Slashdot Mirror
Opera Denies Microsoft Edge Battery-Saving Claims (thestack.com)
An anonymous reader writes: According to the makers of the Opera browser, Microsoft's recent claim that its Windows 10 Edge browser is more power-efficient than Chrome are erroneous. Running its own tests with Opera, Edge and Chrome, the company finds that Opera runs 22% faster (with a battery life of 3hr 55m) than Edge (3hrs 12m). In Microsoft's own tests, Google's Chrome browser was the first to completely exhaust the battery, closely followed by Firefox and Opera. In May, Opera added a power-saving mode, but any advantage it can be verified to have in the energy-efficiency stakes may be more due to the native adblocking feature it introduced this year.
Me, I'd rather sacrifice some runtime so long as I don't use a Microsoft or a Google product.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
Any test needs to include uBlock Origin or the equivalent. Even Edge supports it now, I read. Otherwise any test data will be corrupted by random advertising altering the content and wasting as much battery power as it can.
Also, it's not clear that they even tried to match the laptop batteries. Maybe some where lower capacity than others, due to manufacturing variations and lifetime degradation.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
This is a false argument introduced by Microsoft in order to divert attention away from just how much Edge sicks for its main function: web browsing.
Battery life is indicative of overall efficiency. A browser that's more CPU efficient and more memory efficient will likely draw much less battery, too, as it's the battery that provides the energy for the CPU and memory. That's why measuring the battery usage is a good measure of efficiency: it collectively represents all of the actions the browser takes.
This graph from the article about the testing is really interesting:
https://winblogs.azureedge.net/win/2016/06/browser-power-consumption-tests.png
The interesting part isn't Edge. The interesting part is actually how Firefox is the worst performing browser in the test.
We often hear from Firefox supporters that Firefox is more efficient than Chrome and other browsers. They talk about how Firefox is supposedly faster (that is, uses less CPU) and how Firefox supposedly uses less memory. They vehemently deny that other browsers, and especially Chrome, are more efficient.
Of course, these claims from Firefox's supporters contradict experience. People who try Firefox and Chrome, and who aren't driven by ideology, readily admit that Firefox feels a lot slower than Chrome does.
Now thanks to these tests we have more evidence to show that Firefox is less efficient than Chrome and other browsers are.
The real question is, when will the Firefox supporters finally admit that Firefox needs some really serious work to become competitive with the other browsers? We aren't talking about wasting time and energy on something like Servo, too. We're talking about improvements to the Firefox browser that an ever-dwindling number of users are using. When will they admit that Firefox has poor performance and that this poor performance needs to be fixed immediately?
Edge is an improvement over the new Chromium-based Opera.
The first time someone picks a browser based on power usage...lemme know.
Did you read what I wrote? I care about not running a program made by an evil company first, and about efficiency a distant second.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
How do they run these experiments? I don't see any deviation in the measurements. It'd be nearly impossible for the thing to run exactly for 3h55m every time. It should probably be given as 3h55m +/- X minutes. Also, do they swap laptops to eliminate innate differences in batteries between the machines? Say laptop A running Chrome dies after 3h00m and laptop B running Edge dies after 3h10m. Do they run the test again with laptop A running Edge and laptop B running Chrome and get the same results? Do they repeat the runs on different laptop models?
But instead they decided to work on hello and pocket, making your Gnome 3 and SystemD powered Linux box even slower.
The interesting part is actually how Firefox is the worst performing browser in the test.
Did you not even read TFS? Because:
Google's Chrome browser was the first to completely exhaust the battery
Which leads to:
We often hear from Firefox supporters that Firefox is more efficient than Chrome and other browsers.
It's clearly more efficient than Chrome, as per the results of this test.
I can't afford to be ideology-driven, as I'm a web developer and must test my work in all browsers. I'm comparing them side-by-side, day in and day out; yes, if you manage to wrangle all of Chrome's little sub-processes and add up their memory usage, Chrome is using more memory about half the time. Sometimes it swings in the other direction, and by about the same amount, so I'd say, honestly, on average they use about the same amount of RAM.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
I'm letting you know. Battery life doubled over Chrome.
"Science is the power of man"
AdBlocking does not only load pages faster, helps block tracking and other semi-malicious activities but it also saves battery time!
Ads are big revenue streams, someone has to bury this before the word spreads to the uninformed masses.
A 'singular oddity' is an event that cannot be explained and only happens when you are alone.
To respond to "Edge uses less power than Chrome", they compare Edge to Opera and declare Microsoft's statement false. Despite the fact that looking at the link shows that Edge does beat Chrome. That this build of Opera beats both should be the half-story, but no one is pointing out the reality of the test.
