Question: are we talking captcha or REcaptcha?<div>My understanding is that REcaptcha is better than old school captcha </div><div><br></div><div>Have we evaluated it as an option? <a href="http://www.google.com/recaptcha/intro/index.html">http://www.google.com/recaptcha/intro/index.html</a><span></span><br><br>On Saturday, April 16, 2016, Ben Gamari <<a href="mailto:ben@well-typed.com">ben@well-typed.com</a>> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Niklas Hambüchen <<a href="javascript:;" onclick="_e(event, 'cvml', 'mail@nh2.me')">mail@nh2.me</a>> writes:<br>
<br>
> Hi Ben,<br>
><br>
> Could we not have a captcha instead of a reject, to avoid false<br>
> positives? That would require no training.<br>
><br>
> Since I assume most Trac spammers are extremely unsophisticated, a<br>
> simple hardcoded question like "What programming language is GC all<br>
> about?" may be sufficient.<br>
><br>
The CAPTCHAs being broken are the reason why this incident occurred.<br>
I have added some more CAPTCHAs to try to dilute the pool of answers<br>
that they already know, but they still seem to solve them easily<br>
enough regardless. I can only imagine they have some sentient beings<br>
sitting at computers solving CAPTCHAs.<br>
<br>
I don't really feel like we can make the CAPTCHAs themselves any more<br>
difficult without excluding real new users, which I really want to avoid.<br>
<br>
Regardless, my goal here is to error on the side of less filtering, not<br>
more, even if this does mean more manual maintenance. To this end, I've<br>
configured the filters such that the probability of legitimate activity<br>
being suppressed should be negligible,<br>
<br>
* I've been careful to only train the Bayes filter on obvious spam;<br>
I have tested it against various snippets from the wiki and mailing<br>
list and have yet to see it score anything legitimate with a spam<br>
likelihood > 5%.<br>
<br>
* Even if the Bayes filter does deem your content to be spammy enough<br>
to warrant further attention, you will merely be asked to solve a<br>
CAPTCHA. Posts will not be outright rejected unless it is quite clear<br>
that they are spam.<br>
<br>
I am optimistic that the filtering will have negligible effect on<br>
legitimate traffic. As a smoke test I managed to create a new account,<br>
open a new ticket, and start a new Wiki page without even needing to<br>
solve a CAPTCHA.<br>
<br>
Cheers,<br>
<br>
- Ben<br>
</blockquote></div>