Taking a lead from LJer rhiannonstone, I took the MIT blog survey tonight.
Looks like it uses some Python:
"[24/Jun/2005:23:40:09 -0400] "GET HTTP/1.0" 200 89322 "-" "Python-urllib/1.16""
It was a little funny. One of the things it does is reach out and grab some random links from your blog. Then it asks you to categorize the links. Problem is, the first link that it grabbed was the footer link to Alexander Limi's page. It then asked me to say how I came upon the link.
- Someone told you about it
- Saw it on another weblog
- Saw it on a bulletin board
- Saw it on a news site
- Found it through a search engine
- Stumbled upon it
- Wrote it myself
- Can't remember
None of those are really appropriate for this situation. It's part of the copyright notice in the footer. I chose the first one, because I guess in a sense, someone did tell me about it. It came with the default Plone footer template.
The next link it asked me about was the link to the story about Buddhist monks being able to suppress the startle reflex from the recent entry on meditation. This one was also problematic, because the question is worded, "How did you first hear about this story/website?"
Those are two different questions. Apparently, they are assuming that the first time you heard the story, it was in a form that had a link, because the only answer options they give (other than Can't Remember and Someone Told Me) all have to do with Web sites and the Internets. In this case, though, I knew about the story beforehand, from a magazine (the kind you hold in your hand). So, should I be answering the first place I heard about the story, or the first place I found the link? I found the link by searching based on my previous knowledge of such a story from a magazine.
I answered, "Found it through a search engine", because at least that's how I found the link, even if it's not how I found the story.
Unfortunately, the third link it asked me about was another name from the footer. Same story as the first, only this time it was Runyan. Strange. I would think they could have done something to screen out the footer links. Shouldn't be that hard. Perhaps there is a control reason why they did not do that.
The next link it asked me about was to BrainyQuote.com, because I had written a response to an Art Quote of the Day. The questions seemed to fit this one OK.
The final link said, "undefined". And it was not a link at all. Unfortunately, I still had to answer the normal questions about it. I hope they do something at the end to remove all of the "undefined" links, because obviously my answers in this case were totally made up.
I wondered how that happened, so I ran my page through the validator, and sure enough there are a bunch of errors currently. I just upgraded Quills yesterday, and it looks like there are some errors in the templates, and it also looks like some characters aren't being escaped properly in links for whatever reason. So, things to fix. Sorry to screw up the survey. I'm surprised though, that if it could tell that a link was "undefined", it didn't pull another link from the blog to try. Sorry it couldn't cope with my currently ugly-ass blog.
The rest of the questions were more survey-oriented. There is a set that asks about the content of the blog entries. It's kind of hard to say what percentage of posts relate to your profession if your profession is writing. While writing isn't my only profession, I would consider it to be part of my profession, as it's what I was professionally trained for. Supposedly.
Finally, I find it a little funny that they allow you to log back in and change your answers. That's fine, but they also allow you to view how you compare to the results so far. Isn't that kind of strange? I guess it depends on what they are actually studying, but that seems to be kind of unscientific to me.