imc just posted about weird spam — that is, spam which doesn't seem to benefit the spammer. There's an example linked from his post.We've had similarly weird spam attempts[0] on RGL, for example:
Comment added by gjtbx wicbr: vqkmli ijxdfazsn narjbcsxm lyoizrd gmivefaxp wjmivg okyatsjf http://www.bdiup.fohmiwpxd.com
Just nonsense words; not always with a URL included, and where there is a URL, it's nonsense too.
fanf suggests on imc's post that the purpose could be to see how diligent you are about deleting spam, but even so, if the content differs significantly from what you would actually want to post, it's still not a fair test since nonsense could slip through spam filters where real spam would get caught. Why not just test with actual spam?Anyone know what's going on?
[0] We have automatic spamtrapping, but these were quite hard to write rules for. I eventually went with "two-word lowercase username plus two occurrences of q followed by not-u", which caught a fair proportion of them.
