de-xhtml is an XHTML-to-HTML XSLT tool by Joseph Walton. Three things to consider:
text/html is largely useless.Joseph's tool helps you realize the benefits of points one and two without the hassle of point three.
More info on Joseph's site.
So, the TypeKey FAQ has done little to assuage Shelley Powers's fears about a centralized comment authentication system. We're not nearly as worried about it, but with all due respect to Mark Pilgrim, we'd like to point out that TypeKey is fundamentally different from Technorati, Feedster, Blogdex, Bloglines, and, for that matter, probably any existing centralized weblog service you can think of. Each of those services, though centralized, affect a blogger's day-to-day operations very little. They are, essentially, weblog observation tools. They track, search, rank, and aggregate, respectively, but what all of them have in common that makes them different from TypeKey is that they don't directly interact with the weblogs they encounter. That is to say, it's possible for us to run a weblog without worrying too much about a sudden change in policy by those services. Contrast this fact with TypeKey, which, if we decide to support it -- and let's face it, we don't allow comments now at all, so this is purely hypothetical -- requires from us an implicit vote of trust in Six Apart to handle something that directly affects our weblog and its readers.
Additionally, those who chastised the community for criticising TypeKey early may want to read the FAQ themselves. In answer to the question, "Why did you announce TypeKey before its launch?" Six Apart answers that they "wanted feedback" and that such feedback "has helped [them] make some additional design and communication decisions." Turns out they wanted to hear what people had to say. Imagine that.
In September 2003 we emulated Mark Pilgrim's slugs -- short, easily-customized, cruft-free URLs. Doing so effected a rather negative change: our referlog -- a tool designed to store and display incoming referers -- no longer tracked individual archive pages. Why? Referlog retrieves visitors' referers via an SSI instruction located in each archive file. Files without the .html extension aren't parsed for SSI and, furthermore, can't be parsed as such using the common Apache server-parsed handler.
What to do about it?
It turns out that it is possible to parse files without an extension for SSI using the appropriately named Apache XBitHack directive, which is invoked by simply adding XBitHack on to your .htaccess file. The Apache tutorial, "Introduction to Server Side Includes," explains:
XBitHacktells Apache to parse files for SSI directives if they have the execute bit set. So, to add SSI directives to an existing page, rather than having to change the file name, you would just need to make the file executable usingchmod.
In other words, the XBitHack directive gets Apache to parse all executable text/html documents for SSI directives. Read that last sentence again because it contains the catch. This hack only works for documents served as text/html. Under our current configuration this means that referlog will track Internet Explorer users only; users of standards-conscious browsers, since they get proper XML documents, will not be recorded.
You may be wondering why you should care about a shortcoming in our referer tracking script. You shouldn't. But keep in mind that this means SSI is non-functional within all extension-less XHTML files served properly as application/xhtml+xml.
We'll admit it's a bit of an edge case, but it's not as though we needed more proof that embracing XHTML is a dubious adventure.
Matt May's fantastic notes on SXSW are especially useful to those of us who can't be there. Readers of this site may be particularly interested in Tantek Çelik's "CSS: The Good, the Bad, and the Ugly" slides (Matt's notes), and Douglas Bowman's "Hi-Fi Design with CSS" (Matt's notes), but it'd be a waste to stop there.
Kevin Hemenway's asking interesting questions about Six Apart's forthcoming TypeKey service, described by its creators as "a free, open system providing a central identity that anyone can use to log in and post comments on blogs and other web sites." We believe those that know them when they say the folks at Six Apart are good people with good intentions, which is why we're hoping the TypeKey debate won't be centered around willful abuse on the part of those fine people. The key here is that, even in the hands of a company that gets weblogs, that wants to do the right thing, centralized systems are subject to a long list of technical and social problems. Quoting Kevin:
What happens when the TypeKey server goes down? Can spammers comment, or would everyone be shut out? If 20 people report Dave Winer as an idiot, who decides whether to revoke his rights? If Kung Lui Pao Kang spams my blog, but has posted 100 valid messages elsewhere, what does that mean for my redemption?
Kevin says he won't use TypeKey. Shelley Powers, either. We're not as sure as those two, but they've definitely made us think twice.
Update: It looks like we were late to the party. Adam Gessaman's got a thoughtful piece on the larger TypeKey backlash, including a lot of links to some good discussions, and even a few answers from Six Aparters to Kevin and Shelley's questions. Wish we'd have seen this thirty minutes ago.
If, like us, you've been unable to resist the masochistic urge to roll your own blogging tool (or if you're the type who, you know, directs their time and resources toward improving existing open source tools for the benefit of everyone), The Weblog Wishlist Manifesto is the type of thing you might want to read. Lisa Williams and Shimon Rura have done an excellent job summarizing and enhancing the huge list of responses to Dave Winer's request for ideas on the future of weblogging tools.
Gotta' love that weblog-as-printer-driver idea.
We mentioned recently our plans to include in this weblog posts-by-citation and posts-by-abbreviation features using meaning-rich markup culled from within our weblog entries. The W3C's Semantic Data Extractor has us thinking more about what we can do with simple X/HTML enhancements. The best part? Creating a service like the W3C's is easy, (they're using a single XSLT stylesheet) since the markup's already in place doing most of the work.