What HN Users Use (45% Mac, 63% Chrome) | Hacker News Reader

What HN Users Use (45% Mac, 63% Chrome)

dangrossman.info

83 pointsjohnx123-up43 comments

To be clear it's what HN Users that are interested in a web technology use.... and surprise surprise there are a lot of Macs in that sector.

If you posted something interesting to games developers I'd wager a lot more Windows PCs and a lot few Macs.

Quite true.

I remember in the Games Developer Conference Europe 2009, all the Macs I managed to see, where actually running Windows.

Haha, I love that. Had a friend who did the same thing. But seriously: why not get a beefier computer for your buck and if you need OSX, spring for a copy and install it on a virtual machine?

"if you need OSX, spring for a copy and install it on a virtual machine" ... because that's against the license agreement?

Except that in most countries EULAs are actually void.

Really? I wouldn't have guessed that.

I also figured this person was in the States. I didn't even think to check

Because beefier != better in all cases.

More accurately it would be "analysis of user-strings of people's browsers whose referrer was hacker news when visiting my website"

Exactly. There will be sampling bias. Opera in particular makes it very easy to turn off sending referrer information (one click on the quick preferences menu), so Opera will be underrepresented in any data corpus derived from the referrer.

Or worse, Opera users will change their user agent string to Mozilla so as not to have to deal with ua-sniffing websites that work fine on Opera but warn you anyway about using a unsupported browser.

I saw the story, but didn't click the link. So it's definitely not going to match what pg and team see when they examine their logs for actual HN Readers.

it would be great if pg published these stats

And limited to the subset of HN users interested in the topics he posts about, too.

As with the recent post on HN's age/gender, this does not reflect data obtained ~5 months ago through a HN poll [0]. On one hand, the poll probably has some self-selection bias, it is only representative of registered users and it asks "what's your primary OS" rather than "what OS are you using right now". On the other hand, the "date range picker for Twitter Bootstrap" post was probably heavily biased towards people who use Bootstrap and it measures what OS visitors were using when visiting which is not necessarily their primary OS. Anyways, here's the results from the HN poll:

    OS          Points
    --          ------
    OSX         3252 (40.9%)
    Linux       2666 (33.5%)
    Windows     1729 (21.7%)
    iOS         104 (1.3%)
    Android     99 (1.2%)
    Other Unix 
    variant     73 (0.9%)
    Chrome OS   20 (0.3%)
    Other       15 (0.2%)

[0] http://news.ycombinator.com/item?id=3786674

From my Hacker Newsletter project (http://hackernewsletter.com) I can say that about 70% of HN users use Gmail.

And probably a lot of people not using Gmail use Google Apps. I rarely use my @gmail.com addresses, but all my domains with mail are managed through Google Apps.

Internet Explorer didn't even make the list for browsers. I'm not really surprised, but as a web developer it's always nice to be reassured that it's dying/dead.

You can't conclude that from his data. His audience is web developers, so obviously IE will be very low.

I see HN as a fringe audience. Trends in the technology world usually happen first here. I do see your point though, and I'm not drawing concrete evidence from this thread.

IE 9 and 10 have been good to me. I just hope we keep a good balance of browsers going forward.

Internet Explorer is still the most used family of browsers. I also produce this Global Stats report from usage at over 50,000 websites:

http://www.w3counter.com/globalstats.php

I'm surprised Windows users outnumber Linux users. I'd like to see a breakdown as to how this changes by time of day. I'd suspect Windows users are in second place because of the workplace. At night I bet linux gets a boost.

In the other stats thread[1], users self-reported themselves as Linux 34% and Windows 22%, which I was highly doubting. There are probably a lot of closet Windows users here, or it's more l337 to claim Linux (even on an anonymous poll?). I also understand there's a selection bias related to the article's content.

I won't say polls are useless, but I keep being surprised how unreliable they are.

[1] http://news.ycombinator.com/item?id=4421971

I could use Linux but then HR wouldnt be able to schedule me for sensitivity training. (Snide but true.)

I'd be interested in this too since I use Windows at work and Linux (Mint 11) at home.

I was surprised to see that iOS is beating firefox. While this is just one website, I hadn't realized firefox had dropped so much in the ranks. It's still my browser of choice.

It's not a global sample. It's a sample of HN readers only. That means it's dominated by a population focused on web development, which means it skews heavily OS X and Chrome (and to a lesser extent iOS). Firefox remains popular among the general public, though it's dropping fast in the face of some really great work by Chrome. Obviously IE remeains popular too, and it doesn't even register on the linked chart.

The data is presented in a neutral way, but from the comments I get a sense that people are happy that other people use what they use. Why is an Apple (or Chrome) echo chamber desirable to having discussions with people that use a variety of technologies and software?

And people without javascript or blocking w3counter are also not shown

There's a noscript fallback. If w3counter does it's job right, people without javascript should be counted.

Those who explicitely block w3counter wouldn't, but that's few people, unless it's in an adblock/ghostery/... blacklist.

I'm using RequestPolicy, which I'd assume also hinders tracking even without being somehow explictely blocked. No idea how wide-spread those extensions are though.

Will this still collect the data if someone is running no-script?

Your browser (unless configured otherwise) will still send the User-Agent string with the request headers, regardless of script settings.

I am surprised that the estimate for Mac is so high. I would have never expected that. It could be the topics that he is posting. Either way interesting read.

Interesting disparity between OS X users and Safari users.

Less than 1% IE users is great news!

Dan, the sidebar makes this page unusable on mobile (Android).

You're not the first to say that, but I don't know why. The site's fine on my Android phone and tablet. The markup/CSS is so simple it's hard to imagine any browser having trouble with it.

I've already spent enough time abusing the display model Android tablets at my local big box stores to test Improvely without buying a bunch of devices. I think if I go back again to try to live-fix my blog theme, they might kick me out. ;)

So in Chrome for Android the problem comes only when you zoom in by double-tapping. Your site is quite readable without zooming so it's mostly fine, but it does definitely break.

For your pleasure, some screenshots:

* Chrome: https://dl.dropbox.com/u/1853222/dan/Screenshot_2012-08-24-1...

* Chrome zoomed: https://dl.dropbox.com/u/1853222/dan/Screenshot_2012-08-24-1...

* Browser: https://dl.dropbox.com/u/1853222/dan/Screenshot_2012-08-24-1...

* Browser zoomed: https://dl.dropbox.com/u/1853222/dan/Screenshot_2012-08-24-1...

For reference: Galazy S running Cyanogen10

It also breaks when zooming in ios, so if it's the same problem and you're on OS X you might be able to fix it in the simulator?

I don't own an Apple anything, unfortunately.

The problem is display:fixed. http://stackoverflow.com/questions/4889601/css-position-fixe... has a good overview of why it happens.

hope to see all the web like this!

Reply on news.ycombinator.com