Monday Mode

Calling all converts in big corporate environments: is not just about your time and your tasks. If a lower daemon of some misbegotten lower dungeon can't get his/her act together, it is your solemn duty to point out what they can do to improve things for themselves.

For instance:

  • I've got a "Getting Started" entry in my personal TiddlyWiki with links to 43folders, the wiki and GTDTiddlyWiki, which is outdated but easier to get into. These days I just copy and paste it into an message (which renders the HTML as clickable rich text) and send it off.
  • I've started publishing a heavily modified TiddlyWiki (with a stylesheet) as a sort of internal "public" notebook, listing project issues and outlining solutions as we go along. With the EncryptionPlugin at YATWA, I can encrypt sensitive data and share different passwords with different people, which solves a lot of the niggling "but should everyone know about this?" issues that mire corporate sensitivities.

Statistics are Meaningless

Update: In all fairness to Technorati, they seem to be improving their indexing. I just now received the midnight referrer report, and was utterly amazed to see a referrer from here, pertaining to my . There seems to be some confusion regarding what a post is (Technorati thinks headlines with links are post boundaries), but the only thing that really needs fixing is the "days ago" bit...

Not much news today, so I'll run with for starters.

There seems to be a decrease in the signal-to-noise ratio regarding the ego trips in the Portuguese "blogosphere" (i.e., the noise is increasing), and almost everyone seems to be missing the point that whatever the stats counters you use, technically they're all meaningless, because all of them are fallible and (this is the bit that politicians really don't get), none of them have any real relationship to actual eyeballs (there's caching, browser variants, NAT, proxies, you name it).

"Link indexes" like Technorati aren't much more reliable either, since they are demonstrably limited in their ability to index content correctly (Technorati can't make head nor tail of unconventional sites like mine despite my recent HTML tweaks, and it doesn't look like the competition is much better).

Again, statistics are meaningless, and you don't have to have a technical background to understand why. I guess it's one of those common sense things that get hammered out of you after you've been through politics boot camp or something...

For instance, I use Sitemeter, but it is only useful to have a general idea of traffic (or, when compared to server logs, as a way to figure out who syndicates your content). It's so fallible it's laughable (and so are most others), and I've stuck with it mostly because it gives me an idea of trends without having to go to the trouble of processing server logs.

(Besides the syndication data, the monthly deltas are the best information I get out of it, and I take the absolute values with a large pinch of salt.)

Obviously, over time, relative measures achieve some degree of precision, if only because the error rate is the same for all the sites involved - so you can get some information from Sitemeter, but only if you compare listed sites within the same time periods.

But (and, again, this is the bit you really want to let sink in) if you believe any sort of site rating, then you simply don't know enough about how the Internet works, and ought to keep your mouth shut in order to avoid making a fool of yourself.

Personally, the only thing I like to keep track of are Referrers, and even those are constantly being - the only useful input I get from them is who quotes my posts (hit counts are also essentially meaningless), and, like John Gruber, I happen to think they're vastly more useful than Trackbacks.

Oh, and do take a moment to read the before making a nuisance of yourself. I am aware that derogatory comments are two-a-penny in Portuguese blogs, but I prefer to have constructive feedback. So if you want to express an opinion, at least make it a civilized one.

This page is referenced in: