[austenallred]

28.3% on IE, 41% on Windows 7. Wow.

It's odd, being in the tech bubble, you forget how many people there are that really don't think about their technology (or work somewhere for whom the most important things are cost and stability).

It would be interesting to take a poll of what people think the numbers would be like before they saw them. I would have guessed a lot more Chrome/FF.

[driverdan]

What's wrong with Windows 7? The majority of the world runs Windows and 7 is a fine choice. It's rock solid and there's really no reason to upgrade to 8.

[bduerst]

Maybe they're worried that Windows 7 is only 41%?

I think W7 is the de facto for businesses now.

[austenallred]

Nothing is "wrong" with it, I'm just surrounded by people who obsess about tech all day and go after whatever is newest and shiniest - if you're not running chromium on your 15" retina macbook pro then you're obviously incompetent.

Turns out for most people stability and cost are much more important than having something incrementally better.

Not passing judgment on one or the other, just thought it an interesting realization.

For what it's worth, I run Windows on a virtual machine to do some stuff and I use 7.

[slayed0]

Turns out for most people stability and cost are much more important than having something incrementally better

Or, rather, people have different definitions of better. For many people, cost and stability are huge components of their personal calculation. I think it would be more appropriate for you to say:

having something incrementally NEWER

[tpurves]

I work in tech and take the latest shiniest things very seriously. But when it comes to PCs I don't put Windows 8 in that category, for me it's an egregious regression in UX quality and pleasure of use than Win7. Hope springs eternal for windows 10, but for me Win8 is so packed full of annoyances that I would never consider it to be an upgrade from 7.

[sheltgor]

Have to agree here... my company even develops primarily for Win8, so its not as if I'm not used to using it, but I still vastly prefer Windows 7.

[DanBC]

I'd love to see the stats for HN for browser use.

[joedavison]

On February 20, I had a story hit the front page of HN on a brand-new blog that didn't receive traffic from anywhere else.

Here are some screen shots from my analytics.

OS: http://i.imgur.com/oZNOxCK.png

Browser: http://i.imgur.com/84DGWHz.png

[hackmiester]

For the amount of *nix people I see showing out on here, I'm surprised that it's only 10% - although for the Web in general, that's an absurdly high number, still!

[nitrogen]

Some percentage of Windows visits could be browsing at work, maybe?

[dublinben]

>stability and cost

If this were truly the case, they wouldn't be using Windows. It's mostly inertia in large organizations and for the general public.

[NeutronBoy]

I'm not a sys admin by any means, so I may be out of the loop here - but is there a viable alternative for management of tens of thousands of systems and users, outside of Active Directory?

As good as *nix/Apple stuff is, is it really feasible to run it at scale across an org, without building custom deployment tools each time?

[nitrogen]

There were giant UNIX networks before there were giant Windows networks, using things with names like NIS/YP and NFS. IMO building custom deployment tools is kind of the point of using not-Windows, but I'd wager Canonical, RedHat, and/or Oracle have something for central management and deployment of Linux systems.

[higherpurpose]

I'm actually surprised at how low the IE number is - for government websites. Obviously you can't compare it to the traffic on HN or Techcrunch, where like 70 percent must be on Chrome and 5 percent on IE, but seeing how some analystics sites still put IE at 50 percent market share, I think that's quite impressive.

Or NetApps' market share numbers are crap, and StatCounter's (and others') numbers are a lot closer to reality - which is that Chrome dominates the web browser market.

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

[Hovertruck]

I slapped this together recently using Chartbeat's realtime data (so the sources are major sites – nytimes.com, foxnews.com, espn, etc): http://percentoftheinternet.com

http://percentoftheinternet.com/windows http://percentoftheinternet.com/chrome http://percentoftheinternet.com/ie

...etc...

Raw data is here: https://api.chartbeat.com/cbtotal/?v=2

[acdha]

NetApps numbers appear to be based on a small number of corporate sites, which are skewed towards US media companies. They weight the numbers by the CIA World Factbook data for internet usage in a country but they've never given a reason to believe this approach doesn't exaggerate the distortions present in their sample of data.

If you want reliable numbers, nothing beats direct measurement for a comparable site. I've never worked on a site where traffic was anywhere near NetApps’ but I've found Akamai’s numbers to be reasonably close for sites which target a general audience:

http://www.akamai.com/html/io/io_dataset_v2.html

[thejosh]

IE8 is just 3%, great news. On par with what we see.

Most people whinge about IE8 support, yet their mobile experience is awful. 25% of these hits are mobile, we see close to 40-50% for us, with much higher for hospitality clients (bars, restaurants), though they usually get <3k sessions a month.

[kevan]

Similar IE8 and mobile numbers here. IE8 support used to be a main focus because 3% of a really big number is still a pretty big number. Then we started paying attention to mobile stats and realized how much we were missing in the mobile experience.

[JimmyM]

What would someone who really thought about their technology run? Why would the most important things for an operating system not be cost and stability?

e: Realise this comes off as confrontational. Genuine questions!

[vacri]

Stability and features tend to fight each other.

[Bedon292]

It is interesting to see the IE numbers, I suspect it is because many government computers do not have other options. When I was in the military it was all we had. CAC cards also only work properly in IE, on government sites at least, so that will likely also skew the numbers. We of course were running extremely outdated versions too, so that may be part of the higher 8/9 numbers then you might expect.

[_ea1k]

As someone that has had a CAC, it does work on non-IE browsers (including on non-Windows OSes) for most government sites.

[zackify]

It does but half the crap they have us use runs on flash and only works in ie8...

[_ea1k]

I would have guessed much higher for IE (closer to 50%). I'm glad that it is ~30%, as that means we have a relatively competitive market with a diverse array of quality products to choose from (and yes, that includes recent versions of IE).