the URL level. You can bookmark a Google Docs document, for example, and then switch your accounts. Going to the bookmark (or just going back) to the Docs page gives you a "permission denied" error because you're using the wrong account. Same thing with sharing URLs — if someone at work sends me a Google Docs URL, it's bound to not work because it'll try to use what I try to keep as my default account, which is my personal one. Sometimes I just munge the "authuser=1" part in the URL.
It's amazing that a Google app can't simply resolve the union of identities you're signed in as. I have two accounts, work and personal, and one of them has access, after all.
I have this problem with JIRA: ever since Google authentication was enabled there I can not login on two separate JIRA instances with the same browser.
I dont want this, it's better to be explicit with the accounts. However they are just using cookies and Chrome gives each "profile" a separate cookie container which works well and I prefer it to the mess it would be at the tab level. I have separate window profiles for work and personal and can keep everything (including passwords) completely separate.
I use Safari, which doesn't have profile handling at all. The web should not rely on clumsy browser hacks. Chrome profiles are way too coarse-grained — all the settings and state (history, bookmarks, etc.) are separate. It also interacts badly with everything else on a computer. If I click on a Slack link to a Google document, then nothing knows which Chrome window to open it in.
Google's mistake was the account switcher. Supporting multiple accounts is fine, but pretending they support concurrent logins is nonsensical.
But I have no choice to use multipel Google accounts, since my company uses Google Apps or G-Suite or Google For Work or whatever it is called today, and I'm consulting for a third company that also does this.
This kind of defeats the purpose of multiple accounts.
Remember, we started to have them to separate our work/home/personal lives, to build separate digital profiles that companies (like Google) store on their servers. Now you beg Google to recognize and unify them (which I guess they have almost done already by themselves).
I'm not begging anything of Google, but I would want them to make this work better.
Google accounts are not just about data and personalities, but also about permissions. As long as I'm "logged into" multiple accounts, it should be able to map the content I'm trying to see to an account. Or at least offer to switch to that account. For example, if I'm trying to view a company document, but my default account is my personal account, then I don't want to see an often obscure "permission to view denied" error, which always freaks me out and makes me think I've been locked out or something.
Better yet, build this badly-needed auth stuff into the browser and allow apps to negotiate access with the browser.
I just setup multiple profiles in chrome which is one thing it seems to do better than FF imo. You can select a profile in the top right corner. Everything about them is separate. cookies, history, extensions , theme so I give each one a different theme make it clear what profile I'm using.
in FF afaict you can only use one at a time and you have to choose at launch
Your comment has convinced me to do this. This is pretty fantastic. It was driving me nuts being signed into both work and personal gmails to access either my files or adwords. Not to mention YouTube which would open under the work account sometimes or drive. Plus now I can rip out all the work bookmarks and shove that into a work only profile.
My only concern now is with using multiple devices. Do I just sign into to both profiles on each device? I only worry because there was a big snafu that occurred with their password manager that synced nothing and everything disappeared. (I no longer trust google with that stuff).
Chrome profiles are too coarse-grained: Not only is the entire window associated with that profile but each profile also has separate settings, histories, bookmarks, everything. All I'd want here are separate cookie spaces.
Firefox profiles are similar to Chrome's. But the container tabs that people are talking about in this thread are better than Chrome profiles. They're per tab, and last I checked they only separate out cookies and local storage.
Yeah, but they probably get punished by managers for implementing unnecessary features. And that's probably why, even if we have one, the multiple accounts feature sucks so much.
Right, Chrome can't silo things like that between tabs, but can between windows using separate chrome profiles. Not as handy as container tabs but not horrible either.
Agree that you shouldn't try mixing personal and Google apps account in the same session...it always breaks eventually. (Why can't Google get that right?)
> the thing that drove me to Firefox was the excellent Account Containers system
Though unfortunately, history and bookmarks are not isolated to their individual containers, and Firefox's profile management system is a PITA compared to Chrome's easy button in the title bar.
I switched cold turkey to Firefox a few months ago. It was fine for a bit but eventually became unbearably slow. I made the switch on both my work MacBook Pro (retina) and my personal MacBook Air. Both had similar issues. That and the fact that it ships with things like Pocket made me jump back to Chrome.
What you didn't find to be true in your case - Firefox becoming slower over time (as OP says) or Firefox shipping with Pocket? Or both?
Personally my reason was Pocket incident and since then I never went back to Firefox. Checked few months ago and they still had Pocket. Besides I am pretty happy on Safari except the lack of extensions which I am kind of used to by now.
They still have Pocket, but it's both less in the way and now actually owned by the Mozilla Foundation though that's not really made obvious.
The most obvious place you'll see it is on the default new tab/blank tab page where top trending stories being saved on Pocket are listed. It's a crowd-curated list of stories that people have found interesting enough to save to read or return to later, almost like a technology news site where people vote stories up...... But no commenting.
The way Pocket is integrated in Firefox is a problem for me, may not be a problem for others and that's completely understandable.
For me Mozilla was the "open" browser (and "clean" too). Integrating Pocket in it by default, in a way that it cannot be "completely" removed, put me off. Heck, they could just ship the browser with it installed as a "normal" extension that can be "completely" removed like any other extension. And maybe disable this extension by default, or keep it enabled they'd really want it that way.
And no, if I want curated trending stories I will go the website I want to read those from. Or I'll install a relevant extension of my choice; maybe Pocket.
I use it as a daily driver since quite a long time (2 years I guess (* )) and I find the performance not to be an issue (MBP retina late 2013), especially considering having open most of the time 4 windows with about 50 tabs. It can warm the machine on opening 5-10 amazon pages at once but there is no slowdown.
As for Pocket - I don't use it, but prefer this non-inclusive addon instead of Google spying on my every move (yes, you don't have to log-in to google account and you can install ad-blockers but still). And also I don't want to end up with "Best viewed with Inter^WChrome" Internet.
