I am an unabashaed Memeorandum fan and booster, and in the past I have said: I would praise the “impressive signal-to-noise ratio, but the fact is, there’s no noise.”
But as of late Friday/early Saturday, this is the top story:
Actual header on the HuffPo report:
Bill Kristol To Become New York Times Columnist In 2008
Atrios is listed first among “Sites Linking to This Page…” so that probably has something to do with the mishap, though it doesn’t actually explain it. As for “Jesus H. Christ,” your guess is as good as mine. Same for the “(2)” bit, whatever that refers to.
Here’s the Atrios post linking, for what it’s worth:
Your Liberal Media
Publishing lying conservative psychopaths since I can remember.
No Jesus there, nor at HuffPo.
I’m sure whatever this is will be fixed before long, but it suggests a problem for Gabe Rivera’s meme-tracker (and its sister sites) in the near future: as blogs continue to be added to the Memeorandum list, and those sites change trade in their simple CMSs for more elaborate ones, Memeorandum could get noisier.
For awhile now I’ve noticed an existing issue where authors of posts are sometimes misidentified — especially blogs at National Review Online, where even posts on Jim Geraghty’s Campaign Spot are often listed simply as “NationalReview.com” and attributed to Katherine Jean Lopez. I searched the December Memeorandum archives for an example, but came up empty, so you’ll have to take my word for the moment. My impression is that posts on The Corner are properly attributed, but the other NRO blogs are not.
Don’t get me wrong, Memeorandum is still the best gauge of what’s happening in the political blogosphere right now, but it isn’t 100%. It is still something like 99%, and I hope it stays at least that good.
Update: Gabe responds in the comments:
Thanks William. Made a manual fix after your post tripped one of my alerts. Not sure what the underlying problem was…I’m not able to reproduce it. Hope it doesn’t happen again. I guess this shows days go by when I don’t even look at memeorandum. That is definitely the case. As for the bigger picture, notwithstanding this, I think my system is always improving at extracting headlines, bodies etc. There has always been an error rate, and it’s being gradually reduced, but will always be non-negligible.
Thanks William. Made a manual fix after your post tripped one of my alerts. Not sure what the underlying problem was…I’m not able to reproduce it. Hope it doesn’t happen again. I guess this shows days go by when I don’t even look at memeorandum. That is definitely the case.
As for the bigger picture, notwithstanding this, I think my system is always improving at extracting headlines, bodies etc. There has always been an error rate, and it’s being gradually reduced, but will always be non-negligible.