So... RDOS 3.1 is still under way. In fact it's code complete. I've just been seriously taking my time testing it. (No, not that I've been testing it THAT well, just taking my time to test it... if you follow my meaning.)

I decided to ditch Virtual Path Provider support, though I'll probably just throw it back in a bit later just for the SubText folks if they still want it. (It's a bit too complex at this point, and not suited too particularly to most end-users).

Otherwise, one of the big changes is the addition of rules. Previously RDOS just used filters - patterns that you'd set up in an attempt to block spam. That works pretty well, especially if you've got the time to keep updating them, or are lucky enough to get a single pattern that traps most types of spam. But it doesn't work well when you're just trying to protect certain directories from certain types of traffic. So now RDOS has support for rules - which basically say: "open season on anything that tries to POST to such and such directory". The rules can be pretty complex, but I've found them insanely useful at this point.

In fact, I left my old blog up on AngryPets.com as a type of 'honey-pot' for spammers. Spams against that 'honey-pot' site were coming in at roughly the rate of 500/day. (Weekends were a little worse actually). I just threw up RDOS with a rule against that blog, denying POSTs, and as soon as it went up, my inbox went quiet - no more spams.

I've almost got a whole battery of functional tests (too complex to be considered unit-tests, and they don't test code but outcome) ready for prime-time. Once I've got those done, I'll release them, along with some beta versions of RDOS 3.1...

There-after I'll look in to doing some stuff catered to SubText, and likely just turn RDOS over as an open-source app...