Mollom

Thanks for the feedback

Thanks for the feedback, D'Arcy. I've also pointed Matthias Vandermaesen, the author of the Mollom plugin for Wordpress, to this forum topic. Hopefully he can confirm and fix those bugs.

From a UI point of view, it would be good to see which comments triggered a CAPTCHA and which comments didn't. Like that, you can see whether spam was entered manually, or whether spam was entered by an automated bot. I'm going to add a similar feature to the Mollom module for Drupal.

Normally, there should not be false negatives. This is by design as we return a CAPTCHA in case we are not 100% confident in our decision. That is (part of) the beauty of Mollom. Systems like Defensio and Akismet merely turn the problem around. Instead of looking for spam in your ham, they make you look for spam in your ham. Mollom addresses this problem by eliminating the moderation queue. Even if spam gets through occasionally, there should be a lot less to moderate and review.

feedback

Hi,

Thank you for your feedback! Over the past days, I've been working on a new release. So I'm going to take this into account

On the mollommanage page, it would be good if it showed which post the comments were added to - makes it easier to see if the comment is wildly off-topic as part of an attempt to game the whitelist.

The commenter's URL isn't hyperlinked. Not a big deal, but it's easier to just command-click a URL instead of selecting, copying, new-tabbing, pasting in order to see if a commenter is an evil spam roach.

Will implement these. I'm trying to make the Mollom Manage panel as verbose as the default Wordpress one.

The batch operation doesn't seem to work - if I select a few comments by their checkboxes, then hit "spam" at the top of the list, I just get a blank page.

I'm aware of this bug. I've fixed it over the weekend. There's only one bug left that need fixing before bulk moderation is fully operational.

There's no list of comments that have been flagged as spam - there's no second chance. That gets a little scary, in case of false positives. Is there any way to keep spammed comments around for a number of days before purging in case of a need to recover any?

Are you referring to comments that were blocked as spam without a CAPTCHA? Or comments you flagged as spam, unwanted,... and you want to keep those around for a while?

In case of the first: the idea behind Mollom is to get rid of the moderation queue. The very first versions of the plugin did keep spam around, just in case. But after discussion with Dries and Benjamin, I dropped this. Comments flagged as spam by Mollom don't even make it to the database in the first place.

In case of the latter: after sending feedback, the comment is indeed deleted. It's possible to keep comments that were flagged seperate for a while. But deletion of these will still need to be triggered manually.

I'm more thinking in this direction: after sending feedback, the user will get the chance to delete or keep the comment. If he/she keeps it, the feedback buttons will become unavailable (feedback can only be sent once) and the comment can only be deleted.

I also need to implement the 'approve/unapprove' options to keep back comments without deletion in the Mollom Manage panel.

It _looks_ like some comments are getting past the captcha, even though they are likely posted by bots. Is there a hole? Are the spammers just offshoring their comment posting to actual verifiable human beings to get around captcha?

It's likely that a small percentile of spammers are humans that post spam around blogs manually.

Anyway, thanks for the feedback!