Facebook Develops HTML5 Gaming Benchmark 84
An anonymous reader writes "A couple of Facebook engineers are developing an HTML5 gaming benchmark. They write, 'Two weeks ago Bruce and I released JSGameBench version 0.1. Today marks the release of version 0.2, a much faster and cleaner version. We continue to learn both from tightening the code and from the strong HTML5 community. Version 0.2 reinforces our belief in HTML5 as a strong, horizontal platform for games and highly interactive applications across the web.'"
Why is this tagged Chrome (Score:2, Informative)
Re: (Score:2, Funny)
When firefox kicks the shit out of it in the benchmark [facebook.com]
Because Chrome fanboys* are a lot like Apple fanboys* -- unconcerned with objective evidence of any kind. They are unable to appreciate Mozilla's accomplishments and strengths because they weren't done by "their team" and anyone not on "their team" is TEH ENEMY. Course, "their team" is sort of like the fat couch potatoes who see their favorite football team win a game and say "fuck yeah, we won" and the only correct response is "really? I didn't see you out there on that field". So they won, you shallow
Re: (Score:2)
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Not quite territorial. Tribal. There's your tribe, and the Others. And you have to protect your tribe from those Others....
Re: (Score:2)
hell even ie9 scores better. chrome outhta be ashamed.
Re: (Score:2)
Well, Chrome 11 is still in alpha though.
And from what I noticed, while Chrome seems to "lag" other browsers when you compare versions, but due to it's ridiculously fast development cycle it almost always catches up if not surpass it's competition.
People have been saying, IE9 has this, FF4 has that, since ... Chrome 7, and they both still aren't out yet.
Re: (Score:2)
No, Webkit is slow at compositing images, which I'm sure is what this benchmark reflecting. Firefox is really fast. Interestingly, this seems to be the case whether you use a bunch of HTML elements or canvas. The image rendering in Webkit needs an overhaul.
I found this out when I was making this [ontographstudios.com]. Check it out in Firefox or Opera compared to Safari or Chrome.
Re: (Score:2)
Interesting.
It's quite a lot more jerky in Safari and Chrome.
Looks like this is something the Webkit guys need to work on.
Chrome 11? (Score:2)
Didn't they just release Chrome 10 beta? It's still on the front page of Slashdot (at this time).
Re: (Score:2)
Re: (Score:2)
You can assume they're using a version currently under development.
All OSX browsers are really slow here (Score:3)
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
Chrome 10 is BETA; Chrome 11 is more like nightly/alpha.
Re: (Score:2)
Re: (Score:2)
My point was that Chrome 10 is still Beta on Windows. Absolutely agree with your main point that the other browsers deserved to have their dev versions tested too.
Re: (Score:2)
The fastest non-BETA browser is Safari 5 on OSX. Your point?
Re: (Score:1)
They didn't bother testing non-beta versions of Firefox or Chrome or *anything* on Linux or WinXP. This article is useless.
WinXP is like IE6 - people still staying with a decade old system are clearly not interested in being up to date, so out of scope for this comparison.
Cost to switch from IE 6 vs. Windows XP (Score:1)
WinXP is like IE6
No it isn't. There's a lot more switching cost involved. Switch from IE 6 to Firefox or Chrome, and the vast majority of web sites still work;* in fact, most work better. Switch from Windows XP Service Pack 3 to Ubuntu 10.10 with Wine from Software Center, and half your applications won't work. Unlike web sites, native applications are rarely tested by their developers on alternative implementations distributed as free software.
* Except possibly niche sites powered by ActiveX and visible only on your com
Re: (Score:2)
TFA is not the one you've linked there. TFA: [facebook.com]
Benchmarking
In order to talk about browser performance, we needed to standardize. We now have two machines that will be our testing machines:
For OS X: a MacBook Pro laptop, currently OS X 10.6.6, 4GB of RAM, 2.66 GHz Intel Core i7, and NVIDIA GT 330M with 512MB of RAM.
For Windows: a Lenovo T410s laptop, currently Windows 7 Enterprise, 4GB of RAM, 2.53 GHz Intel Core i5, and NVIDIA NVS 3100M with 512MB of RAM.
Both of these laptops are significantly less powerful than the Mac Pro the original tests were run on. In addition, the 3100M offers approximately half the performance of the 330M.
Additionally, even in the article you linked to, Safari was the fastest non-beta browser.
Re: (Score:1)
I have first hand experience that browsing is sub-par on OS X using any mainstream browser - Safari, Chrome, FF. I recently bought a top end Macbook Pro (Core i7 Nvidia 330M) and was very unimpressed with browsing performance - scrolling and page rendering mainly. I installed Win 7 and both Chrome and FF4 beta really shine in performance department - so much so that I am very reluctant to go back to OS X. FF4 it is explainable that the hardware acceleration may be making a difference but Chrome by default d
Re: (Score:2)
Isn't this also because FF4 and IE9 both use hardware acceleration on Windows? (+ The points made already about comparing dev builds vs. stable)
Re: (Score:2)
So? Pretty bloody obvious why it's worth noting: the task is wholly graphical in nature, so obviously hardware acceleration will give IE9 and FF4 an advantage on Windows (and FF4 is not HW accelerated on OS X). So it's kind of a no-brainer why there's a huge discrepancy
Re: (Score:2)
Yes, which is precisely the point. On Mac there is no way to easily do 2d hardware acceleration at the moment.
