(I'm on Yosemite.)
Google Sheets and Google Docs have been horrible on anything than Chrome for a while as well, and I guess they have no incentive to make that situation change. Firefox is dealing with it better than before I think, but I won’t hold my breath for parity to ever come.
If your cpu does not have a webm decoder in hardware, it has to decode them in software which is inefficient.
P.S. macOS 10.13.6.
Go to chrome://flags, search for "UI Layout for the browser's top chrome" and pick "Normal". Restart and enjoy the old design.
Running my mouse down the list in about:config feels like I'm drunk.
This is on a brand new maxed out i7 MacBook Pro.
Now if I could just get it to be reasonable about search engines...
It's pretty bad. Firefox cuts my battery life in half doing nothing.
It can be mitigated by setting
gfx.compositor.glcontext.opaque = true
in about:config, but the fix is invasive, and has taken years to get prioritized and worked on. I've seen reports that it should finally be done in FF64 or 65.EDIT: It seems AdBlock Plus has now come out with a new plugin that conforms to Apple's new API.
I believe all of the people who report this, but it's not something I experience. Here's my usage pattern:
- 50+ hours in a browser each week
- Regular use of both FF and Chrome with very occasional use of Safari (I'm a developer of web apps)
- FF has been my primary browser for 15 years or so
- Used on a variety of Macs (2006, 2008, 2011, 2013, and 2015 MBPs with 8-16GB RAM)
- Running uBlock Origin in all browsers FWIW
...so, anecdotal, but pretty extensive.
I never had major problems with FF in the pre-Quantum days. Chrome always felt snappier than FF, and Safari felt snappier than both, but the differences were not huge and FF was "fine." And now since Quantum, FF is on par with Chrome for me in general.
The one time FF feels like a pig for me is on Google-owned web apps like Gmail.
For many years, Gmail and Gsuite apps were lightning fast on FF. But in the last few years it has gotten slower and slower on everything but Chrome. Hmmm, wonder why.
- uses 100% CPU even when idle
- frame rate drops to 2-3 per second for 5-10 seconds when switching tabs
- causes other programs to not work properly because it is using too much CPU
I have tried uninstalling the browser and re-installing OS X but it didn't solve those issues. I read somewhere it is a bug that pops up when you have display resolution scaling on and I think it applies to me because I have two 4k monitors set at 2x scaling.
uses 100% CPU even when idle
That sucks.Anecdotally, I've never seen this, even when I had 1 external 4K and 1 external 2K connected. The whole OS was kind of sluggish at that point, but that was definitely asking a lot out of a puny integrated 3-year old laptop GPU so eh.
Just did a quick informal test with my current setup:
- MacOS High Sierra.
- 2015" MBP with no discrete GPU.
- Single 4K external monitor, 3008x1692 scaled resolution.
- uBlock is off.
With only a single blank tab open FF uses 0% CPU as expected. I opened three tabs: Facebook, Espn.com, and MSNBC.com. There was a brief flurry of CPU activity as the sites loaded, and FF's processes are now idling comfortably at 0-1%.
I repeated the process with Chrome and results were similar. One difference is that Chrome spawns 10x as many processes, but they seem to consume less memory each.
In both FF and Chrome, there's a palpable (500 or 750ms?) delay when switching tabs. In my experience a lot of Mac apps behave that way at scaled 4K resolutions.
I repeated this informal test in Safari and it "feels faster"; less CPU spike on initial site load and switching browser tabs feels close to instant.
I don't know if Google has gone too far with the A/B split testing and are picking winners solely on a most ad revenue metric or if they abandoned UX testing altogether. I suppose it could be a bit of both. May be their decision making AI is secretly optimizing Google in a destructive direction.. who knows.
Makes me wonder if companies aren't just slowing down their competitors..
Also the parent comment is about Google Voice, which is not a browser and therefore is not constrained by Apple's restrictions on 3rd-party browser engines.
All this unfixed for years now.
I'm trying to switch to Firefox but it certainly doesn't feel as snappy as Chrome when doing dev work. I do like some of the tools though. Still undecided.
I've never been a huge extension user (ad blockers and some development tools), so moving to Safari was easy. The big thing I missed was favicons, and those have been in the technology preview for awhile now.
I may explore FF for development, but I'm not sure there is a need since I've already relegated Chrome to a specific task.
I have noticed a few quirks in Safari that I have to work around. Like not rendering things exactly the same way as chrome/firefox. Bit of a pain, but there is never anything that's really "broken".
I'm puzzled as to why people seem to dismiss the "built-in" browser. It's the fastest, smoothest, best integrated, and least power-hungry browser on the platform. I regularly use all three major browsers (for testing, I write web applications) and I consistently switch back to Safari for all my non-special browsing.
I keep switching away from it due to lack of favicons and lack of extensions. I don't personally mind paying for a dev license to release an extension, but it definitely takes a toll on the extensions that get released.
Safari has a new extensions model and will be deprecating the legacy one in a year. Hence the push to migrate to newer extensions.
I made an extremely simple extension using the new model around 2months - before I try migrating an older "actually useful" extension. But theres just no way its possible to migrate.
