Skip To Navigation

<a ping>

Hmm. I'm undecided.

Firstly, I don't buy the privacy/security/bandwidth concerns. ping isn't the only way to do Bad Things, and it's probably not even the simplest way.

Secondly, whilst valid HTML is usually something I stick to, I'm not that bothered about this not being sanctioned by the W3C. Whilst it would be nice if the WHATWG stuff was implemented in a namespace on top of XHTML 1.1 with the full backing of everyone else and served with the right content type and character encoding, that's just not consistent with this planet. What is consistent, at least in recent history, is adding stuff that works, doesn't break old browsers, and has at least some consensus. It's the best we could hope for, really.

It's good that there's a pref, and that the spec requests UI, but I'll wait to see that before jumping to conclusions about what the "average user" thinks.

Distinctively cliché bottom line: how does this help me? Does it help users get laid? Hmm, maybe not... but seriously:

  • As a visitor of websites, and occasional clicker of - Steve forbid - adverts, this has a minimal effect on me. Sure, my page loads might get a split-second faster, but to me the only benefit of asyncronous tracking is if the tracking server is down, I can still follow links.

    And of course, some sites may end up giving me a choice on whether I get tracked or not.

  • As a bog-standard web developer, I can now track which external links some of my visitors click. But I don't really want to.

  • As a hypertext application coder, or as I like to put it, HAC, things are slightly more interesting. I track link clicks in my feed reader so the code knows when I've read them. Currently I do it with an ugly redirect, and whilst swapping invalid HTTP for invalid HTML is a bit academic, I'd be willing to experiment.

    The problem, of course, comes when I want to use my own site in a browser that has the tenacity to not support ping. Say... Links, IE, Openwave, Opera, or (so far) SeaMonkey. There's no way I can detect whether the browser supports ping, unless you include Javascript and/or User-Agent sniffing, which I don't. The attribute is useless in cases where the ping is a desired feature.

So, in what seems mandatory for most blog posts recently, some predictions:

  • Firefox 2 will be released, weeks after the target date that wasn't really a target date but somebody on the Neowin forums said it so it must be true. It won't have UI for the enabled-by-default browser.send_pings pref.
  • An extension will be released to flip the pref, with a name like "Evil Tracking Defender". Maybe that but without spaces, since it's cooler. It will quickly be updated with a new version including a statusbar indicator and sound alert for pages that use the attribute. It will get rave reviews on download.com, and, somewhat ironically, Firefox Addons.
  • The extension will be posted to Slashdot, "dugg", and integrated into MR Tech Local Install.
  • Tracking companies will add the ping code to their links in addition to their obfuscated redirects.
  • The comments on Bug 319368 will start to rival the MNG bug in quantity, and lack of intelligence.
  • Google will experiment with the feature, first adding it to their Personalized Search service, then to their advertising. The latter change will cause AdSense users to complain of clickfraud and lost earnings.
  • The rest of the real world will continue as if nothing happened.

My honest opinion, after all these diversions? I'm really not bothered. And since I was bored, and Greasemonkey scripts are leet, here's one I made earlier: Pingping. For every page you load, it picks a ping attribute at random, and pings it after a random delay. Useful? No. Amusing, in a "nuts to both of you" sort of way? I suppose.

Attachment: pingping.user.js / 0MB / application/x-javascript

⇐23 Jan 2006 - Getting cynical... / 12 Jan 2006 - wtf.⇒

Feedback

W / Comment from Gill on 19 Jan 2006 at 17:29.

T / Comment from Gill on 19 Jan 2006 at 17:29.

F / Comment from Gill on 19 Jan 2006 at 17:29.

Sorry, you'll probably want to delete those :P - i did try to understand it - then it went on for a loooooong time / Comment from Gill on 19 Jan 2006 at 17:30.

Whaddafu?! Don't make the feedback links link anywhere, use javascript to write that link with the href javascript: void(0); and use <noscript> to give a link to the new page. Otherwise Mozilla jumps to the top of the page. / Comment from Elliott Hird on 21 Jan 2006 at 11:23.

No, Mozilla 1.7 jumps to the top of the page due to a bug in Mozilla 1.7. Start typing and it jumps back down again.
Now get off my internets until you actually understand the code. / Comment from Robert Marshall / [Admin] on 21 Jan 2006 at 15:06.

So I'm *not* the only person confused by the apparent lack of possibility for this to be used...
And, taken together with the post above (no prefs), this won't be pref-able since most of the world doesn't even know what <a> is right? :)
(I guess I'm guilty of the lack of space in naming... but I think that was because I was lazy in coming up with a chrome name. Nice touch about Mr Tech :) ) / Comment from Mook on 24 Jan 2006 at 10:53.

Heh, I suppose I did turn into "attack everyone mode" towards the end there. Never mind, I think I still made my point. ;) / Comment from Robert Marshall / [Admin] on 25 Jan 2006 at 19:31.

Great idea on the Greasemonkey script! I must still be waking up, because it took awhile to understand why it was amusing, but I laughed once I did get it.
And good opinions. I think I agree with you - all the tinfoil hat-wearers are stupid not to realize this sort of thing is already pretty simple to do. I'm also a little disappointed that it is non-standard (for the moment, at least) and still being implemented. It will probably just end up "muddying the waters" for sites that really need such a feature and need to be cross-browser compliant as well. / Comment from Joey Hewitt on 06 Jul 2007 at 16:52.

Thanks. One thing to add is that since I posted this, the WHATWG draft has become the working document for the new W3C HTML WG. So it's at least closer to being a REC than it was before. / Comment from Robert Marshall / [Admin] on 08 Jul 2007 at 15:48.

Feedback is closed. Feel free to contact me privately.

tH