do YOU have javascript disabled?

December 12, 2005 12:07 AM

How many browsers really have javascript disabled? I want to implement a stats program on a client's site that uses javascript to collect stats, but don't want to exclude a big chunk of people

In case it matters, it's a music downloading site. right now gets about 10,000 hits a day according to server logs. Also, it is mostly aimed at youngsters.

posted by menace303 to computers & internet (12 comments total)

Sorry to point out the obvious, but why don't you let your stats package tell you?

Or write a teeny Javascript insert that calls an invisible GIF, and work it out yourself.

My observations last time I cared about this kind of thing is that such people are vanishingly few. Certainly big orgs I worked for used packages that relied on JS without qualm (and come to think of it, pretty much relied on JS being present for full functionality anyway).
posted by i_am_joe's_spleen at 12:10 AM on December 12, 2005

I use a proxy that selectively disables javascript; it disables all javascript that's encrypted or any that contains certain keywords. Those keywords include anything that looks "nosey".

On my browser, most javascript detectors will work, but certain operations won't.

In general, try to find a way to do what you want without javascript.
posted by orthogonality at 12:20 AM on December 12, 2005

The point isn't who's got JavaScript on or off, it's if you code a site which requires JavaScript, you have to inform people without JavaScript what's going on. Otherwise it just looks like your site doesn't work.
posted by AmbroseChapel at 1:02 AM on December 12, 2005

I have javascript normally _disabled_ and enable only selectively for sites I "trust" - if there's no other way. I always find it very very poor from a site when there's no fallback (and no message indicating that javascript is needed).

Also: I observe that Opera sometimes modifies a site's javascript code "on the fly" - but havn't looked deeper into it.

So I strongly agree with orthogonality: better find a way to do it without javascript, or provide some fallback. Example: If your stats wants to know the screen size and color depth: add a noscript section with a form where the visitor can tell you. You'll be surprised that even IE gets it wrong sometimes.
posted by nlgw at 1:16 AM on December 12, 2005

According to W3, as of July 2005, 90% of internet users have JavaScript on.
posted by panoptican at 2:10 AM on December 12, 2005

You should use Unobtrusive JavaScript to enhance the user experience, but you shouldn't require it.
posted by Sharcho at 3:24 AM on December 12, 2005

The original poster was talking about stats, not usability.. therefore, the only people who'd disadvantage would be those who relied on the stats, rather than all visitors. In that case I'd advise tracking who doesn't have JS on by using a pixel graphic in a noscript, and then multiply up the other stats by the ratio (or just don't include them).
posted by wackybrit at 4:31 AM on December 12, 2005

One of the most important "users" that doesn't have JavaScript turned on are search engine bots. If you're using JavaScript to create essential things such as navigation, a search engine bot won't be able to follow a JS link and spider your site. Jeremy Keith's DOM Scripting is a fabulous book, should you be interested.
posted by TheDonF at 5:56 AM on December 12, 2005

What kind of stats are you looking to collect that can't be found in your server logs?
posted by odinsdream at 6:57 AM on December 12, 2005

Server log stats almost never line up perfectly with page tage-based stats. See Web Analytics Demystified. Presumably you're looking at implementing a commercial tool that relies on page tags. It's a safe technique, for the moment. The only concern is that a future browser release (that disables javascript by default or changes cookie handling) could lower the accuracy of the technique.

odinstream, page tags track individual user activity much better than server logs, which only can see to the IP address level. Users on DLS or dial-up may rotate through IP addresses, while multiple users behind a NAT-based firewall will all be lumped together with server logs.

The W3 Schools data says 90%, other users say that they get over 99% on their site. You can always try to test this yourself by setting a cookie in JS and seeing if the cookie ever reappears in subsequent page requests.
posted by GuyZero at 8:19 AM on December 12, 2005

Note that many of the current "fancy" stats packages (like Google Analytics) use javascript includes on the page. So they don't even track non-javascript users!
posted by smackfu at 8:58 AM on December 12, 2005

Be wary of sites reporting 99%+ of visitors have JS enabled, they could be self selecting for JS users. IE: if you can't get past a splash screen without JS you aren't ever going to return to the site except by accident and there for aren't going to show up in their logs as a non-JS user.
posted by Mitheral at 5:05 PM on December 12, 2005