It looks like my captchas are not entirely working, which sucks. One of the posts on the wife's page that had been giving us trouble awhile back are doing the same thing now that they have been reactivated with captchas. I am not entirly positive that the posting of these comments is automated, but it seems unlikely to me that people would actively spam old posts and not new ones. As for the spam, we have been trimming the spam down as they come in, but we are only so willing to keep doing that.
I have changed a couple of things involved in the posting so that they are flagged in the raw accesslogs better. I reevaluated a couple of areas in the posting code to make sure the captchas are being enforced. There are some SQL injection issues I need to doublecheck to make sure things are behaving the way I think they are behaving. I have also added a logging mechanism into the mix that I hope will help pinpoint the issues...and I will probably be having some fun with this feature later. I figure if all else fails, I can use this new logging mechanism to start banning ip addresses if the posts are frequently from the same ip or agent. I have a few other ideas floating around in my head that may or may not help, and I will probably be testing them over the course of the next couple of days. One of which is a new algorithm for the characters on the captcha that will have upper and lowercase chars, the full alphabet, and be a little less predictable...hopefully.
A minor change to the Recall feature I added recently, probably for my own benefit more than anyone elses, will affect if the recall link is displayed. Previously, the link had been a hard coded link that called a javascript function. Now, there is a piece of javascript that writes the link to the form. While this won't change how it acts for most people, some of us run without javascript enabled, and this will prevent a link that does not do anything form popping up.
