> - need some sort of a spam robot protection function like we have now at the > bottom of each post (little blue box) or like the number system in this > example from a commercial site: > http://japan.axxs.org/5114

What you're talking about is a CAPTCHA - for various reasons, these are not as effective as they might seem to be. There are even ways of diverting the CAPTCHAs to *human spammers* on another website so that they read the digits/letters and can give a correct reading every 5 seconds or so and earn some money (or avoid being sent to a reeducation camp?) while not actually themselves going to the website that getsspammed...

There is already an antispam feature in place, based on a list of regular expressions from a centralised collection http://arch.thinkmo.de/cgi-bin/spam-merge plus a local list of expressions.

No antispam method is perfect. The present antispam method means that *sometimes* an anonymous user will type an innocent message and be considered as a spammer by the system. There are two "short-term" solutions: (1) log in and then you can even type in v..... and words about various human intercorporal intercourse without getting considered to be a spammer, or (2) record the exact text you type, and send it by some method to the sysadmin (tech group on axxs, or kaz or i) and we can work out what set off the antispam trap.

The problem with (2) is that anonymous users will usually not bother to do this.

To test the mechanism, look up some of the latest spam on the dada site (if you can find it) and try publishing it (withOUT logging in) on http://japan.axxs.org.

For the moment i consider this point to be under discussion, rather than on the todo list, but anyone feel free to edit the todo lists - they are wiki pages. https://docs.indymedia.org/view/Local/ImcJapanTech