(* ) (using Firebird, then long time Opera user until version 12, then made a switch to Firefox, then to Vivaldi for a 1-2 years and now back to Firefox again)
How is Pocket any different from Safari’s Read It Later, or all of Google’s equivalents? And unlike the other browsers, Pocket while included by default, is still an extension instead of baked into the browser.
It really isn't, but neither Chrome nor Safari are in the habit of making obnoxious UI changes, force-pushing addons I don't want, putting ads on the new tab page (which you would think would be more Chrome's bailiwick, but they don't), and so forth.
I don't want to be using Firefox when whatever the next boneheaded move they make lands. Mozilla burned every bit of goodwill I ever for them had when they destroyed an odd decade or so of muscle memory and customization of a tool I use every day. Yeah, so they're supposedly pro privacy. Neat. Whatever. They also have the organizational attention span of a toddler on Red Bull.
If Firefox wants to be taken seriously they have to stop showing ads in the home page. Period. It looks way too unprofessional. Like some adware browser you’d have to uninstall from your grandmas Windows XP computer.
And there is only one thing which I don't like — high CPU usage compared to chrome. It makes my old MacBook air very noisy because of the fan spinning. When I use chrome it's completely silent.
Do you have details? I noticed that Youtube on Firefox makes the fan spin up on my 2013 rMBP a lot. I currently use Firefox but as a heavy Youtube watcher this is a pretty big negative.
Same here. I have been a longtime Firefox user and don't recall having performance issues with YouTube on macbook before Quantum. But now if I'm watching anything more than a short clip on YouTube, the MBP's fans start to annoy me and it feels like it is going to burn a hole in my lap. YouTube in Chrome is totally fine. YouTube on Firefox on my Windows Intel NUC desktop is also totally fine. It seems to be a mac only Firefox issue. Would really love a solution!
Firefox uses H.264 for YouTube on macOS, so it can use the system's H.264 hardware decoder, which should be faster and use less power than a software decoder. Chrome uses a VP9 software decoder on YouTube. You can try VP9 in Firefox by setting the about:config pref "media.mediasource.webm.enabled" = true, but it is disabled by default for good reasons. :)
This is because Youtube pushes VP8/VP9-encoded video by default to save on bandwidth. Unfortunately, absence of a hardware decoder consumes a ton of battery. Both Firefox and Chrome receive VP9 encoded video.
Safari doesn't understand VP9. So Youtube pushes H.264 video which can be hardware accelerated.
Interestingly, for me Chrome is the browser that causes my machine to sound like a helicopter on take off (though maybe my usage is atypical). Safari still has the best power efficiency on the macOS, but between Firefox and Chrome anecdotally it seems like a toss up.
Would be curious to see some actual benchmarks comparing the latest browsers, does anyone do those on a regular basis? A quick google search didn't turn up anything more recent than last december: http://themainframe.ca/comparing-firefox-chrome-and-safari-b...
I do. For me, it looks like Firefox is struggling with rendering when it comes to any kind of animation/effects, e.g videos, images, as well as js powered animations using SVG/canvas. Even simple things make my laptop boil.
Maybe I'm wrong about CPU and it's all about integrated Intel GPU.
I'm having similar issues on a fairly modern laptop with i7-7700hq and GTX1050. After a few minutes of watching YouTube, it gets all jerky and not smooth. I'm guessing something is throttling, though video playback isn't taking a high percentage of either the CPU or the GPU. I tried forcing the builtin Intel GPU on firefox.exe, that didn't change anything.
YouTube prefers VP9, when available. You can try disabling VP9 in Firefox by setting the about:config pref "media.mediasource.webm.enabled" = false. Depending on your GPU, Firefox may be able to decode H.264 using a hardware decoder.
You can check in about:support whether it is disabled. There's a section at the bottom of the article about how to force-enable various features if you want to see how glitchy it really is on your machine.
I would really love if Firefox didnt make me wait until it install an update while I need to quickly test something on an alternative browser. Can I be the one who decides if I want to install the update right now or do whatever I wanted to do in the first place.
The chrome print dialog is totally broken for us. It always tries to use the wrong size paper from an empty paper tray. We always need to click through to the system print dialog. I need to tell people to do that every single time because they forget every single time. I should switch everyone to FF just for that.
On Android the performance is comparable to Chrome, but I can never get used to some of their UXes (pressing the X button doesn't clear the address bar, instead closes it). Also no seamless page opening with Google search app (although I suspect they can't do much about this)
I've never found that confusing. Tapping the address bar opens a whole new navigate/search view. X closes that view. The most common use case is entering a new search or url, just start typing, the existing text is already selected so it will overwrite. On the off chance you want to edit the current URL, it's there. I can't think of a more streamlined workflow. Whatever you're used to is only good because you're used to it, but it's adding extra steps and/or unnecessary UI elements.
I keep trying to switch back to it on my Mac Pro (I've tried three times in the last 12 months) but after a week or two I always end up switching back to Chrome because the CPU/battery impact it has.
I can have the same amount of tabs/workload in Chrome and my laptop copes fine with good performance. If I do the same in FF (with likely less plugins installed) it often spikes to 100% CPU usage, chews through my battery like nothing else and just gets bogged down/slow. I _want_ to use FF but can't sacrifice having a burning hot laptop with terrible battery life on my lap.
Exact same deal here. That plus the lack of keyboard shortcuts for many extensions keeps pushing me back to Chrome. But I love the account containers enough that I keep trying again. I imagine it will get there eventually.
Yeah, I mean I only switched from IE because it froze my computer if it ever opened more than one window (this was a Windows ME system with about half the ram ME claimed as required). I've been Mozilla since ~2002. The idea of switching for anything short of the apocalypse just seems weird, like getting rid of a car because it's a couple years old.
I have been mainly using Firefox for the last 3 months on desktop, and for over a year on phone - and it's amazing. Highly recommend it. I even uninstalled Chrome now on my main machine- sometimes you run into what appears "Chrome optimized" sites, like back in IE6 days but its rare.
Chrome is still king when it comes to fast JS execution and modern technologies. I'm talking about real world use, I know some benchmarks put FF and Edge ahead.
