WP Mollom likes your comments
Over the past couple of days, it became apparent that WP Mollom suffers an issue where it “eats” your comment when a CAPTCHA is shown. The result is that only half of your comment is saved. Not good of course. Apart from that, there was also a problem with character encoding.
Of course, this issue needs solving. I could recreate the problem on a testsetup and found the culprit. This issue rears its’ ugly head when a commenter uses double quotes in his/her comment. Because of the way Wordpress implements commenting, I have to embed the commentdata in the CAPTCHA form as a cluster of hidden fields. The handling of the encoding was a bit wonky here which causes data to get corrupted.
I fixed the issue and commited the code in the development version of the plugin (trunk). I’m not commiting it to a stable version yet because the adjusted code needs testing against foreign non-western character sets like simplified chinese and such. If your blog is set to use UTF-8 encoding (which it should!), you shouldn’t notice big problems with this update.
If you’re really anxious to get your hands wet, you can download the development version here.
Op 11 mei 2009 om 6:06 pm
Question: on most of the blogs using Mollomw that I know of, I have to enter a captcha every single time I post a comment. Is it possible to make Mollom remember my e-mailaddress? I don’t like captcha’s :)
Op 11 mei 2009 om 6:07 pm
Point proven (captcha again :P)
And I’m 100% sure that even though I just posted an approved comment, I’m gonna have to enter a captcha for this one as well.
Op 27 juni 2009 om 2:46 am
Have there been any updates on this issue? Has anyone tested it with non-Western character sets yet?
Op 27 juni 2009 om 10:47 am
Hi. I didn’t receive any feedback, but it would be great if you’d take the time to give it a spin. If it works for you, I’d happily commit it to a new stable version.
Op 28 juni 2009 om 12:55 am
I would be quite willing to test it out if my blog were using a non-Western character set, but it’s already UTF-8.