Hacker News new | past | comments | ask | show | jobs | submit login
Bye, Chrome: Why I’m switching to Firefox (fastcodesign.com)
91 points by ProAm on May 30, 2018 | hide | past | favorite | 44 comments



Just to offer a sample point:

Switched to Firefox Quantum when it came out, and never looked back. Never had a problem with any website I used on any of my machines (which include the i7 16GB RAM work laptop, Core2Duo 3GB home desktop, and wife's old Thinkpad still running XP). Everything runs faster than Chrome, especially on the older machines, and I can comfortably open more tabs.

Granted, my top 4 sites are probably HN, Reddit, Facebook and Youtube, and the first two aren't very demanding of the browser - but the links from these sites go all over the place, and I simply never had to launch another browser ever since.

I find Firefox more feature-rich, and use the Reader View very often (especially at work, where visual clutter is the last thing I need on my monitors). MRU tab switching order out of the box is also great to have.

</anecdata>


> Article: Firefox doesn’t always work better than Chrome

I assume the author isn't on Quantum because my experience (developer edition) matches yours. Firefox is mind bogglingly fast, which is precisely why I dumped Firefox for Chrome when Chrome first launched. My extensions seriously bloat memory usage; memory is cheap, time is priceless.

Features and privacy are great, but the performance is simply so incomprehensibly far ahead (and we don't even have Servo yet). The omnibar is also default now and is more reliable than the extension - that was one crucial Chrome holdout for me.


The reader view is really something. One of those things I necer thought about til I used it, now I miss it whenever I'm in chrome.


The point of Chrome is not that it can spy on you, it is a vehicle that allows Google to control the evolution of web standards. Still a good reason to use something else though, it’s very important that there are competitors.


Hmm, there is something contradictory to me in the funding of Firefox. If it's funded by big search engines paying to be the default for Firefox, then Firefox's revenues are coming from Google or Baidu's revenues. These search engines' revenues come from advertising and tracking, which is what Firefox is meant to be against. I'm not sure how meaningful it is, but it seems like a strange loop to me.


It is somewhat of a loop, but it doesn't really play into things, because:

1) The search engines pay money and receive something in return. That's a closed off transaction. It's not like Mozilla needs to keep them happy, so that they come back. Sure, it would be nice, if more than one search engine continues to exist, so that there's competition for that spot, but that's about it.

2) Search engines are not actually that dependent on tracking. The ads can be placed based on the search query, and a browser can't block them from tracking that search query.

3) There's competition between search engines, so even though they share a common interest in being able to track users, it's not like they'll all agree to give up that spot. When one search engine is pissed off enough with Firefox to not want that spot anymore, another one will gladly fill in.

And 4) there's already other politics that prevent Mozilla from blocking all tracking. Webpage owners need to be kept happy, too. If you block tracking, those make less money off of your browser, are less likely to spend money to fix or develop things for your browser and so on. With so many users voting for Chrome, too, there's little incentive for webpage owners to spend anything on supporting Firefox already, so Mozilla can't be too aggressive with taking tracking away from them. And I feel like that's much stronger than the search engines' stake in Firefox.


There's a lot of strange loops like this is free and open-source software generally. For instance much free software is partially funded by proprietary software companies who want to use it internally, or is open-core but not fully open-source. GitHub itself is a walking contradiction of license philosophies for instance.

Not quite the same thing, but there's (almost) always a weird interplay and tension whenever something actually needs to get funded or make money.


GitHub's on-prem enterprise offering is certainly a contradiction but I don't think their hosted service is. It's code they wrote running on servers they own/rent. FOSS software might come with an ethos of publishing useful things one makes but it's certainly not an obligation.

I will admit that there is issue regarding non-free client-side JS but a lot of projects tend to ignore that.


switched to firefox a few months ago but i keep going back to chrome everytime i need to watch a video online. Firefox videoplayer just doesn't work. The video lags very frequently and the cpu is always overheating.

Only when that matter is fixed will i be able to completely ditch chrome.


Yeah, I do video production work and Firefox is useless when it comes to video. If it even can play half the things, it's going to suck up a lot of computer power to do it.


Hmm that's odd, I have the exact opposite issue. For me and a number of people I interact with, the chrome video player will abruptly freeze the video while the audio continues to play. It only really seems pronounced on laptops with a second monitor attached though.

Seems the same issues plague both browsers and one will just arbitrarily work better for some people.


Same. I actually use FF for most browsing and I haven't really noticed a huge problem with the occasional youtube video. But I watch a lot of twitch streams on the box hooked up to my TV at home and Chrome seems to be the only option for getting 60fps at the moment.


I haven't had this experience, but internet bandwidth tends to be a serious bottleneck here, so I rarely feel game enough to exceed 720p.



This article resurfaces every time a new version of Firefox gets released. I'd love for someone more technically inclined to explain how the new iteration of Firefox is better than the current build of Chromium for Linux (66.0.3359.117).


(I use FF developer edition and chrome, but not chromium)

I find Firefox Dev tools to be a bit nicer,esp. the network tab. Firefox also has a few experimental features that are okay, but the fact that they're experimenting is worth something (eg the JavaScript scratchpad feature).


The biggest selling point for me is that Google doesn't have their fingers all over Firefox.

As to your question, there are lots of articles comparing the new FF to Chrome..

https://fossbytes.com/firefox-quantum-vs-google-chrome-compa...

https://www.zdnet.com/article/firefox-quantum-vs-chrome-are-...


You didn't get a whole lot of responses, but honestly, I'd bet money that you would get even less, if you posed the question the other way around. Because how is Chrome/-ium better than Firefox?

To give some rough image from what I know about the browsers and what I've heard other people say:

Performance: Roughly equal, Chrome seems to still be more consistently fast, which Mozilla is still cleaning up after that big architecture change. Mozilla also has more in the pipeline, which I'm not seeing as much from Chrome.

RAM use: Firefox is still considerably lower here, even though the Quantum iteration needs more RAM. There's little motivation for Google to have users use software outside of their web browser, they can't display ads or gather data in those, so there's little motivation for them to not eat up all of the RAM.

Customizability: Clear win for Firefox. You can drag UI elements everywhere you want, color the whole UI with extensions as you like or even fuck around with CSS to alter its look.

Extensibility: Chrome still has more extensions in numbers, mainly because Mozilla does not allow telemetry in add-ons (unless the users opts in). Firefox extensions are more capable, though.

And Chrome's extension store is a dumpster fire, filled with malware. Mozilla vets extensions with actual human beings, which Google doesn't consider an option.

Security against script-kiddies: Also roughly equal. Chrome has to a minor degree still a more secure architecture (sandboxes each tab individually most of the time, whereas Firefox sandboxes them in groups of how many cores your CPU has, for performance reasons), but Chrome on the other hand has some glaring idiocies here and there.

For example Chrome's autofill fills in data in all input fields on the page at once, meaning that it will also fill input fields that you as a user can't see, so you might send off your address to a sketchy site without knowing about it. Another example is them shipping the WebUSB-API in a form that made Yubikey Neos completely exploitable, as webpages could literally just connect to the Yubikey on their own and read out the secret, bypassing the U2F API that Google had built into Chrome.

Security against Google and in extension US intelligence agencies (and in extension non-US intelligence agencies): Well, you can probably guess by yourself. Chrome Sync by default uploads your browsing history and such to Google's server in decryptable form. So, Google can access it, and because of US law, the US intelligence agencies can grab it from Google's servers, too. And because those intelligence agencies are friends with other intelligence agencies (Five Eyes etc.), those likely have your browsing history, too. Obviously depends how much you consider these a threat, but it's certainly not in your favor for these groups to have your data.

You can bypass that, by enabling end-to-end-encryption, which Google requires a second password for, so it's not necessarily an argument when you know about it, but that brings us to what the article mentions, too.

Defaults: Firefox Sync is end-to-end-encrypted by default. Only one password needed. Firefox's Private Browsing mode ships with Tracking Protection, no (potential malware) extension needed to block trackers, not that Chrome even allows extension to run in Incognito mode.

And these are just superficial examples. We're talking about millions of lines of code, tens of thousands of design decisions. In one case made by a non-profit, that always tries to protect users while trying to not piss off webpage owners too much, in the other case made by a company that always tries to satisfy its own needs in the hope that users don't notice or don't complain too much. And again, millions of lines of code. Lots of shit goes under the radar that no journalist reports about. Even in Chromium.

This is for example a project that tries to fix Chromium and it admits that it's an uphill battle: https://github.com/Eloston/ungoogled-chromium


Thanks. I'm a big fan of Chromium's UI because outside of enabling flags I don't have to customize it to get it where I need it to be.


Nice article for laymen. On HN I want something more technical, more in-depth.


I don't necessarily want something more technical, but for some reason I was anticipating that the reason given would've been more of a technical one, like "Firefox is way better with WASM" or something of that nature.


Firefox still haven't fixed a privacy bug of keeping a record of all your closed tabs in private mode. You can all try by opening several tabs, going to any sites of your choosing and closeing them. If you invoke, reopen last closed tabs, all of them returns.

This simply doesn't happen in Safari and Chrome. I reported this issue some time back and someone at Firfeox said this is not an issue.