The recent post about real time pose estimation was a pretty example of that - pretty unusable in FF, runs great in Chrome (on Windows at least)
Pretty much any page with paralax scrolling or great ammount of images - and you can barely scroll the page. It's gets better with 62, but Fx still struggles with pages like https://www.hellomonday.com/
I never really "switched" to any particular browser, I use Safari/Vivaldi/Firefox concurrently, but have noticed how much better Firefox has gotten recently, after years of painful decline (Speed, and resource hogging being my biggest peeves).
The only major thing I wish it did better is if it conserved battery power as well as Safari does. If I want to maximize my laptop's battery life I don't really have a choice other than Safari on my MBP.
I've been using safari for a few months as my main browser and while it does feel quick, I can't say I've noticed a huge difference in battery life compared to chrome or firefox.
Are there any (recent) tests you know of that compare battery life between safari and the other major browsers?
Firefox has enough features and customization available that it's the only non-native browser that I'll use. I don't think Chrome or the others are worth it.
Otherwise in general, the native browser is usually the best bet for battery life. They render differently and have tighter integration overall as you noted. Firefox or native is the way to go, and if native browsers keep improving, even Firefox should watch out.
Firefox also seems to have better support for WebVR than Chrome, at least on Windows. As a very rough measure, half of the A-frame official demos don't render for me on Chrome, but all of them render on Firefox. They're also choppy on Chrome compared to Firefox.
I've switched back from chrome for the quantum release. Overall, it works fine 99% of the time but since chrome became the new IE there are a few sites that break in non-webkit browsers. Not enough of an inconvenience to make me go back to handing over all my info to Google.
I switched back a few months ago on all of my devices and only go back when I need Chrome devtools, Firefox is absolutely back in its prime. Feels just as speedy as Chrome on desktop and mobile, and handles large numbers of tabs better.
Ironically, or maybe intentionally, Google services are the only thing degraded. Google even refuses to give you the regular home page. Thankfully you can just get an extension that fakes Chrome useragent on the offending pages, but it's the darkest pattern I've ever seen
Of course there’s no sponsorship disclosure. That’s not what the OP means.
This article was seeded by a PR firm. It’s not directly sponsored by Mozilla. It’s been professionally “placed” in the NYT.
Then again, I don’t know why I’m arguing with you. Judging by the style of your comment, I’m guessing you didn’t really go back to check if the article contained any sort of sponsorship notice (because, seriously, what real person would do that and then come back and leave a comment?). Your account is probably owned by the PR team that got the article placed and then pushed it to the top of HN. #AmIRight
I did give it a try. I used it like I use Chrome. Many tabs open and keeping them open for a couple of days. I did that a couple of times. It, of course, crashed every single time and for some reason made my 4GHz quad-core i7 with 64Gb memory freeze for 5 minutes where I couldn't even ^d in a terminal.
If I would have ever switched to Chrome, it would have been to save probably less then one minute per day (when Chrome was faster)... not worth it for me.
That would be great. However, the webs standards are bloated as hell. I know you were probably being factitious. However, I think it's a big problem that web standards have gotten so complicated.
The browser would also have to backward compatible, and handle mal-formatted and bloated code written by god who knows. There are features that are the antithesis of privacy and disabling some of the features breaks quite a few sites. Sadly, the web is no longer linked documents, but has full blown apps. It's honestly freaking insane.
This sums it up nicely. I refuse to use Firefox until they remove Pocket, it's an absolute tragedy that it was ever included by default and forced onto a user base that never asked for it.
It was acquired so that Mozilla could create a new revenue stream with recommended & sponsored content. To rephrase, they forced a feature on me so that they could sell ads.
First, they didn't own Pocket when it was first integrated. Second, for years Firefox removed features like tab groups in order to make the browser easier to maintain because that functionality could be added as a plugin. Then they force everyone to use a plugin for a proprietary service? It was very hypocritical of them and I still have a bad taste in my mouth from that.
At the risk of stating the extremely obvious to the Linux user that likes to build their browser from source: I don't think the advice in the article was intended for you.
1. Yes. The bug you linked to is prominently marked "RESOLVED WONTFIX."
2. We are not sending DNS data to third parties "for analytics." We are openly and transparently conducting a DNS-over-HTTPS (DoH) experiment with a very small portion of users running Nightly test builds of Firefox. The experiment places a massive notification banner in the browser window with prominent buttons to disable or accept the experiment.
3. Firefox is Free Software, under an OSI-approved license. You're not "required" to build pocket, nor can or would we "forbid" you from removing it when you build Firefox.
Still seems like a really bizarre thing to get hung up on when the extent of "Pocket in Firefox" is a toolbar icon and a section of the new tab page, both of which can be hidden with 1-2 clicks, after which they're completely inert. But hey, whatever floats your goat.
I switched from Chrome to Firefox a few months ago and am mostly happy, with just a few annoyances:
There's a setting to prompt "are you sure" before closing Firefox, but it doesn't actually work. I really miss this feature (which works as expected in Chrome) since unless I'm turning my machine off a pretty much never close my browser intentionally
Shortcuts in text input also work differently than every other program on my computer. Command-backspace deletes the whole text field, not just the text before the cursor. ^t doesn't transpose, ^k behaves differently when text wraps, etc
Yeah, there's some arguments about that on their bug tracker page and it looks like someone tried to change that code but it never went anywhere. If you horde tabs like I do it can still waste a lot of time to restart the browser and wait for the session to be restored.
Pretty consistent. I haven't timed it, but it certainly feels like more than 15 seconds for my 50-100 tabs. I'm not in quite as pure of a situation as that benchmark though-- I have plenty of other programs using memory and I'm waiting for the page to actually render. During some of my Firefox launches lately I also had a bunch of other applications launching at the same time since I was trying to debug a WebGL page that sporadically hangs/reboots my machine (in both Firefox and Chrome)
Firefox was never away, at least for me. Therefore the expression "Firefox is back" sounds a little odd to me.
I have been using it since 2006 (Firefox 2.0). Back then, the presence of the Firebug extension for Firefox made me switch to Firefox as my primary browser.