Apple doesn't care about browser extensions now, it seems they want everyone to move to the new model, so they can collect revenue on sales on ad blockers.
:(
The web is terrible without a serious adblocker and the available alternatives (paid or not) are unable to take away the same amount of junk. People will simply start using another browser.
https://developer.apple.com/documentation/safariservices/saf...
More specifically, if you can inject scripts I don’t see why you couldn’t call window.close()
This is a tradeoff Apple made valuing privacy and performance over user freedom. The new content blocking API is safer and faster.
To use Safari, you pretty much need to decide that performance / battery life are more important than anything else which only describes my needs when I need to milk my battery when it hits 10% with no electric outlet in sight.
This is how to enable it: https://lifehacker.com/how-to-enable-safari-favicons-on-mac-...
I do use Chrome during development though.
Safari lets you peek at the previous page, and makes it very obvious that you are about to go back, without any performance hit.
It also comes with reading mode, which chrome doesn't. Even with an ad blocker, many pages are still layed out in ways that can only be summed up as reader hostile.
So for reading and moving between pages, I feel like it beats both ffx and chrome. So I disagree with your "more important than anything else" part, but agree that performance and battery is among its top features, and if you don't care for that, safari does become a hard choice.
But it needs more than that when my battery already lasts 6+ hours under heavy use and my laptop is plugged in almost all day anyways.
Perhaps you can see how a reading mode that we already had extensions for in other browsers and sexier prev/next gestures damn it with weak praise.
As I said, it provides the best performance, experience and battery life.
We have no particular stance against it, though I agree with the other poster that WebGL is more the future.
edit: reading all the comments in this thread, geez, I finally feel like I'm not a stupid user. :) The issues are real!
"Faster tab switching in the majority of cases for our macOS users"
"Improved the reactivity of Firefox on macOS"
Unfortunately, while Firefox has made significant leaps the last couple of years, it's still not a viable alternative to Safari for me. It doesn't feel particularly native (neither the "chrome" nor form controls etc.), and the UX is still clunky compared to native Mac-first apps.
One particular way in which Safari absolutely beats Firefox is the address bar. Most of the autocompletions that Safari provide are exactly what I want. As an example, if I type "mapq" into Safari, it suggests mapquest.com (a site I've never visited before, so this is not based on anything from my history). For Firefox, it suggests a Googles search for "mapquest", which is just stupid. Aside from suggesting web sites, Safari is superb at showing good results (Wikipedia is often a top hit), and it is awesome at providing autocompletions from my bookmarks and history.
Firefox also seems a bit lost when it comes to innovating, and keeps coming up with weak concepts that don't get any tractions; Personas comes to mind. Container tabs are a similarly interesting idea with a weird implementation that requires you to micromanage your tabs, which I certainly don't want to, and provides a "techie" solution to something that should just be invisible and default (i.e. all web sites should be "contained").
All websites contained would be an interesting default. With ways to “re-open current tab in container [X]” and “merge current tab with container [Y],” it would involve less micromanaging. That’s still not invisible, the container process and UX warrants plenty of thought.
The other is what could be described as "focus" or "modality". A lot of people use windows as a poor man's workspace. For example, if I'm researching where to travel, I'll have a window open with tons of tabs — Google Flights, Kayak, Tripadvisor, Booking.com, all jammed into one "workspace". If I'm comparison shopping for one specific thing, there'd be Amazon, Jet, Etsy, eBay, etc. This is how people tend to work: Windows separate modes, tabs separate units of focus within that mode. And yet there's no browser that supports such a way of working. We have to accept that people now "live" within the Internet, and need to support different modes. Rather than think of the browser as a shell, why not think of the browser as a world in which you can open up different types of interacting with that world? For example, a Google Doc tab is not something that works very well as a tab. It's a document, representing its own modality. Slack is an app, also its own modality. A "workspace" is something else again. And so on. These are all "applications" in a classical sense, yet work within the context of a user.
I'm not suggesting a ChromeOS, just that what we think of as "the web" could be rethought in terms of different variations — ways to work — with the same thing.
Also, in terms of privacy, why are we still using cookies? Why are we "logging in" and "logging out" of web sites and having password managers that automatically fill out form fields? Why can't the web site negotiate its session with the browser ("I want an identity for gmail.com, please give me one"), which can already know who I am?
You might want to take a look at the Firefox Multi-Account Containers Add-on, it lets you set domains to always open in a given container, which cuts down on the micromanagement a lot once you get it "trained". [edit: This is not well-advertised, it's accessed through the context menu (right-click) on a page when you have it open in that container]
https://addons.mozilla.org/firefox/addon/multi-account-conta...
> i.e. all web sites should be "contained"
It's possible to enable this for most intents and purposes with "1st party isolation", which came from the Tor browser. Some info here:
It's still too difficult and cumbersome to use. I might want to have one tab with gmail signed in but another to make anonymous searches on Google, both are on google domains so i have to manually fiddle with the containers again. Then you click a link in an email or search result and you don't really know if this tab should be in the mail-container, in the default container or in the domain-matched container.
