It's worth repeating that it's possible to greatly enhance the value of Web content using currently available semantic markup. It's what Mark Pilgrim was talking about when he implored us to "try pushing the envelope of what HTML is actually designed to do, before we get all hot and bothered trying to replace it, mmmkay?"
On a related note, Jon Udell notes that "ubiquitous but underexploited technologies (XPath, XSLT, XHTML) can make our everyday information more useful." His summary, which prompted this post, effectively simplifies the message:
When you work publicly, in blogspace, as I have been doing, reaction to your work is exquisitely measurable. And when I take the pulse of that reaction it's clear that I'm miles away from proving three points:
- Ordinary Web content is already full of metadata,
- which can enable powerful queries,
- which, in turn, can motivate us to enrich the metadata.
We mentioned recently Jeroen Sangers's discovery that the latest version of Movable Type generates invalid markup. It's been fixed (via Matt Burris).
Movable Type developer (and Six Apart co-founder) Ben Trott:
Update: We've released 2.661, an update to 2.66, to fix a problem with invalid XHTML in the comment redirect script, along with using comment IDs instead of URLs to fix an open redirect problem.
Note that we're not sure when this update was released since the fine folks at Movable Type aren't using the ins X/HTML element, a subtle semantic enhancement, which, when used with the datetime attribute, not only alerts users to additional content, but more important, indicates when the content was added.
As mentioned previously, the new Movable Type release uses a proxy script to generate URLs on comment pages. It's not a perfect solution, to be sure, but it does severely limit the effect spammers' URLs have on search engine results. Which makes us wonder: are we the only ones concerned that this proxy script also limits the effect legitimate links have on search engine rankings?
In the past we've used Google to dig up comments we've left on other weblogs and to locate comments posted by other bloggers. This second usage is important. Blog comments provide a pretty good overview of an individual's thoughts on subjects he may not discuss on his own weblog. Comments create an even more useful record of Web activity for those without a weblog. Individual blog comments do not domininate search engine results, but it is possible to find one or two comments per results page, especially when looking for posts by more prolific commenters.
Is stopping spam worth losing this feature? We don't think so.
Joi Ito has some interesting thoughts on classifying comment spam. Knowing where to draw the line, as he puts it, is the first step in maintaining an equitable deletion policy. The resulting discussion is worth a read.
Of particular interest: Adam Lasnik's initial response to Joi's post strengthens the case against URL proxies by emphasizing what we lose:
Many bloggers' admirable sites have become popular from the GoogleJuice (PR) they've attained from their posted comments. And this, IMHO, is a good thing -- everyone has ended up a winner. These thoughtful bloggers have contributed interesting and useful commentary to blog entries, and in return, they've enriched those blogs... and also boosted the popularity of their own blogs.
[R]emoving the opportunities for the 'legitimate' folks to enjoy promotional side benefits from their sincere contributions is more unpleasant, I believe, that the original spam issues...
Jeroen Sangers reports that Movable Type's latest crack at stopping comment spam -- a URL proxy script designed to limit the effect unwanted messages have on search engine rankings -- generates un-encoded ampersands, which render XHTML layouts invalid.
There's a simple solution, which Jeroen provides:
I did a quick search in the new files, and found the error in the file lib/MT/Template/Context.pm at line 1189. Just add 4 characters, and MT produces valid HTML again.
That is, replace the ampersand at line 1189 of Context.pm with &, or &, its numerical equivalent.
Update: This error has been fixed.
Tantek Çelik ponders modifying permalinks to recently posted material:
If a permalink anchor changes and there was no one hyperlinking to it, did the permalink really break?
There's no way, of course, to be absolutely sure that a document is un-linked. Referrers are the only universal, immediate record of a link, and they're only sent once the link is actually visited. The moral: always follow the links you create and keep URLs cool by not changing them.
Tantek's weblog structure overhauls are instructive despite this minor indiscretion -- and the koans are free. So far: better URLs -- but still no slugs -- and an XFN enhanced blogroll. If you haven't already, view his source for an anatomically interesting weblog post structure (built using (X)HTML lists).
We were hoping our foray into blogging this new year would contain original substance -- i.e. that it'd be more than a roundup of what others are currently saying -- but we can't wait for what could be another year to feel that elusive creative spark. So:
Jeremy Zawodny labels the social software nay-sayers point-missers:
Get yourself out of the mind set of social network software for the sake of social network software and start thinking about how adding a social networking component to existing systems could improve them.
It's a pretty good point, though we must admit we're pretty fickle when we don't really know what we're talking about. To wit, Richard Stokes's response seems as practical:
I have hundreds of contacts, but the value I derive from introducing people far exceeds any advantage I would gain by entering them into a system somewhere. Moreover, the value I derive from my hard-earned network is sacrified for the "good" of the system. If anyone, or even just my associates can find out everyone who I know and everything I know about them, I am no longer indispensible. What would possess me to give away my personal "competitive advantage"?
Furthermore, it takes a fair amount of actual work to get a good social network in full swing. We watched every one of our friends this summer take to Friendster feverishly, entering personal details, submitting testimonials, and creating elaborate profiles. It took about a week to get everyone interested. It took about as long for our group to abandon it. So, it's clear to us that Jeremy's on the mark:
The real value of this stuff comes from integrating it with services like Amazon.com or Google or your favorite on-line movie tickets site. Like many things on-line, it will move from novelty to utility.
If you really think that Friendster scored over $10 million in VC funding on the premise of being nothing more than a "browse your friends" site, get real.
Incidentally, Jay Allen's recent networked address book proposition is a great idea.
When I move, I update my contact information and suddenly, every one of my friends has the new information in a central place. No need for mass emails that people have to hunt to find. No more question marks next to possibly outdated phone numbers. No more bounced email because someone decided that they didn't want to use Hotmail anymore.
Go a step further now and integrate this with my email application, my local address book application (for when I am offline) and my Bluetooth enabled cellphone.
We recently changed mobile phone numbers. The current (low-tech) way to distribute this new information among friends is extremely slow and tedious. We're forced to update each friend's address book, as we see them, individually. We may not see some friends for weeks, which means it could be weeks before they get our new number. Worse, not having our new number likely exacerbates the problem by extending the time between visits.
Jay's system neatly solves this problem -- by creating a host of not-so-easy-to-allay privacy concerns. Matt Haughey's recent thoughts on social software, too, include a few great ideas challenged by seemingly insurmountable privacy and signal-to-noise concerns. We'd love to hear more thoughts on these dangers.
All this is to say that our New Year's resolutions are not, in fact, to intensify our tendencies to ramble and conjecture, despite any hint of the previous in this entry.