Two years later, Chrome arrived but Firefox remained as my primary browser because I wanted to continue using the Vimperator extension for Firefox that allowed mouseless browsing with Vim-like key bindings and commands.
Firefox still remains as my primary browser with Chrome being an additional browser that I use sometimes. These days, I use the Vimium extension for both to use Vim-like key bindings. I occasionally use Safari and qutebrowser too.
And don't forget about those thousands (millions?) of users who suddenly found Chrome installed on their computer because they updated their Virus Scanner or installed some Freeware where it was bundled with Chrome.
Still an upgrade from IE8 the same group would be using. I'm not advocating bundled installs, but installing secure (compared to old IE) browser alongside antivirus software is one of the very few cases where it makes sense.
From my private IT support perspective: I hate it!
My "customers" have FF installed and are every time utterly confused missing their favorites on their "Internet Program".
There is so much wrong with this bundling but yes, it's probably still better then IE...
I have a very similar background with Firefox / Firebug and Vimperator, but there was a period where - despite all JS benchmarks seemingly proving otherwise - Chrome's real / perceived performance completely blew FF out of the water. Specifically things like startup time, opening a new window or tearing a tab off one, time to first content displayed when clicking a link, etc. As a result from about 2011 to 2017, I kept periodically checking back with FF, but was effectively a Chrome convert (Vimium was a good enough replacement for Vimperator). Once Servo hit the FF mainstream, it immediately felt faster, and I switched back. I still find myself switching back to Chrome for web dev debugging, probably for familiarity reasons. In this sense though, I'm sure I'm not the only one for whom "Firefox is back".
Oh really ... Not having DownThemAll just breaks my entire workflow, and there's no replacement. They should have at least considered the "most popular add-ons" to be in working condition before even releasing WebExtensions to all.
> High on my list is for the extensions to work on [...] addons.mozilla.org pages.
There's actually an about:config setting that fixes this already. I put a helper in Tridactyl for it called `:fixamo` [1], but you can easily do it manually by setting
Have you tried adding the setting that doesn't exist manually? There's a little + button you can click on about:config.
---
That's a particularly unhelpful error, sorry about that. It usually means that Tridactyl can't access the native messenger. Did you install it with `:installnative`? What is the output of `:native`?
> That's a particularly unhelpful error, sorry about that. It usually means that Tridactyl can't access the native messenger. Did you install it with `:installnative`? What is the output of `:native`?
I did not, but now I have.
:fixamo now doesn't return any error but still doesn't work.
Adding "privacy.resistFingerprinting.block_mozAddonManager" manually and setting it to true did the job. Thanks!
So far I like Tridactyl, especially the useful commands, but since it seems to be incompatible with other addons that overwrite the NewTab page. I think I will have to disable it.
I also miss the smooth scrolling experience with j,k from VimiumFF, jumping lines like Tridactyl does makes me loose the point were I left of. I might check back how the progress is going tough.
> :fixamo now doesn't return any error but still doesn't work.
You need to restart the browser afterwards. Sorry, people usually find the command through the help page where it has all of these caveats.
We have smooth scrolling, you just need to do `set smoothscroll true`. It's a bit rubbish, though.
You can often fix the new tab page by disabling and re-enabling the new tab add-on you want. Every time Tridactyl is updated it will steal the page back, so I'd suggest that you don't use the beta releases and instead use the normal ones.
We could reasonably easily provide a build that didn't use the new tab page (we occasionally get people who feel very strongly about this) but I'm afraid it's quite low down my list of priorities. If you're feeling up to it there's an issue open: https://github.com/cmcaine/tridactyl/issues/534
But I won't hold it against you if you use Vimium. It's very well polished : )
> You need to restart the browser afterwards. Sorry, people usually find the command through the help page where it has all of these caveats.
:help fixamo
Does not state that I need to restart firefox afterwards. [1]
I did restart firefox after I installed `native` and before I run :fixamo. :native also returned that the it was installed correctly before I run :fixamo.
[1]:
fixamo
fixamo(): Promise<void>
Defined in src/excmds.ts:290
Simply sets
"privacy.resistFingerprinting.block_mozAddonManager":true
"extensions.webextensions.restrictedDomains":""
in about:config via user.js so that Tridactyl (and other extensions!) can be used on addons.mozilla.org and other sites.
> If you're saying that a restart didn't make fixamo work, would you mind filing an issue so we can try to fix it?
No I haven't tried that. After I run :fixamo and that didn't work I just opened about:config and change those settings myself without restarting firefox.
Edit:
Since I now continued to test tridactyl I found that this newtab overwriting function gets pretty annoying from a usability perspective. When I start a new Tab with [Ctrl]+[t] and immediately type my desired address part of it will be overwritten when 'about:blank' is inserted. I guess you want people to use just a binding to :tabopen, but I do like my options. Also since :tabopen do not contain my bookmarks I prefer using the address line.
> Since I now continued to test tridactyl I found that this newtab overwriting function gets pretty annoying from a usability perspective. When I start a new Tab with [Ctrl]+[t] and immediately type my desired address part of it will be overwritten when 'about:blank' is inserted. I guess you want people to use just a binding to :tabopen, but I do like my options. Also since :tabopen do not contain my bookmarks I prefer using the address line.
Huh, that is annoying. I hadn't noticed that. There's not much we can do about it though, short of just disabling the new tab page.
The lack of bookmarks in tabopen is annoying, I agree. We rushed to get the completions out with Tridactyl, and did it shoddily and haven't bothered to rewrite it yet.
There is bmarks which just completes from bookmarks, but that opens things in the current tab.
Breaking "old" add-ons with no working replacement and no way to implement quite a few of them killed FF for me. What those idiots at Mozilla don't get is that without the add-ons their stock browser is no better than any other.
I can also use it on android with ublock etc.
Chrome hasn't even been installed in my main computer in maybe a year, firefox works for everything I need, including google apps for business mail and youtube.
My biggest reason for using it is the fact that it is open source, the other is that ad-block software actually block calls to adds not just the display of them as in chrome.
