Slashdot Mirror
Google Chrome 27 Is Out: 5% Faster Page Loads
An anonymous reader writes "Google on Tuesday released Chrome version 27 for Windows, Mac, and Linux. The new version features a big boost to page loads (now 5 percent faster on average) as well as significant updates for developers. The speed improvement is thanks to the introduction of 'smarter behind-the-scenes resource scheduling,' according to Google. Starting with this release, the scheduler more aggressively uses an idle connection and demotes the priority of preloaded resources so that they don’t interfere with critical assets."
Come on, USA! Catch up to the rest of the world.
Loaded 27 onto my laptop. It was so fast, the computer launched itself out of the house at FTL speed and is now tweeting from somewhere around Alpha Centauri.
Guess I'll replace it with a Chromebook.
CPUs are magnitudes faster today than they were 10 years ago. Why is it that pages still take seconds to load? Go back 10 years and they still took the same amount of time. Why?
Only the State obtains its revenue by coercion. - Murray Rothbard
Is "'smarter behind-the-scenes resource scheduling,'" a codeword for 'not loading huge fucking flash objects from shitty overloaded ad servers'? Because that really helps with load times...
The hideously poor performance that I observed had nothing whatsoever to do with Chrome or the browser, the problem was that in order to paint a simple page, my browser was also sent to the following hosts: a.fsdn.com, b.scorecardresearch.com, ad.doubleclick.net (47 times), fls.doubleclick.net, ajax.googleapis.com, www.google-analytics.net, libs.coremetrics.com, edge.quantserv.com, js.bizographics.com, ad.yieldmanager.com, r.twimg.com, and several connections to facebook and twitter, which are really puzzling since I have no facebook or twitter account. After about 3 minutes, something in the world of TCP/IP finally closed a couple of the doubleclick connections and the browser painted the page!
In that case, it wouldn't provide much of a benefit over what I already use: an extension that enforces a click-to-play policy on plug-ins. Such extensions go by names such as Flashblock.
Chrome is optimized for Comic Sans. Other fonts are for pretentious hipsters.
Its memory usage that is such a great problem for me, not really the issue of CPU time. If chrome is constantly cuasing disk caching because of the enormous memory usage, that is going to cause massive speed degredation, which is far greater than any 5% decrease in CPU time by an algorithm. I wish Chrome had a feature for not storing uncompressed copies of image if they are off screen and would fix the massive memory holes. Really no reason a browser should use more than 5-10 MB of RAM per open tab.
5% actually makes a huge difference. "Latency matters. Amazon found every 100ms of latency cost them 1% in sales. Google found an extra .5 seconds in search page generation time dropped traffic by 20%." [link]. These statistics would not be true if the average Joe would not notice them. He notices, he just wouldn't phrase it as "this site was 100ms slower than usual so I didn't buy from it."
Sure, if you like Google knowing what you're browsing. I just dumped Chrome after several years for Firefox.
It's too easy to use Google for everything.
But wait...it's a client speed up optimization. All in all it would cause the same amount of traffic just reordered to start pages render sooner.
As for Joe Human, his reaction time is at about 20ms. Hence @5% improvement would be theoretically "noticeable" in a 2 seconds page load. But unnoticeable if Joe Human would have to observe it relative to 2 seconds total. Likely even with a stop watch Joe H. would be in "error area". And 100ms would be an improvement on a 20s load which would challenge patience of any Joe H.
From improvement description:
"The primary drivers are: preloading images sooner, more aggressive use of idle network time, dynamically changing resource priorities, deprioritization of preloaded resources, and reduced bandwidth contention among images."
From TFA:
"when you add up those saved seconds across all Chrome users, it totals to more than 510 years of people’s time saved every week."
Now that's convincing argument.
I admit that squeezing that 5% of already well optimized code is a great achievement (technically) that may be head-deep in the law of diminishing returns (practically). But G$$gle can afford it.
4wdloop
Links (elinks I think is the package name) is a console browser with some CSS layout support (unlike lynx when I replaced it).
Wow, sent an e-mail as suggested when clicking on "use classic" banner, and got a fast response that addressed my msg
If you want a Google Chrome like browser I would recommend Chromium, which unlike Google Chrome is open source and doesn't track you as much as their proprietary product. You will miss out on some of the extra features available only in Google Chrome, but most of it should be the same.
Then you might want to use Chromium instead. Chromium is open source unlike Google Chrome, and doesn't include the same tracking system that Google adds to its proprietary product.
Page used to load really fast in 1990s in mosaic then, Netscape as long as you had something like a T1 connection. Now, funnily enough, the software layer involved in serving dynamic content and all the xml, third party sites and what not network calls the browser has to make before actually counting the page as loaded make it seem like the software layer has become the bottle-neck. This sounds silly to me, maybe we over did a bit?
Everything I write is lies, read between the lines.
As for Joe Human, his reaction time is at about 20ms. Hence @5% improvement would be theoretically "noticeable" in a 2 seconds page load. But unnoticeable if Joe Human would have to observe it relative to 2 seconds total. Likely even with a stop watch Joe H. would be in "error area". And 100ms would be an improvement on a 20s load which would challenge patience of any Joe H.
Reaction time is how long it takes you to process a stimulus and initiate a physical action in response - it's not how long it takes you to notice something. Joe Human's sitting there wondering wondering why he's letting this site steal the precious moments of his life away from him loading ads for stuff he doesn't care about. He might not have time to hit the Cancel control before the page finishes loading but he knows his time's been wasted.
One of the penalties for refusing to participate in politics is that you end up being governed by your inferiors - Plato
It's a continuous curve. If it took 10 minutes, few would stick around. If instantaneous, everybody would, and only quality of results would matter. For each fraction of a second delay in-between, you get a nice sine curve of people giving up.
This was the source of an interesting analysis that forcing airlines to give free baby seats would cost more lives than it saved by driving a small fraction of people to take the car instead. You can't get something for free, and the additional dimes added to everybody else's tickets would, incredulously, drive some, subconsciously, to take a car instead, crash, and die, as car travel is much more dangerous. For each penny of increased price, you get a diminishing number of buyers. With an annual granularity of hundreds of millions, this could be tens of thousands of people at each penny at the price points.
Meanwhile, free baby seats (where babies could be strapped in instead of being held) would have only made a difference in two known cases, and one of those the baby survived anyway by flying into an overhead bin.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
Also, I think Google Chrome is first browser to implement click to play flashblock in browser, and that is a good thing.
Settings -> Content Settings -> Plug-Ins and select "Click to play". You can also make exception like PDF reader to allow always.
I think Google and FB and others like them have a lot of blame to share for the web needing a 10X fatter pipe to get the same speed: if every freaking page didn't have to talk to Google Analytics, send your cookie to FB for tracking etc either before (likely) or during page load perhaps you could actually enjoy the content you are there for in the first place on a slow connection. Now you need the fast pipe just to get all the preamble out of the way to all parties interested.
I think the proper person to blame for embedding GA and FB tracking is the webmaster, not Google and Facebook.
True but claiming to save 5% of load time by making a browser while at the same time marketing products that slow down the page load in the first place seems kind of circular.
Well back in the 1990's the common web page was text with a few hyper links, and if you were really fancy you had a picture.
The bottle neck was the speed of the line.
However html has transformed from a way to displaying documents, to more of an application platform.
Complain if you like about it, but it is here to stay, and modern heavy html has solved a lot of problems. Such as platform independent programs, universal access to a program, easy deployment, etc...
Yes we have sacrificed speed for convenience, but I think it is worth it.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.