Prior to this, the blog spammers, people that post either poker referrals, drug purchase links, 'hot' stocks, online diplomas [almost as bad as the USC DEN program!], or money making schemes all seem to never have posted on this '...' blog. I used to think that it was because it had custom CSS that I use to display the pages, and that comment bots couldn't figure out which link to click when they came to my page.
That can't be the case now though; I think that more and more pennies/day workers are being paid to drive views to sites rather than bots. I have to say that is the case, I had 15 comments my museum post which is pretty crazy considering I get like 1 unique visitor a month here and almost no real traffic [maybe I should talk about the iphone too!?].
To fight this I've turned on the blogger captchas, but still allow anonymous posts. I've also upgraded to the new blogger, but I can't tell any difference (other than the label/tag support [for which I didn't add to my CSS? hrrm]). *shrug*
Comments:
not to take anything away from your awesome stylesheet and its sweet bot fighting abilities, but i don't think the bots care or have ever cared about it. I bet they just automagically post, via "/comment.do", in N time after getting pinged from a ping server.
I could probably write the script in almost no time.
I thought about that, but I think the blog to post id map should take care of that (by returning null results). but then i have no idea how the 'postID' blogger variable is generated. meaning, as i post this i am posting the comment.g script on https://www2.blogger.com using the two variables
blogID=14597313
postID=7237064045425469409
Which means that if you have an automated script, then how the hell can it match up the postID to the blogID? Try other combinations, they won't work.
Unless they loop through all those 18 digit integer hashes to try and find the one that works.
there is a one to one static correlation between your post permalink and the blogId/postId parameters.
they could request the permalink and extract the blogId/postId values.
which means there is a small amount of html scraping, but no css impact.
Uhh, the point is that there is no mapping. There is no way to extract the postID from the tmarthal.blogspot.com URL back to the blogger.com comment URL. It's just not feasible.
However!
What you say is correct, and leads to my original point. Since I did not use an existing CSS Template when creating these pages, you can't use any sort of typical scraper to scrape the html!
However, they had to've updated thier stuff to use more robust scraping which was the point of the whole post.
uhh, mistaken, the point is that there is a mapping. there is a way to extract the postID from the permalink. It is feasible.
Since we are not communicating in english, lets try ruby. 45 minutes of coding produces a way to map them.
http://www.thirstymind.org/scripts/bloggerMapping.rb
They must have changed things when I updated to the blogger beta. if you notice in that link that you scrape out of the html, you are looking for a "comment-link" class. on my previous blog, I removed that class generated in the css... but they must have put it back (since I am using thier $BlogItemCommentCreate$ variable). :\
I since re-removed it: the html generated by the blogger software doesnt label the 'Post a Comment' link with "comment-link" class. That should help. Maybe I'll turn off the captchas again to see [they are turned off]!
So we will see.
I only used the comment-link class as a simple way to find the url.
I could have also looked for any href with a baseurl of http://www.blogger.com/comment.g (which I assume sploggers would do, instead of rely on a class identifier) and still successfully map a permalink to the postID.
CSS has nothing to do with being able to map a permalink to a postID. It can be mapped with or without CSS.
your site is loading rapidly