Isn’t it essentially the prototype for all modern browser dev tools though? As in, they are all substantially similar to what Firebug was like, and Firebug existed before any of the others?
I could be getting my history wrong, but from what I remember Firebug was absolutely mind-blowing.
752 comments
[ 2.2 ms ] story [ 427 ms ] threadWant a bind to open something in mpv/youtube-dl `";v": "hint -W exclaim_quiet mpv"`
Want to open a text field in gvim? Hit `ctl+i`
Want to remove the extra chrome: run `:guiset gui none`
I stuck on the ESR of firefox for a long time because I wasn't willing to give up vimperator. Its not the same, but its pretty close. There are 2 main pain points, which I don't know it will be possible to address. First the page has to before its functional. Second currently domain keybinds take priority (looking at you github).
> Want to remove the extra chrome: run `:guiset gui none`
Top bar with so much padding is one of the main reasons I am not using Firefox.
Also the awesome command bar at the bottom is something that I've always found missing in vimium, (which has an omnibar but it's placing feels awkward.)
I think I am going to make firefox as deafult.
You know how mathematicians get excited by exp(-i*pi)+1=0? This is like that but for all of the things I'm embarrassed about in Tridactyl:
- our documentation is not great, particularly of current configuration/possible configurations, hence why the GP didn't know about it
- hints don't support pipes - obviously this should be `hint | ! mpv`
- hints don't support pipes because of a bad architectural decision: we have most stuff living in a background script, when really we should probably have most stuff run on the web page where all the objects live, so that passing them around is easier
- we don't have an argument parser, so silent versions of commands have to be separate functions
...but I'm very glad you like it!
Regarding your pain points:
1. Tridactyl actually works really soon after the page starts loading, but only for commands that don't add UI. Personally I use quickmarks to escape from slow loading pages. Otherwise, this will probably never be fixed.
2. This isn't actually true. Tridactyl does override page binds. The reason you think GitHub is stealing a Tridactyl bind is because GitHub steals `/`, but Tridactyl doesn't bind anything to `/` - that's a default Firefox bind! (And whoever decided it would be a great idea to use that in GitHub should probably test their code in more browsers).
Yeah Tridactyl is great. While I'm not a great JS/TS developer I might be able to offer a couple of pull request for documentation. Is there anything I can help with currently?
> 2. This isn't actually true. Tridactyl does override page binds. The reason you think GitHub is stealing a Tridactyl bind is because GitHub steals `/`, but Tridactyl doesn't bind anything to `/` - that's a default Firefox bind! (And whoever decided it would be a great idea to use that in GitHub should probably test their code in more browsers).
Didn't know that. Interesting.
The next big win in documentation is configuration docs, but there's a type doc bug I'm waiting on first before I can merge it.
However, the `tutor`, new tab page, help page introduction and addons.mozilla.org page (stored on repo as amo.md) always fall out of date, so time spent reading through those and adding things that are missing or out of date is always time well spent. The changelog is probably the easiest source of truth to work from.
Thanks!
Some things have Bugzilla tickets to fix in a future release, but something just won't be possible as it stands today.
Don't mean to discredit or invalidate your feelings, I just feel the need to disclaim that it's impossible to meet people's need with that granularity.
The alternative is that they keep XUL, make 0.1% of the install base happy, and ditch the performance increases that Quantum/WebExtension have given us... at which point there wouldn't be a NYT article and we'd all just be stuck in a browser monoculture.
I do not believe that was the only alternative. I have spent a lot of time reading bug reports / feature requests relating to this functionality, and it certainly seems that much of it could be enabled, but apparently never will. The developers seem to think that giving extensions "too much power" would be dangerous, and thus it is best not to let anybody choose power over safety guarantees.
It seems a little too harsh to bring up Linus Torvald's "If you think your users are idiots, only idiots will use it" quote, but it tends to come to my mind whenever I read the abovementioned discussions.
(Chrome has no solution for containing accounts across tabs, last I checked. You have to spawn entire new Chrome instances.)
ADDED: In general a lot of services don't handle multiple accounts terribly well. I've had issues with Wordpress as well. I actually have two different soft-key authenticator programs related to making personal and work accounts play nice.
In particular: you can log into different profiles in the browser (each window can be associated with up to one profile, and they can be different), by clicking the little "Sign In" tab in the upper right corner. See also https://support.google.com/chrome/answer/2364824?hl=en&co=GE....
You can then open tabs under the different profiles with right-click, "Open link as", and selecting the appropriate profile.
(Personally, once I tried profiles, I never touched multiple-accounts-logged-in again.)
Except this has been a problem for years, and according to a friend of mine who's had a "personal" Apps account from back when it was free, never has been a problem for Apps users, so I'm not sure what's going on on their end.
fwiw, I use multiple Chromes, each with its own project / client, own password manager, etc. I don't think I could maintain the silos I do with FF. I wish I could sack Google (Chrome) but I can't, afaik.
You can type "about:profiles" into the URL-bar to manage your profiles (which you can also bookmark for easier access).
You can also run "firefox -P" in a terminal to access a similar GUI. With "firefox -P profile_name", you can start a specific profile. To run multiple profiles in parallel from the CLI, you have to start further profiles with "firefox -P profile_name --new-instance" (Linux) or "firefox.exe -P profile_name --no-remote" (Windows). This is because when you click on a link in an application, it needs to know in which browser instance to open that link.
As for resource usage, the profiles separate everything. Browsing history, bookmarks, logins, add-ons etc. It's essentially like a second Firefox installation. This also means that it actually runs two instances, so twice the RAM usage and somewhat more CPU usage to move that data around.
The Container Tabs on the other hand only separate what needs to be separated in order to look towards webpages like two different browsers. So, there's a different cookie jar, different localStorage, different HTML5 Web Storage etc., but other than that it's the same as just opening two tabs. So, probably not even 1% more RAM usage than using Firefox without any separation.
For launching multiple profiles in parallel from the command-line, I already tried to explain it in my previous comment, though it is more finicky that way.
