Slashdot Mirror
Mozilla Tests Firefox 'Tab Warming' (bleepingcomputer.com)
Catalin Cimpanu, reporting for BleepingComputer: Mozilla is currently testing a new feature called "Tab Warming" that engineers hope will improve the tab switching process. According to a description of the feature, Tab Warming will watch the user's mouse cursor and start "painting" content inside a tab whenever the user hovers his mouse over one. Firefox will do this on the assumption the user wants to click and switch to view that tab and will want to keep a pre-rendered tab on hand if this occurs. "Those precious milliseconds are used to do the rendering and uploading, so that when the click event finally comes, the [tab] is ready and waiting for you," said Mike Conley, one of the Firefox engineers who worked on this feature.
Let's bloat the browser down EVEN MORE rather than making something efficient that people want to use...or cleaning up the UI to make it clean and not confusing.
Firesux still leads Chrone on trashiness.
A realize it is probably a different team but they could spend some time improving the Android version -- it is too damn slow. I really don't have a problem on Windows and if they're trying to eek out milliseconds in UI response there, maybe put the effort to shaving seconds off of the Android interface.
In other words waste memory, bandwidth and energy in a pissing contest that started with google compromising on security by disabling ocsp, and Mozilla already lost. Who is asking for this? No one, that's who! People want back their extension API!
... what could possibly go wrong?
I actually tried out Chrome for a bit after Mozilla pulled it's Mr. Robot stunt, but came back to Firefox after noticing how much better it performed than Chrome, which was somewhat surprising to me. I had assumed they were at performance parity. At this point, I think Mozilla has the top performing browser by metrics that tend to matter in real life. The one I notice the most is that Firefox's UI rarely stutters when loading a page, while Chrome hitches and hangs in short bursts, making things feel sluggish. I think that makes a huge difference in the perception of speed and performance.
Tabs switch almost instantly for me, and that's on a nine year old PC with a moderately slow internet connection. So while I'm glad Mozilla is looking at important things like performance (instead of yet another pointless UI revamp), it almost seems unnecessary at this point. Has anyone else noticed any sort of delay when switching tabs?
Irony: Agile development has too much intertia to be abandoned now.
Opera just cached the rendered version of all open tabs. This is part why it were the fastes browser of its time.
And they even cached the rendered version of pages in the history. A faster back button is not possible.
"...do the rendering and uploading..." -- Does he mean uploading of the user's information to the website, or does he mean automatic downloading of site elements if you so much as put your cursor near the tab? This sounds like not only a waste of bandwidth and resources, but a security and privacy problem as well. Imagine the fun a malicious actor could have with these features.
bitcoin mining will warm up your tab before you type the url - brilliant!
My browser better not start uploading anything when I hover over a tab.
Maybe they should invest in a new PR person with those millions they have in cash. He doesn't even know the difference between an upload and a download.
Just how many tabs are people keeping open at a time that this is considered a good feature? I mean, at home or at work, I only ever have maybe a half-dozen or so tabs open at once. Whereas an old roommate of mine used to have dozens of tabs open at the same time.
But I don't recall him ever complaining about clicking on a tab and it not rendering immediately. It was more of a "which tab was it again"? problem as he looked through the ones he had open.
Mr. Hu is not a ninja.
I only use AMD, so nothing's going wrong here. My internets are a bit slower, but at least they are safe!
Escher was the first MC and Giger invented the HR department.
There is no way the user can get to those pages without typing a long boring url or using a bookmark.
Or do something simple like press Ctrll-Shift-T to open the most recently closed tab that you accidentally closed — a nice feature saving you from looking up a long, boring URL for a page that you didn't have bookmarked.
Whether it leaks memory or not is a different issue and you may be right... but the basis for your argument is incorrect.
I want less predictive behavior not more Mozilla. Firefox doesn't need this anymore then I need a pizza guy to show up when he thinks I want a pizza.
I'm gonna go way out on a limb here and suggest that this predictive option will be.. optional. We should be able to turn it off. Perhaps by moving the mouse in the general direction of the "settings" button.
I will continue to use Waterfox instead, and also recommend Pale Moon and Basilisk
Why? They're just older, slower versions of Firefox. They all depend on Firefox for upstream development. Might as well just use Firefox.
Tab warming is related to tab switch animations, not actual tab loading. It works only with your existing, already-loaded tabs.
I only use AMD, so nothing's going wrong here.
There are (at least) two attacks based on speculative execution - and disclosed at the same time.
Meltdown is Intel-specific. Spectre runs just FINE on AMD - and some high-end ARM cores, too.
(It's beside the point in this case, though. Speculative loading and rendering/re-rendering/activating animations of a page when the mouse hovers over the tab will leak the same information regardless of whether the browser is running on Intel, AMD, ARM, or whatever.)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
For decades, browser scientists warned us this was coming. We had simple static pages, but no that wasn't enough for us. We needed dynamic content. We needed javascript.
Suddenly we had all this free computation. It was exhilirating. We could make hampsters dance and punch the monkey to win. But that computation had a cost. We kept burning more and more CPU cycles.