Who knew?
LOOK AT THE FUCKING GRAPH FROM THE MICROSOFT ARTICLE ABOUT THIS TESTING!
https://winblogs.azureedge.net/win/2016/06/browser-power-consumption-tests.png
LOOK AT IT! Look at the data! Don't go by what some shitty Slashdot summary says. LOOK AT THE DATA! We are scientists here. We LOOK AT THE DATA!
It shows average power consumption by browser. As the title clearly states, LOWER IS BETTER. What is Chrome's number? 2819 mW. What is Firefox's? 3161 mW. The other browsers are less than Firefox, too.
Firefox uses the most power on average. That means it will exhaust the battery quicker, regardless of what some fucked Slashdot summary says.
You're arguing against physics and math here! You say you're not "ideology-driven", yet you've clearly let ideology override your ability to LOOK AT THE DATA and perform an objective analysis of it.
Seriously, think about this! You're assuming the Slashdot summary is correct. Why the fuck would anyone do something that stupid?! The data clearly contradicts what the summary is stating. Holy fuck, this is Slashdot! The summary is probably the last place to find correct information!
Microsoft measured it with the do you want to upgrade to Windows 10 dialog open.
Yeah sure. It's like going from having a minor cold to getting AIDS.
The test they performed is pretty much invalid because it's such a narrow case... they played a movie until the battery died. I'd really prefer to see a test where they script browsing or something... so that there are a range of behaviors tested, rather than just a movie.
Which has more power: the hammer, or the anvil?
According to the makers of the Opera browser, Microsoft's recent claim that its Windows 10 Edge browser is more power-efficient than Chrome are erroneous. Running its own tests with Opera, Edge and Chrome, the company finds that Opera runs 22% faster (with a battery life of 3hr 55m) than Edge (3hrs 12m).
What a surprising result. It's strange, when I tested the browser I wrote myself, "Pseudonymium", using my own hand-picked test criteria, which I call the "Pseudonymo" benchmark, the results said that my browser was, and I quote "one point nine and one-third percent times more betterer" than all of its competitors.
It's almost like comparison tests administered by the producers of a product versus competing products in the same market segment are inherently untrustworthy.
I found in my experience Firefox uses far less memory than Chrome for displaying content, however Firefox has this tendency to not free up memory sometimes. I find after having both Firefox and Chrome open the entire day without closing doing various opening and closing of tabs alone, by the end of the day when you open the same content side by side in Firefox and Chrome, the former would use far more RAM.
But really none of that concerns me. RAM is cheap. Now Firefox seems to take longer to open tabs, grinds to a halt far more easily, and overall feels sluggish. Whether it is or not is not really relevant since I am not a benchmark.
At least it was about 5 versions back. I finally had Firefox crap itself during an update and the solution was to nuke my profile. So I did by switching to Pale Moon.
Firefox does definitely have a memory leak; for me it takes a couple weeks, maybe a month, to become an issue. In that same timeframe, I often catch Chrome using the same amount of RAM, or more, with the same tabs open (testing the same work on the same sites); eventually, Firefox does crap itself and nearly double its RAM usage, though I'm not sure if that's Firefox or one of my add-ons. I could disable add-ons to find out, but I don't want to be without them for a month or longer; it happens so rarely that it's not an issue worth my time to track down. I just kill it, reopen it, restore the session, and carry on.
For me, it only seems to get sluggish when I have a broken or poorly-coded add-on, or when it craps itself after a few weeks. I've also gone into about:config and turned off all the superfluous crap; and I'm running the developer builds, so that might have something to do with it, as well.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
Did you read what I wrote? I care about not running a program made by an evil company first, and about efficiency a distant second.
So you won't run EDGE because it's from an evil company. So you run Opera instead. On your WINDOWS 10 machine. Dude, credibility is not yours for the day.
Saw this in the sidebar. Opera still calling BS, 11 years later.
Opera: Firefox User Figures 'Inflated'
Has this been independently verified by an impartial group or are you just going off of what MS says?