And containers in firefox are simply amazing! I use it very often to access "admin accounts" on the occasion without the hurdle to use private browsing.
https://tweetdeck.twitter.com/
I think, the Windows Task Manager groups processes in the most stupid of ways, where when a (parent-)process has spawned many child-processes, then only the parent-process and its resource usage is shown. You have to go into the "Processes"-tab to see all Chrome-processes and manually add up their resource usage.
Ugh, this is a common problem apparently. Why can't they just make it open files from my org domain using the only org account I'm logged into. >:(
I wouldn't call it 'infuriating', but it's something that I can't figure out "how does this of all things break this of all things?"
First, it almost never remembers what account you used with what service. If I do some Google Cloud stuff under my work account, I don't want Google Maps using my work account, or vice versa. The list of accounts in the upper right corner, and what's signed in/out, seems completely random and broken.
The second problem is that Google's various apps are wildly inconsistent in how they deal with multiple accounts. Some (like Maps) have apparently been "modernized". Others (like BigQuery; I've had issues with Google Drive, too) don't understand multiple accounts at all. It doesn't matter that you sign into a work account. Open that app and it will use some other account.
A sign they've not mastered this at all is that it doesn't work at the URL level. You can bookmark a Google Docs document, for example, and then switch your accounts. Going to the bookmark (or just going back) to the Docs page gives you a "permission denied" error because you're using the wrong account. Same thing with sharing URLs — if someone at work sends me a Google Docs URL, it's bound to not work because it'll try to use what I try to keep as my default account, which is my personal one. Sometimes I just munge the "authuser=1" part in the URL.
It's amazing that a Google app can't simply resolve the union of identities you're signed in as. I have two accounts, work and personal, and one of them has access, after all.
I might consider FF just to work around this.
Google drive is the biggest annoyance though with multiple accounts.
I dont want this, it's better to be explicit with the accounts. However they are just using cookies and Chrome gives each "profile" a separate cookie container which works well and I prefer it to the mess it would be at the tab level. I have separate window profiles for work and personal and can keep everything (including passwords) completely separate.
BigQuery has a new UI built into the console: https://console.cloud.google.com/bigquery
You might need to sign up for the alpha/beta test: https://docs.google.com/forms/d/e/1FAIpQLSf6hyfvoWZ8eUbbKWq9...
Google's mistake was the account switcher. Supporting multiple accounts is fine, but pretending they support concurrent logins is nonsensical.
But I have no choice to use multipel Google accounts, since my company uses Google Apps or G-Suite or Google For Work or whatever it is called today, and I'm consulting for a third company that also does this.
But if google knows who you are and you’re not logged in at be moment? Hidden behind login page.
Remember, we started to have them to separate our work/home/personal lives, to build separate digital profiles that companies (like Google) store on their servers. Now you beg Google to recognize and unify them (which I guess they have almost done already by themselves).
Google accounts are not just about data and personalities, but also about permissions. As long as I'm "logged into" multiple accounts, it should be able to map the content I'm trying to see to an account. Or at least offer to switch to that account. For example, if I'm trying to view a company document, but my default account is my personal account, then I don't want to see an often obscure "permission to view denied" error, which always freaks me out and makes me think I've been locked out or something.
Better yet, build this badly-needed auth stuff into the browser and allow apps to negotiate access with the browser.
in FF afaict you can only use one at a time and you have to choose at launch
That said, do try multi-account containers, unless you really rely on extensions and theme being different per profile - they're great: https://addons.mozilla.org/en-US/firefox/addon/multi-account...
My only concern now is with using multiple devices. Do I just sign into to both profiles on each device? I only worry because there was a big snafu that occurred with their password manager that synced nothing and everything disappeared. (I no longer trust google with that stuff).
Firefox profiles are similar to Chrome's. But the container tabs that people are talking about in this thread are better than Chrome profiles. They're per tab, and last I checked they only separate out cookies and local storage.
Agree that you shouldn't try mixing personal and Google apps account in the same session...it always breaks eventually. (Why can't Google get that right?)
Though unfortunately, history and bookmarks are not isolated to their individual containers, and Firefox's profile management system is a PITA compared to Chrome's easy button in the title bar.
Personally my reason was Pocket incident and since then I never went back to Firefox. Checked few months ago and they still had Pocket. Besides I am pretty happy on Safari except the lack of extensions which I am kind of used to by now.
The most obvious place you'll see it is on the default new tab/blank tab page where top trending stories being saved on Pocket are listed. It's a crowd-curated list of stories that people have found interesting enough to save to read or return to later, almost like a technology news site where people vote stories up...... But no commenting.
(https://techcrunch.com/2017/02/27/mozilla-pockets-pocket-in-...)
Mozilla rewarding everyone involved in the Pocket debacle by buying the company was weird.
Or just open the Options page and click Home way over on the left.
For me Mozilla was the "open" browser (and "clean" too). Integrating Pocket in it by default, in a way that it cannot be "completely" removed, put me off. Heck, they could just ship the browser with it installed as a "normal" extension that can be "completely" removed like any other extension. And maybe disable this extension by default, or keep it enabled they'd really want it that way.
And no, if I want curated trending stories I will go the website I want to read those from. Or I'll install a relevant extension of my choice; maybe Pocket.
As for Pocket - I don't use it, but prefer this non-inclusive addon instead of Google spying on my every move (yes, you don't have to log-in to google account and you can install ad-blockers but still). And also I don't want to end up with "Best viewed with Inter^WChrome" Internet.
(* ) (using Firebird, then long time Opera user until version 12, then made a switch to Firefox, then to Vivaldi for a 1-2 years and now back to Firefox again)
Mozilla owns Pocket (Read It Later). It's sort of like me complaining Safari comes with iCloud support by default.
So sort of not.
I can't stand Chrome on OSX, I only use Safari when on battery and FF when plugged in. Chrome just brings my Mac Boook to a grinding halt.
https://support.mozilla.org/en-US/kb/disable-or-re-enable-po...
How to remove the "Recommended by Pocket" stories on the about:newtab page:
https://support.mozilla.org/en-US/kb/hide-or-display-content...