It's a tough problem, and I agree the grandma test is a good standard, sometimes it can't be easily done.
Firefox, on the other hand, only shows you searches. It might offer "Isaac Newton" as a search suggestion, but that just triggers a search. It has no knowledge of other sources of information (beyond bookmarks and history).
For me, I find them very useful. It's convenient, for example, to be able to have two side-by-side Gmail tabs open. e.g. one for work email, and one for personal email.
But I do see your point that they shouldn't be necessary.
Works for me when I flip it on https://i.stack.imgur.com/0ZfvX.png . I keep it off because I don't like having my every keystroke in the address bar sent to Google.
Firefox only does search suggestions. In other words, anything you select from the suggestions just go to search results. Safari does that, too, but it also provides actual results. Very often the top hit is the right thing.
Firefox it is, and they have done a great job with the new version.
After this incident, I switched to FF Quantum permanently. It has come a long way and honestly I feel more at home in FF after a day than I did in Chrome after everything they've tacked on the browser.
Having read these threads, and personally knowing people in the Chrome team that read these threads, I am almost certain that Google will backtrack this. I know more than a dozen developers that switched from Chrome to FF Quantum this week.
I for one am glad that it exists, I love Pocket, the only annoyance I have is that the Firefox integration is not as complete as the Chrome add-on, probably because Firefox users keep bitching and moaning about it.
It's great that you like Pocket. I think it's a bit rude to act like it's unreasonable to have qualms about how it was added (certainly pre-Mozilla owning them) or to be confused about it's inclusion while things like RSS reading/folders, (a not-uncommon browser feature) are removed.
This entire conversation wouldn't need to happen. There wouldn't be unhappy users. There wouldn't be "another side" calling criticism "bitching". They'd get active, enthusiastic user feedback, etc, etc.
Whenever suggestions pop up for migrating from freaking Chrome, which if we are honest is the new IExplorer 6, to Firefox, somebody has to mention freaking Pocket. And it's tiresome and I don't think it is legitimate.
The presence of Pocket's integration in Firefox can't possibly be the reason for why somebody doesn't want to migrate from Chrome.
This describes their approach to basically every new thing they come up with, and as a long time exclusive Firefox user it's the one thing I really don't like about it. I find it to be a great daily-driver browser but every now and then they'll try to cram some new garbage down your throat.
At least they don't surreptitiously start siphoning your data with without asking first, though.
Edit: As for the screenshot thing being poor, I agree that labeling the upload button 'Save' was quite a faux pas but at least the icons are pretty clear. The 'download' icon is the same as the download manager icon, and the upwards arrow pointing to a cloud is a reasonable representation of 'upload'. I might even use this thing someday.
Chrome initially was faster than Firefox because it lacked features. This is a common pattern if you've been in this business long enough. Firefox itself was originally the fast-but-cut-down version of Mozilla. Like Firefox, Chrome quickly gained features and bloat and became just as slow as Firefox and everything before it.
Recent attempts I've been making to help teach web developers about the dangers of referers, before they add more tracking pixels or leak reset password links get deleted by Mozilla technical writers because
" We don't think it's appropriate to have a red warning banner at the tops of the pages. That kind of design element is one we try to avoid on MDN unless it's highlighting the very first thing you need to know about the item, which we don't think it is in this case, although we do appreciate that it is important."
Great idea, let's leave the wet floor warning as a note at the end of the corridor.
See here https://developer.mozilla.org/en-US/docs/Web/HTML/Element/im...
I certainly trust the MDN team to understand how to arrange their content to match their audience.
I also believe web developers should be more informed about the privacy & security issues of their work. The content you tried to add was verbose without any technical detail or links, and the MDN revision history isn't a great space for content discussion.
Have you tried filing a content bug? It's much easier to converse on bugzilla than thru edit battles.
https://bugzilla.mozilla.org/form.doc?bug_file_loc=https%3A/...
I feel like Firefox is back in the same boat it was in the 90s, but this time fighting Google Chrome instead of IE. But it's a bit different since you have to explicitly download Chrome. It's a bit different since you see an ad to download it right when you hit Google.
Firefox is a truly great browser, they really made amazing improvements on all fronts in the last years. And I say this as someone who never used Firefox as their daily driver until half a year ago or so.
Exist a way to make it work? (ie: Maker a goblal zoom and a per-domain/page)
[1] https://www.bleepingcomputer.com/news/software/chrome-and-fi...
whenever a page loads all the elements move around for a second or two before they settle in their right place. that would crack me up if I had to put up with that all day
>So many sites don't work correctly in Firefox
I've not experienced that.
Example: https://i.imgur.com/ZfHnNmX.png
Maybe one day, mozilla. Maybe one day.
I've been seeing people repeat that for the past 5 years. It turns out to be false though.
It's actually a selling point for me, I just can't stomach all the material design craps that Chrome forces on me where a page lost like 70% of usable space to useless whitespace.
At least for me, on Linux, it’s always been opt-in, and plenty of the versions of Firefox packaged by distro maintainers get rid of it all together.
[1] https://bugzilla.mozilla.org/show_bug.cgi?id=1362065#c57