Big update coming tomorrow 13 Feb, 2005
I can't say much about it now, but there's a bit of information I'll be posting tomorrow, hopefully with multimedia.
I can't say much about it now, but there's a bit of information I'll be posting tomorrow, hopefully with multimedia.
I’m working on revamping a number of my older RSS feed generators, and came across an interesting tidbit of information. The particular piece I was working on was my atom-to-rss parser, which ties together a number of custom-written components. I was checking to make sure that the cache-to-domtree portion was working and noticed a bunch of weird elements while testing the Googleblog feed (which is, of course, Atom, so it will likely display all weird in your browser).
The weird part is specifically relating to post by a Sean Knapp, one of their employees who (as best I can gather) is involved with their expansion to other countries. You can read it here. Likely you won’t notice anything out of the ordinary, but check the source code. Here’s the relevant portion:
there are country codes like .es (Espana, or <st1:country-region><st1:place>Spain</st1:place></st1:country-region>)
Perfectly normal, you may say? What the heck is an st1, I wanted
to know, so I checked the namespace declarations hoping to find a DTD of some
sort. Nothing in the HTML, and nothing in the Atom feed.
As the plot thickens, I googled st1 country-region, which is something
which seemed rather unique to this vocabulary. Turns out it’s related
to Microsoft Smart Tags which
I have never seen embedded in a page like this. “Oh great,” I
thought, “he’s using Office or *gasp* IE!” Further evidence of Microsoft
products leaving behind signs is <o:p></o:p> which occurs
later in that same post.
The really ironic part, I believe, is that the only reason this came up is
because my DOM parser is trying to make sense of the code, inside a
<div xmlns="http://www.w3.org/1999/xhtml">, which as you may have guessed, is a feature specific to Atom. If they had an RSS feed, I wouldn’t have noticed this namespace issue, because the HTML littered with Office remnants would have been encoded using entities instead of using a namespace.
Google, fix your feed! It doesn’t validate.
Please?
P.S. I’m not the first to discover issues parsing their feed. See http://blog.p3k.org/stories/3428/
What else is new? I especially like how the results on Feedster have nothing to do with what I’m searching for, and nothing to do with Elizabeth II.
Kate’s “Amazing Staff Page” really is quite amazing, and I’ve had this screen capture sitting on my hard drive for ages now. Saved for posterity, you might say. Insert witty comment here.
Theoretically, if something was now 100% cheaper, it would be free, right? I don’t get their math...
I had to install Shockwave (or was it Flash? I forget) on a Windows machine a while back. I had to chuckle at this license agreement. The "I agree" box came preselected. So my action of “clicking” (by my definition, which fits that of wordnet. would actually be to disagree with the terms!
What, precisely, is
the legal definition of “clicked”? To me, the box
below is “checked”, and if I was to “click”
it, on one hand I would be saying I don’t agree to the terms... but
by my action of deselecting (i.e. clicking), I then agree that
I’m bound by those terms that I just tried to disagree to!
I took a look over next to my computer and thought it was somewhat artsy.