I don't want to be using Firefox when whatever the next boneheaded move they make lands. Mozilla burned every bit of goodwill I ever for them had when they destroyed an odd decade or so of muscle memory and customization of a tool I use every day. Yeah, so they're supposedly pro privacy. Neat. Whatever. They also have the organizational attention span of a toddler on Red Bull.
And there is only one thing which I don't like — high CPU usage compared to chrome. It makes my old MacBook air very noisy because of the fan spinning. When I use chrome it's completely silent.
This YouTube page will show which codecs are currently enabled in your browser: https://www.youtube.com/html5
Safari doesn't understand VP9. So Youtube pushes H.264 video which can be hardware accelerated.
To force youtube to send H264 in firefox/chrome you need something like https://addons.mozilla.org/en-US/firefox/addon/h264ify/
This YouTube page will show which codecs are currently enabled in your browser: https://www.youtube.com/html5
Would be curious to see some actual benchmarks comparing the latest browsers, does anyone do those on a regular basis? A quick google search didn't turn up anything more recent than last december: http://themainframe.ca/comparing-firefox-chrome-and-safari-b...
Maybe I'm wrong about CPU and it's all about integrated Intel GPU.
This YouTube page will show which codecs are currently enabled in your browser: https://www.youtube.com/html5
You can check in about:support whether it is disabled. There's a section at the bottom of the article about how to force-enable various features if you want to see how glitchy it really is on your machine.
https://bugzilla.mozilla.org/show_bug.cgi?id=1407536
Honestly, I’ve given up on Firefox.
On Android the performance is comparable to Chrome, but I can never get used to some of their UXes (pressing the X button doesn't clear the address bar, instead closes it). Also no seamless page opening with Google search app (although I suspect they can't do much about this)
AFAIK the implementation of CustomTabs is open and the API agnostic so Firefox could provide FirefoxCustomTabs if they wanted to.
I can have the same amount of tabs/workload in Chrome and my laptop copes fine with good performance. If I do the same in FF (with likely less plugins installed) it often spikes to 100% CPU usage, chews through my battery like nothing else and just gets bogged down/slow. I _want_ to use FF but can't sacrifice having a burning hot laptop with terrible battery life on my lap.
The recent post about real time pose estimation was a pretty example of that - pretty unusable in FF, runs great in Chrome (on Windows at least)
https://storage.googleapis.com/tfjs-models/demos/posenet/cam...
https://news.ycombinator.com/item?id=17283525
The only real world pages that sometimes crawl to a halt on FF are some Google pages like the old Gmail and Youtube.
Usually people who say that are complaining that Gecko is missing some bleeding-edge DOM API that was pushed out by Blink.
If you think that Firefox/Gecko doesn't have any modern tech, then you're not looking closely enough.
The only major thing I wish it did better is if it conserved battery power as well as Safari does. If I want to maximize my laptop's battery life I don't really have a choice other than Safari on my MBP.
Are there any (recent) tests you know of that compare battery life between safari and the other major browsers?
https://bugzilla.mozilla.org/show_bug.cgi?id=1407536
Otherwise in general, the native browser is usually the best bet for battery life. They render differently and have tighter integration overall as you noted. Firefox or native is the way to go, and if native browsers keep improving, even Firefox should watch out.
BTW, haven't used Chrome/chromium in years unless it was for testing purposes.
PS: full time Firefox user here but anyone who follows IBM and apple knows how "tech journalism " works.
Ironically, or maybe intentionally, Google services are the only thing degraded. Google even refuses to give you the regular home page. Thankfully you can just get an extension that fakes Chrome useragent on the offending pages, but it's the darkest pattern I've ever seen
This article is like every other product release article, plus the recent revival effort from Mozilla is news worthy.
This article was seeded by a PR firm. It’s not directly sponsored by Mozilla. It’s been professionally “placed” in the NYT.
Then again, I don’t know why I’m arguing with you. Judging by the style of your comment, I’m guessing you didn’t really go back to check if the article contained any sort of sponsorship notice (because, seriously, what real person would do that and then come back and leave a comment?). Your account is probably owned by the PR team that got the article placed and then pushed it to the top of HN. #AmIRight
Different machines, this one is: Intel I-6700-HQ 2.6GHz 4-core/8 logical 16GB RAM...
2. does it still send DNS data to third parties for analytics. https://www.ghacks.net/2018/03/20/firefox-dns-over-https-and...
3. Am i still required to build Pocket when I build Firefox? are we still forbidden from removing pocket?
4. Is telemetry still shipped on by default? https://www.theregister.co.uk/2017/08/24/mozilla_considers_m...
Then no, its certainly not time for me to try firefox.
So, genuine question, what do you use, then? GNU IceCat?
Chromium is just as open as Firefox.
https://github.com/Eloston/ungoogled-chromium
The browser would also have to backward compatible, and handle mal-formatted and bloated code written by god who knows. There are features that are the antithesis of privacy and disabling some of the features breaks quite a few sites. Sadly, the web is no longer linked documents, but has full blown apps. It's honestly freaking insane.
And thanks for answering and not just dismissing the question.
2. We are not sending DNS data to third parties "for analytics." We are openly and transparently conducting a DNS-over-HTTPS (DoH) experiment with a very small portion of users running Nightly test builds of Firefox. The experiment places a massive notification banner in the browser window with prominent buttons to disable or accept the experiment.
The details on DoH and how it improves both security and privacy are explained by Lin Clark in https://hacks.mozilla.org/2018/05/a-cartoon-intro-to-dns-ove... (HN discussion at https://news.ycombinator.com/item?id=17196415).
3. Firefox is Free Software, under an OSI-approved license. You're not "required" to build pocket, nor can or would we "forbid" you from removing it when you build Firefox.
...but you could also save yourself a lot of time and just toggle the "disablePocket" group policy: https://github.com/mozilla/policy-templates/blob/master/READ...
