Microsoft Joins The Khronos WebGL Working Group

[Marentis]

Which stats are you going by then, because everything I've seen has uniformly had IE over 50%. Take what Ars is reporting, for example, in the global desktop market:

http://en.wikipedia.org/wiki/Usage_share_of_web_browsers

Wikipedia lists 4 different sources, 3 of them seeing Chrome on the first place.
I'm not saying that this has to be correct because the majority says so but it's not as likely that 3 sources are wrong and only one gets it right than vice versa.

[Slartifartblast]

MS

When supping with the devil, use a long spoon.

[Passso]

Wow, just because it's Microsoft you guys are immediately going off into conspiracy theory bullshit land. How about we stop and realize that WebGL has become the standard for 3D rendering in the web browser, and that maybe just maybe... Microsoft wants a say in how a standard they're having to implement evolves. Besides if Microsoft really wanted to disrupt WebGL all they would have had to have done is not implement it, and instead just offer D3DForWebClient.

Like they did for Samba, Exchange, Java, Html (IE), .rtf etc.?

Same strategy as ever :

Hello! How should I help?
Oh man I'm so sorry my version is just buggy...
But wait, I have 90% of the market and majority has the "Holy Truth"
Now the world must change to my version... but they have no source and if they reverse it I put them in jail
WOOT GG

By the way :

Microsoft joining the WebGL working group is an interesting step and hopefully will lead to some open, welcomed contributions from the company

Lol of the year, sorry

[smitty3268]

http://en.wikipedia.org/wiki/Usage_s...f_web_browsers

Wikipedia lists 4 different sources, 3 of them seeing Chrome on the first place.
I'm not saying that this has to be correct because the majority says so but it's not as likely that 3 sources are wrong and only one gets it right than vice versa.

Isn't it fascinating how different the numbers can be just by changing the sites you look at and how you count?

First, I'm going to completely ignore w3counter, because claiming that Safari makes up 16% of desktop users is just utterly ridiculous. Obviously they are focusing on some Mac-centric sites, or something.

Wikimedia is only focusing on their sites, which is obviously not going to be the same as the internet at large. Also, most of them are just counting page views rather than unique users - so it's going to heavily favor power users who open up tons of tabs and refresh sites constantly, while ignoring the light users who look at a page once and then go. I'd say that heavily favors Chrome at the expense of IE.

I'll just quote one of those comments from Ars, and note that NetApps numbers seem to reflect what i see in real life (minus the really old IE stuff, which i assume is mainly coming from China) and let everyone choose to believe whatever they want.

Every other statistic source tells different numbers, because practically none of them count users, but page loads (a method, which can obviously lead to heavily skewed numbers, depending on site usage patterns), and because neither of them correct for sampling bias, which is a standard procedure in statistics to get representative results from raw (and thus unavoidably biased) measurements.

Net Applications does both (ie. it counts unique users, not page loads, and also corrects sampling bias), which is why - contrary to your belief - their numbers are actually the ones that can be trusted most, and to reflect reality as close as it just gets.

Also see: https://yourlogicalfallacyis.com/bandwagon Ie. a lot of other statistics showing different numbers and/or similar to each other is neither proof for those statistics being correct, nor for the one or few outlying being wrong. Actually, they could be all correct in their own way, because - as explained above - they all measure different things, and because most of them are not representative for the whole world, but only for a very specific subset of it, if at all.

[Paul-L]

Isn't it fascinating how different the numbers can be just by changing the sites you look at and how you count?

First, I'm going to completely ignore w3counter, because claiming that Safari makes up 16% of desktop users is just utterly ridiculous. Obviously they are focusing on some Mac-centric sites, or something.

Wikimedia is only focusing on their sites, which is obviously not going to be the same as the internet at large. Also, most of them are just counting page views rather than unique users - so it's going to heavily favor power users who open up tons of tabs and refresh sites constantly, while ignoring the light users who look at a page once and then go. I'd say that heavily favors Chrome at the expense of IE.

I'll just quote one of those comments from Ars, and note that NetApps numbers seem to reflect what i see in real life (minus the really old IE stuff, which i assume is mainly coming from China) and let everyone choose to believe whatever they want.

Yet still doesn't explain how it counts "unique users", just says it does without explaining the metodology; it also does not say whose sites their statistics are coming from.

Wikimedia (with wikipedia) is one of the most visited sites in the world, so it actually makes a good representation of the Internet at large. In w3counter, Safari can make up 16% because it counts Safari on top of iOS platforms, there is good reason why you should count any mobile device, because they are becoming increasingly popular, and they also sum up to the whole internet traffic.

NetApps reflect whatever you want, but 3 sites says it is wrong (and two are dedicated solely to tracking the market share usage), 3 vs 1 can explain why NetApplications can be biased or wrong. Also, unless Ars explains how they get their statistics, if they are not from NetCounter, it also means it could be biased they way you mentioned, "Ars only focus on it's site".

[asdfblah]

http://netmarketshare.com/faq.aspx#Methodology

I wonder how many of those hits are bots...

[smitty3268]

Yet still doesn't explain how it counts "unique users", just says it does without explaining the metodology; it also does not say whose sites their statistics are coming from.

Correct, although that's generally true of all of them.

Wikimedia (with wikipedia) is one of the most visited sites in the world, so it actually makes a good representation of the Internet at large.

Perhaps, but isn't it pretty western-centric? I really don't know, but i suspect it's a better measure of US/European traffic than true world-wide count. Correct me if i'm wrong.

It's also likely highly biased towards home users, as i doubt a lot of people at work are going to be visiting it. Or at least not as heavily as they would be from home.

In w3counter, Safari can make up 16% because it counts Safari on top of iOS platforms

That 16% was marked specifically under a chart labelled "Desktop Browsers", so if they are including iOS in there I trust their measurements even less because they are blatantly lying. Well, maybe somebody at wikipedia just screwed it up...

there is good reason why you should count any mobile device, because they are becoming increasingly popular, and they also sum up to the whole internet traffic.

For normal sites, ok, but it's silly when discussing WebGL. Nobody wants to run complicated WebGL apps on their phones.

NetApps reflect whatever you want, but 3 sites says it is wrong (and two are dedicated solely to tracking the market share usage)

No, 3 sites are obviously testing very different things, as discussed earlier. I think NetApps is measuring the correct info, and the others are just doing the easy (and wrong) measurement.

3 vs 1 can explain why NetApplications can be biased or wrong.

It doesn't explain anything, that's just an observation mixed with an opinion by you.

Also, unless Ars explains how they get their statistics, if they are not from NetCounter, it also means it could be biased they way you mentioned, "Ars only focus on it's site".

It says they are from NetCounter. I think they have shown their own site stats before, and it heavily skews away from IE as you expect on a tech-oriented website. I'm fairly sure Phoronix has < 50% IE users too.