The soylentnews experiment has been a dismal failure.
first, from your text looks like it is a huge difference, but it is inline with the other ones
second, edge is preloaded, chrome tries to have multiple process and be modular, so for this test might not need to load everything... but firefox is monolitic (mostly), so the at startup will load everything. If there is any flash loaded, even worst, as the flash in firefox is still a separate process and will always eat more cpu. Yes, all this are firefox problems, but... read below
finally, mozilla knows that for several years. they already have some code blocks in multi-process (but it was hard, as the base code was build as monolitic since day one) to sustain firefox until the new firefox (called servo) with rust is ready. That one will be more secure, modular since day one and be faster than any current browser. So yes, know problem, already being fixed
Servo should have the first public alpha (beta?) release in the next few months
Higuita
most "memory leaks" today are add-ons related... add-ons had too much access to the firefox internals and simple errors could cause problems.
mozilla tried to limit what add-ons can access and is trying to push then to external process, so it is easier to see where the leaks are coming.
Try to disable add-ons and restart firefox to check where the leak is coming
Higuita
Only if you like giving all of your search queries and links clicked to Microsoft, in addition to having a crap selection of addons and a fugly UI.
And while chrome has the same spying capabilities of edge, chromium does not.
I think I sufficiently explained why I did not, and will not, perform that bit of troubleshooting.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
It should be noted that Edge only runs on Windows 10, Edge is essentially a stripped-down version of IE, and IE
is integrated with the OS. In other words, Edge is probably entirely or mostly loaded as part of the system, just as IE has been since Win98. How does one accurately measure the power drain of specific system libraries as compared to 3rd-party software running *in addition* to those system libraries?
Then don't run gnome. Is there a distro that has KDE but not SystemD, though?
UBlock can't do these as well as (or @ all) hosts do 4 speed, security, & reliability:
1.) Protect vs. bad sites (past ads)
2.) Protect vs. fastflux botnet C&C's
3.) Protect vs. dyndns botnet C&C's
4.) Protect vs. DGA botnet C&C's
5.) Protect vs. downed DNS (reliability)
6.) Protect vs. DNS poisoned dns
7.) Protect vs. trackers
8.) Protect vs. spam payloads
9.) Protect vs. phish payloads
10.) Protect vs. caps
11.) Get past dns blocks
12.) Keep off dns request logs
13.) Speed up 2 ways (adblocks/hardcodes)
14.) Work on anything webbound multiplatform.
15.) Ez data edit
16.) Block ads more efficiently in cpu/ram/I-O use
17.) UBlock now uses hosts (no DNS benefits vs. dns issues) - poor imitation = "sincerest form of flattery"
Hosts = native vs. illogically "Bolting on 'MoAr'" & not ClarityRay blockable like addons.
APK
P.S.=> Hosts (1st resolver) do MORE w/ less in fast kernelmode & before slow usermode addons
Hosts ~3mb vs. UBlock = 64MB -> http://cdn.ghacks.net/wp-conte...
APK Hosts File Engine 9.0++ SR-4 32/64-bit http://www.bing.com/search?q=%...
Less power/cpu/ram + IO use vs. DNS/routers/addons/antivirus (slows you) + less security issues/complexity. Compliments firewalls (w/ layered drivers blocking less used IP addys vs. hosts blocking more used domains) & DNS (lightens dns load). Gets data via 10 security sites.
Ads rob bandwidth/speed, security (malvertising), privacy (tracking) + anonymity.
Hosts add speed (hardcodes/adblocks), security (bad sites/poisoned dns), reliability (dns down), & anonymity (dns requestlogs/trackers) natively. Hosts != ClarityRay blockable (vs. souled-out to admen inferior wasteful redundant slow usermode addons)
Works vs. caps & HTTP PUSH ads w/ firewalls.
Avg. webpage = big as Doom http://www.theregister.co.uk/2... & ads = 40% of the size.
APK
P.S. - Safe https://www.virustotal.com/en/... (Verified by Malwarebytes' S. Burn "I've seen the code & it's safe" http://forum.hosts-file.net/vi... )
I'm going off my battery lasting 8 hours instead of 4.5 since uninstalling chrome. It's been verified by my own experience.
"Science is the power of man"
My program filters above & beyond false positives lists my source already use + users can edit hosts easily themselves. TRY THAT WITH AlmostALLAdsBlocked OR UBlock, & good luck for regular users (that don't understand regular expressions BOTH use).
What hosts blocks, I can't touch & my lists are updated DAILY... this is proactive as it takes time for DNS to propogate changes worldwide & also for the malware makers to propogate their crap too... & users edits can make it immediately proactive easily also (they can't usually DO that with addons).
Stop spamming? I'm on topic - quit using profanity like a frustrated dimwit you show us you are with your erroneous little 1/2 truth to FALSEHOOD BASED lies reply... ok??
APK
P.S.=> Folks can immediately fix hosts themselves, they're EASY TO UNDERSTAND - regex isn't for most users & they'd have to wait out a fix... apk