Browser scientists raised the alarm. All those cycles produced heat. At first our fans dissipated it, but they couldn't keep it. Eventually the heat crept into the rest of the system. They told us it would lead to tab warming. We just laughed and loaded more instagram kittens.
Who's laughing now? Our tabs are getting so hot they overflow into other programs. Their behavior is increasingly erratic and unpredictable. Now we have rogue sites mining cryptocurrency in them. Face it, our tabs are damaged beyond repair, unable to sustain simple online email anymore.
Like Icarus, we flew too close to the sun. We have no one to blame but ourselves.
Democracy is two wolves and a sheep voting on lunch.
Mozilla Tests Firefox 'Tab Warming'
And it's probably a feature affecting the entire browser, so now I have to worry about Global Tab Warming ... (sigh).
It must have been something you assimilated. . . .
Well, since you asked. I remember the days of pop-under ads which had activeX drive by code, or focus stealing, or javascript heavy code that locked up the browser. I remember racing to close new tabs before they could load and crash my session. So lets give the bad guys a head start.
Does this mean whenever I start moving the mouse around Firefox is going to madly start running a bunch of javascript, spinning my CPU up to full?
Ye gods. Delay in switching to a new tab is not an issue.
I use <Ctrl>+<TAB> to switch tabs like a normal person, you insensitive clods!
Google Earth runs in Chrome but not in Firefox 57. What replacement for Google Earth do you recommend?
This has nothing to do with tabs loading. It's about rendering what's already loaded.
Good browser performance already depends on predictive behavior, such as kicking of DNS lookups early when you hover over links. I don't know what you have against predictive behavior, but the alternative is a slower browsing experience.
It all depends on the complexity of the page. It is not difficult to find pages that are slow to render, especially if you are fullscreen on a 4K display, say. Maybe you don't visit such pages, but lots of users do.
Friends don't let friends use Chrome.
Uploading custom emoji to Discordapp.com works in Chrome and in Chromium but not in Firefox 57, where clicking the upload button has no effect. It's even worse in Firefox ESR 52 on Debian 9 "Stretch", with many actions lagging and pegging one core of a Core 2 Duo CPU for one or more seconds, sometimes blanking the whole page for a second. Discord staff has a habit of closing issue reports in Firefox to the effect "Works for me in Chrome. Could you try it in Chrome?" Would you recommend stopping using Discord over this incompatibility? If so, then to what text and voice chat platform with built-in support for retrieval and search of older messages and attachments do you recommend that a community using Discord migrate? (IRC lacks voice chat, retrieval and search of older messages, and attachments.)
Or get back to actually doing your job?
What did you mean by this? A break room computer at work is not the only environment that restricts installation of native applications. Another is a computer at a public library. A third is a computer using an uncommon architecture for which the application's publisher has not compiled the native application, such as GNU/Linux on ARM instead of x86-64. From Google Earth on a Pi? - Raspberry Pi Forums:
I wonder if people at Mozilla realized that this approach will cause an increase in CPU usage triggered by mouse cursor movement, and this will ultimately result in an increase of world power consumption.
There's a "whoosh" playing in my other tab, but fortunately I've disabled the sound on hidden tabs.
Escher was the first MC and Giger invented the HR department.
I tried visiting the first result in Google Search, and it gave me error 502 Bad Gateway. Not a good sign.
Second try: "Jitsi" on Wikipedia. The article states that Jitsi is a native client application, not a server. Therefore, it would need to be used with some sort of server software and a hosting account. It states that Jitsi is available for Linux (by which I assume it means X11/Linux), Android, Windows, and macOS, which would exclude members of our community whose primary device is an iPhone, iPad, or Chromebook. It states that supports IRC and XMPP for text, SIP for voice, and proprietary services of major IM networks (MSNP, OSCAR, and YMSG). Which of these protocols A. have a server available to the public and B. support server-side storage of older messages and attachments?
you cannot install chrome on a raspberry pi
You are technically correct in that Google Chrome is not available for Raspberry Pi. Instead, one would access Google Earth using Chromium, which is the same thing as Google Chrome except without components under a proprietary software license.
What problems have you had getting Chromium to run on your Raspberry Pi?
Stop futzing around with less than necessary trivia. Keep IT SIMPLE Stupids, and keep it efficient.
Efficient also means cpu memory footprint, and computer cycles.
Leslie Satenstein Montreal Quebec Canada
I like Tree Style Tab. Solves all your problems. There are a few other vertical-tab extensions, try them all. I don't see how people can live with 30 tabs across the top :-(
Guess I'll move to Vivaldi, the only modern browser that shows tabs on the side (hopefully in an awesome tree) that old XUL Firefox extensions used to allow.
For many versions, Chrome even had a built-in tabs-on-side feature, but they removed it ages ago.
The only sensible way to manage many tabs is on the side! Or multiple rows on top/bottom if you are in portrait mode!
Sad that WebExtensions can no longer modify the user interface.
R.I.P. power-browsing the web.