The AI Engine

Y'know what would be REALLY great? A "blocked sender" list, to keep kids from coming on with 4+ accounts at the same time, and using them to im me the same message several times at once.

Isaac- Well first i want to tackle the likes of my bots. The likes of the user can always be stored...and I don't think the list for users will ever grow too large so i'm not too worried about that. As for the plug-ins..that would be a possiblity except for the fact that you can't pull up an "if" statement with them can you? What i'm aiming for here is a bot that would mimic actually responding based on its feelings towards the topic.

Professor- Dang...Well while I was hoping for another response I also tackled how it could be handled otherwise. Now the question is when the memory is full will it simply not write to it? Also would tagging a script onto...hmm lets say "xhello" that remembers everything i want get past the user based memory(lol i can only imagine the delay this would cause in the hello statement though). Another question relating to my full memory question, If there is some way to tell if memory is full is there some way to redirect or have a longer check inline so that if say mem-like, mem-like1, and mem-like2 are full it instead writes to mem-like3? Possible example:
<?PF IF (likes) is full; If (likes1) is full; If (likes2) is not full; Remember "(key1)" as "likes2"; ?>

I guess this post is long enough for right now.

Someguy- theres no way to tell how many memories there are in a certain category. When the limit is reached, the oldest memory is dropped off the end, and the new memory is added.

Ouch. This keeps getting trickier and trickier. Still possible though i think.

Is anyone else's bot getting boned by the new "do you" added to the beginning of sentences thing?

Dalchini: Let's find something we have in common - gender, perhaps?
Westerlin Man: I've heard about people who let us find something we have in common...

the keyphrase in question, of course, is 'do you'

I've run into some of that.

Jaden: Hello. My name's Jaden. It's nice to meet you.
Frizella: Really? Billgeratter thinks *you're* your name's Jaden.

Frizella's response is for keyphrase "you are (adjartnoun)". ??

I'm so glad I found this message board, because I ran into another problem If I ask my bot, "I don't know, are you?", he'll respond using the "I (verb)" keyphrase, when I have an "I do not know" and an "I do not (verb)" keyphrase also! In "debug" it changes the phrase to "do I do know". any suggestions?

Ok, so I've already posted this, but it's really beginning to make me want to rip my hair out... >.< lol

i have arrived...there seems to be something going on because of the comma followed by a question. if you want it to respond to the i don't know part of that statement try adding "do i do (verb)" to the "i do not (verb)" keyword. This might not work though because "i do (verb)" might also pick it up, but you might be able to use a regular expression to prevent this. hmmm it should be noted though that "i don't know. Are you?" doesn't seem to have the same problem...

Oh well, I'll just leave it like it is and hope no one uses commas XD

Am I the only one who thinks that using decimals in the rank is a good idea? (hint to The Professor *cough*)

Well, not sure. The engine adds rankings based on the length too, so it's not necessarily the one you ranked highest that will be picked.

I'm already confounded enough with the -50 to 50 range- I hate to admit it but, I'm afraid decimals would make my head explode. lol