Still seems like a really bizarre thing to get hung up on when the extent of "Pocket in Firefox" is a toolbar icon and a section of the new tab page, both of which can be hidden with 1-2 clicks, after which they're completely inert. But hey, whatever floats your goat.
4. Yes, we gather a limited amount of telemetry by default. Unlike other modern platforms, this is always available for your own inspection (see about:telemetry and https://telemetry.mozilla.org/), and you can opt-out with a single checkbox. We actually reduced the telemetry we gather last year: https://medium.com/georg-fritzsche/data-preference-changes-i...
I think you are being defensive and completely underplaying the effort required by saying “1-2 clicks”.
Dude, these things should be opt-in and not opt-out.
There's a setting to prompt "are you sure" before closing Firefox, but it doesn't actually work. I really miss this feature (which works as expected in Chrome) since unless I'm turning my machine off a pretty much never close my browser intentionally
Shortcuts in text input also work differently than every other program on my computer. Command-backspace deletes the whole text field, not just the text before the cursor. ^t doesn't transpose, ^k behaves differently when text wraps, etc
I believe this feature gets disabled if you enable the feature that automatically restores your previous session when Firefox is opened.
https://addons.mozilla.org/en-US/firefox/addon/disable-ctrl-...
Command-delete works as expected for me (in this HN comment box), deleting only the text on the current line before the cursor.
I have been using it since 2006 (Firefox 2.0). Back then, the presence of the Firebug extension for Firefox made me switch to Firefox as my primary browser.
Two years later, Chrome arrived but Firefox remained as my primary browser because I wanted to continue using the Vimperator extension for Firefox that allowed mouseless browsing with Vim-like key bindings and commands.
Firefox still remains as my primary browser with Chrome being an additional browser that I use sometimes. These days, I use the Vimium extension for both to use Vim-like key bindings. I occasionally use Safari and qutebrowser too.
There is so much wrong with this bundling but yes, it's probably still better then IE...
I hope they fix that regression.
High on my list is for the extensions to work on about:* and addons.mozilla.org pages.
I like to use Vimium everywhere and 'Dark Background and Light Text' to also work on the preferences.
Also I really miss DownThemAll... I tried a couple of replacements, but they didn't for instance download all books from a Humble Book Bundle.
There's actually an about:config setting that fixes this already. I put a helper in Tridactyl for it called `:fixamo` [1], but you can easily do it manually by setting
"privacy.resistFingerprinting.block_mozAddonManager" = true
and
"extensions.webextensions.restrictedDomains" = ""
(Unfortunately, this does nothing about about: pages or the PDF viewer which used to work but was "fixed" :( ).
[1] https://github.com/cmcaine/tridactyl/blob/8c49d26340cdc0db3c...
Just changing "extensions.webextensions.restrictedDomains" is apparently not enough.
I also tried :fixamo in tridactyl, but it just returns: "# Error: Attempt to postMessage on disconnected port"
I was using version 1.13.1pre1454 with FF 60.0.2.
---
That's a particularly unhelpful error, sorry about that. It usually means that Tridactyl can't access the native messenger. Did you install it with `:installnative`? What is the output of `:native`?
I did not, but now I have.
:fixamo now doesn't return any error but still doesn't work.
Adding "privacy.resistFingerprinting.block_mozAddonManager" manually and setting it to true did the job. Thanks!
So far I like Tridactyl, especially the useful commands, but since it seems to be incompatible with other addons that overwrite the NewTab page. I think I will have to disable it.
I also miss the smooth scrolling experience with j,k from VimiumFF, jumping lines like Tridactyl does makes me loose the point were I left of. I might check back how the progress is going tough.
You need to restart the browser afterwards. Sorry, people usually find the command through the help page where it has all of these caveats.
We have smooth scrolling, you just need to do `set smoothscroll true`. It's a bit rubbish, though.
You can often fix the new tab page by disabling and re-enabling the new tab add-on you want. Every time Tridactyl is updated it will steal the page back, so I'd suggest that you don't use the beta releases and instead use the normal ones.
We could reasonably easily provide a build that didn't use the new tab page (we occasionally get people who feel very strongly about this) but I'm afraid it's quite low down my list of priorities. If you're feeling up to it there's an issue open: https://github.com/cmcaine/tridactyl/issues/534
But I won't hold it against you if you use Vimium. It's very well polished : )
:help fixamo
Does not state that I need to restart firefox afterwards. [1]
I did restart firefox after I installed `native` and before I run :fixamo. :native also returned that the it was installed correctly before I run :fixamo.
[1]: fixamo
in about:config via user.js so that Tridactyl (and other extensions!) can be used on addons.mozilla.org and other sites.If you're saying that a restart didn't make fixamo work, would you mind filing an issue so we can try to fix it?
Edit: sorry, I should have been clearer: fixamo edits a file which is only read at Firefox startup, so you need to run fixamo and then restart.
No I haven't tried that. After I run :fixamo and that didn't work I just opened about:config and change those settings myself without restarting firefox.
Edit: Since I now continued to test tridactyl I found that this newtab overwriting function gets pretty annoying from a usability perspective. When I start a new Tab with [Ctrl]+[t] and immediately type my desired address part of it will be overwritten when 'about:blank' is inserted. I guess you want people to use just a binding to :tabopen, but I do like my options. Also since :tabopen do not contain my bookmarks I prefer using the address line.
Huh, that is annoying. I hadn't noticed that. There's not much we can do about it though, short of just disabling the new tab page.
The lack of bookmarks in tabopen is annoying, I agree. We rushed to get the completions out with Tridactyl, and did it shoddily and haven't bothered to rewrite it yet.
There is bmarks which just completes from bookmarks, but that opens things in the current tab.
I can also use it on android with ublock etc. Chrome hasn't even been installed in my main computer in maybe a year, firefox works for everything I need, including google apps for business mail and youtube.
My biggest reason for using it is the fact that it is open source, the other is that ad-block software actually block calls to adds not just the display of them as in chrome.
Looking at the past market share we're clearly the minority
I could be getting my history wrong, but from what I remember Firebug was absolutely mind-blowing.
That was equally kind blowing to me.