So if I left my private window open, but with no tabs, someone can always retrace my steps even though I've closed all my tabs. How's that for privacy?


That's not a bug, it's just a different behaviour. I can see the value in being able to reopen accidentally closed tabs, and if you want to get rid of that history, close the window.

Given that Chrome keeps cookies for the window lifetime, I'd say that the fact it throws away history gives a false sense of security - they may not get exact history, but they can get some data still.

If you want the behaviour you are describing, it sounds like the feature should be "private tab" not "private window".


Chrome preserves your cookies across tabs, so it's not really "private" in the sense that you seem to want, either. I wouldn't say that Firefox is less private. I'd say that they made a design decision to allow you to re-open tabs in privacy mode in case you unintentionally close them. In both Firefox and Chrome, you need to close your private window if you really want to erase your data.


To further your point: if you log in to you Gmail account in a private window, close the tab, but keep the window open, someone can open Gmail in that window and will be still logged into your account.

This goes for both Chrome and Firefox.


Doesn't leaving Incognito mode open in Chrome retain cookies until you close it?


Yup.


I've noticed this, although it's weirdly inconsistent, which is possibly worse. Sometimes it hit C-S-t and get the tab I just closed, sometimes I get something from 20 minutes ago.

It'd be one thing if a quick C-S-t within a few seconds could restore something you accidentally closed, but this behaviour is utterly incomprehensible.


Firefox has two hot keys- C-S-t for undo closed tab, and C-S-n for undo closed window.

It’s slightly different from the chrome behavior where both of those actions are treated as the same, and both are on ctrl+shift+t. I switched chrome->firefox several months ago and I’m still not used to it. 99% of the time C-S-t does what I want, that other 1% it’s a different key.


Interesting, although I'm talking sometimes I close three tabs and literally immediately try to restore and get like, the first tab I closed instead of the 3rd sometimes. I don't think I'm just pressing the inappropriate key. But maybe I'll try C-S-n next time this happens and see if it resolves the issue.


I like it that way. If closed by accident can be restored.


One of my email accounts is through my ISP. I can read mail with Firefox, but not read it, the error being

TypeError: this._getDoc(...).body is undefined

I reported this to the ISP some months ago, and was assured that they are working on it. Obviously, it is not Mozilla's fault that this doesn't work, but it suggests that some organizations don't think they need to accommodate it.


Edit: "but not write" it.


I stopped using all Google products in my personal life quite some time ago (including Gmail, Maps, etc.). I gave the latest Firefox a shot though and found it super slow on my Macbook Pro compared with Chrome and Safari, so stuck with Safari.

Would love to return to using Firefox but even without any extensions installed I experience regular lag. Anyone else?


I did for a while. Around version 58 it started performing better again. Then something in v60 caused a bunch of CPU spikes again, so I'm using Developer Edition, which is on v61 and doesn't have that issue.

I would say 99% of the time, Firefox is great for me. There are occasional sites that I open in Chrome - either because of video choppiness or JS going out of control on a page.

However, I would love it if it were less of a gamble when upgrading versions.


You likely won't come back to this, but in the event you do - thank you. V61 dev edition is working an absolute charm for me.


I hear you. I'll give v61 a go.


I don't want to say this, but you'll be back. Putting all cons aside of using Chrome, Google is actually doing a very impressive job which will make everyone considering this switch to go back in no time for some reason or the other.


Can you expand on that? I switched from Chrome to Firefox a few years ago and I've been very happy with that decision.


My Windows 10 machine just did a huge update. Ever since, Chrome closes within a few seconds unless I let it Run As Administrator. Uh, no thanks. I tried various suggested fixes and even did a complete remove and reinstall. So, no more Chrome.


Interesting. I think, Chrome doesn't do sandboxing when you run it as administrator (because it'd be pointless security-wise), so maybe it specifically crashes when trying to spawn those sandbox processes.

Or it's a simple matter of it not having file access permission to a file that it needs and running it as administrator obviously grants it access to that file.

Not that this information is likely to help you fix that issue...


I even tried manually deleting all the files in the user-specific directory; made no difference.

There comes a point when the time investment has to be viewed as a sunk cost, and move on.


I agree with the sentiment of the article. It’s been this way for a while though. Chrome is perceived as the more powerful browser by the masses. I love Firefox for sure, unfortunately it has been hard to convince normal internet users to make the switch.


Show them the Reader View feature, which is there out of the box.

Way too many websites (including CNN) turn a humble article into something barely appropriate by plastering ads, animated ads, video ads, and other irrelevant, bright, colorful mess around the text. Reader View takes one back to text.

Also appreciated by someone who doesn't want to signal "I'M READING A TABLOID" with their screen by simply opening a CNN link.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: