[clicks]

What you say is probably true for IE, AOL browser, etc., but Chrome (when it came out) was really just a better browser. It was significantly faster than any browser of its time, the new interface to increase vertical space was wonderful, the auto-update was a great feature, the tab sandboxing was great. I had been using Firefox for a long long time before I jumped ship to Chrome.

But now I'll try Firefox again because of this post. :)

[StavrosK]

> It was significantly faster than any browser of its time

How was it faster than Opera? I have never seen another browser give you the previous page instantly when pressing the back button. Chrome needs to start its spinners for a few ms and then reflow the page, Opera just does it. I'm hoping that, with the move to WebKit, I can switch back to it.

EDIT: I just checked, Firefox does this too. It didn't, at the time that Chrome was released, though.

[gluxon]

Firefox and Chrome both do this. It is part of the memory caching, which many people disable without knowing the consequences. There were many news articles back around Chrome's release that advised "speed-ups" by disabling cache to free up memory.

And the speed people refer to is JavaScript execution. There is no doubt that V8 was much faster than competitors a few years back.

[StavrosK]

It is enabled for me, pages are served from memory (it doesn't request them from the network again). It's just that Chrome is slow at it, while Opera/FF are instant.

[kbenson]

Maybe this is due to their infinite caching? There was a article on HN a while back about how it impacted performance after a while[1]. Turns out checking a large cache of files (apparently with a poor choice of data structure backing the search) for visited links ends up being non-negligible a year into browsing.

[1]: http://news.ycombinator.com/item?id=5281540 Unfortunately, the original article looks to have been removed.

[StavrosK]

I thought I saw that being debunked by a Chrome developer, I remember him saying that this was changed a while ago, and now the cache has a maximum limit. Clearing the cache doesn't seem to do anything for this problem, though (the page still reflows). It doesn't take more than a few ms, but it's annoying when other browsers do it instantly.

[emn13]

The aptiverse page making that claim was removed from the web rather quickly, so either way, which is a little suggestive.

[codeulike]

javascript - exactly - before Chrome v1 no one was even talking about speeding up javascript in the browser space. Chrome v1 blew everything out of the water with that. And js speed has a big impact on everything these days, pretty much

[BrendanEich]

23 Aug 2008: https://brendaneich.com/2008/08/tracemonkey-javascript-light...

01 Sep 2008: http://googleblog.blogspot.com/2008/09/fresh-take-on-browser...

TraceMonkey was in the works since late Spring 2008. Apple was also doing more advanced JS performance work before Chrome launched.

V8 had the world-class VM team and at least two years lead (Lars Bak went to Google in 2004; I met him in August 2006 when he was definitely working on V8), so it indeed was fastest at the usual benchmarks, but not by the sometimes-asserted 3x factor.

Maciej Stachowiak of Apple and I were both noting back then how V8's advantage seemed more like 1.3x at the time, but I don't have performance charts from Sep 2008 at hand. Perhaps someone reading does.

V8 got faster over time, as did other engines. Again, it's an excellent piece of work and tops by many measures, but not all -- see http://kripken.github.com/mloc_emscripten_talk/#/17 for two large benchmarks of three where SpiderMonkey beats V8 currently.

/be

[mh-]

Speaking as a dev, but from a user's POV, the thing that made jump to Chrome on release day was the immediate recognition that a single bogged down tab did not impact the responsiveness of the chrome (heh) and other tabs noticeably.

I remember, vividly, in Firefox: I would middle-click on a Slashdot link to load it in the background while reading the current one. My focused tab would begin to hesitate and sometimes altogether freeze for several seconds. In Chrome, only the spinning tab would be affected by their bloated DOM.

[Yoric]

In case somebody hasn't noticed, let me emphasize that the author of the previous post is Brendan Eich, the creator of JavaScript.

[codeulike]

I stand corrected - thanks Brendan!

[qu4z-2]

For me the speed-up I liked in Chrome was the UI responsiveness, and start-up time. And it seemed to just load pages faster. My current browser of choice is surf from suckless.org (although it's pretty limited) because it somehow just manages to load pages in a quarter the time of Chrome/Firefox (on my computer/internet connection, anyway). I'm sure it'd lose in js benchmarks though.

[Shorel]

When Chrome launched, Opera was faster than Chrome and everything else.

It took about one year and a half of JavaScript engine development for Chrome to surpass Opera in performance.

Since then, Opera has not been the faster browser anymore.

[papsosouid]

>Firefox and Chrome both do this.

Try opera, there's a reason he's pointing out that it is different.

[khuey]

Firefox has had this caching since Firefox 1.5.

[Mahn]

For me tab sandboxing was basically the reason I switched from Firefox. For years I thought it was normal for the browser to slow down and become increasingly unresponsive the more tabs you had open. It was frustrating, but you had to accept it was your "fault", you can't possibly have 30 tabs open and expect the browser to be cool with that, I thought. Then Chrome came, and I was SO blown away by being able to have dozens of open tabs without the slightest change in the overall performance of the browser, I didn't even care I couldn't use many of the extensions I had on FF. I couldn't help but recommend everyone I knew to do the switch aswell after realizing I would not be back to FF.