Re: (Score:2)
Re: (Score:2)
The problem is that Core Animation is too high-level. It wants to handle the entire animation itself, which makes it suitable for implementing CSS Transitions and CSS Animations (heck, the initial specs for those were basically "do what Core Animation does, because that's how we implemented it"), but not great for handling painting of web pages where you don't know what the web page will do next
To put it in web terms, Core Animation is closer to being like SVG while Direct2D is closer to being like canvas.
Re: (Score:2)
They used different hardware for Windows and OS X. It looks like the Mac is actually more powerful though, so the OSX browsers should score higher if they were equally efficient.
* For OS X: a MacBook Pro laptop, currently OS X 10.6.6, 4GB of RAM, 2.66 GHz Intel Core i7, and NVIDIA GT 330M with 512MB of RAM.
* For Windows: a Lenovo T410s laptop, currently Windows 7 Enterprise, 4GB of RAM, 2.53 GHz Intel Core i5, and NVIDIA NVS 3100M with 512MB of RAM.
Re: (Score:2)
That makes no sense, really. Why use two different machine
Facebook engineers? (Score:1)
Sanitary engineers? Advertising consultants? What engineering does fb do???
Re: (Score:1)
They serve a massive number of complicated pages. They have been hiring people away from Google and such.
Re: (Score:1)
complexity where? what is new in fb?
Google does some cool engineering but much of its work involves only routine IT skills.
Re: (Score:1)
Re: (Score:3)
If you're really so good at making sites that serve billions of page views per month
Creating a routine site which "serves billions of page views per month" is a routine IT skill. Hell, creating a routine site which serves millions of page views requires no expertise at all, and is something I'm sure lots of enthusiasts posting here have done before the days of point-and-click blogs and social networking sites. So, is Facebook routine?
has fairly low latency across the globe despite the huge volume
Fifteen years ago called. They want your cutting edge content delivery network research.
has marketed itself successfully to half a billion active users worldwide
No disagreement here. Facebook knows how to sell itself.
interface with hundreds of thousands of third party apps
Have you actua
Re: (Score:2)
Have you actually used the Facebook API? While this is the only "engineered" component of Facebook, it's (i.e. the Graph API is) basically a frontend to tables of personal information and junction/link tables between them. Again, the skill here is the routine deployment of an SQL database.
They've actually develop their own non-relational database (Cassandra, now an Apache project).
Re: (Score:1)
This is kind of a good point - no-one really cares that much if FBs pages break a little bit (as they often to, friends lists changing at random, comments not showing up, etc etc). You just refresh and (mostly) everything comes right.
I'm not saying that FB isn't impressive, just that they don't have to hit especially high standards of data consistency in order to be a success.
Re: (Score:3)
Think in terms of "software engineers" - 500 million users is a hefty workload for any single site.
HTML5? (Score:2)
Except for the canvas element, there's not much HTML5 to be found here. It's mostly about DOM manipulation using JavaScript and about fancy new CSS styles.
Re: (Score:3)
Except for the canvas element, there's not much HTML5 to be found here.
It's precisely the 2D canvas that makes HTML5 game graphics practical.
Re: (Score:2)
It's precisely the 2D canvas that makes HTML5 game graphics practical.
Wrong, canvas isn't suited for games in current browsers. Moving around divs is much faster and much easier (no need to do image loading manually, no need to handle redraws, etc). The only issue is alpha-aware hit testing, that's actually impossible with divs.
See this presentation [slideshare.net] by Paul Backaus (the guy behind jQuery UI and a javascript game engine that was bought by Zynga) starting on slide 31.
WebGL will change a lot there, when it's finally working in all major browsers (except IE of course).
Flipping (Score:2)
Moving around divs is much faster
One needs to horizontally flip an image when a character in a side-scrolling platformer faces the other way. Which browser can horizontally flip an image in a div? Otherwise, download size for sprite sheets doubles, as the server has to send both the unflipped and flipped versions of every cel.
Re: (Score:2)
> Which browser can horizontally flip an image in a div?
Any browser implementing CSS 2D Transforms (Firefox 3.5 or newer, Safari 3.2 or newer, Chrome at 7 and maybe even older, Opera 10.5 or newer, IE 9).
So pretty much anything on the market that supports canvas supports transforms too.
Re: (Score:2)
Re: (Score:1)
You asked which browser supported horizontally flipping and image, and when shown that virtually every modern browser supports it, you responded with the one that doesn't. It would have been more gentlemanly to admit you were wrong.
I apologize for being unclear (Score:2)
when shown that virtually every modern browser supports it, you responded with the one that doesn't.
I apologize for being unclear. I should have written the following: "Half of users don't use a modern browser. Good luck convincing them to."
Re: (Score:3)
Yes, but the original claim was that you have to do HTML5 games in canvas because you can't flip images otherwise. IE8 doesn't support canvas either, so if you're writing an HTML5 game with canvas you aren't targeting IE8 to start with.
Rationalizing Bad Code (Score:1)
Re: (Score:1)
Please clear this up (Score:1)