Bug Stomp

I imagine you're using a conditional AIScript and haven't provided a response for each possibility. Eg: if you have one response for a keyword tagged only with {?PF if female; ?}, and it comes up in conversation with a male, you'll get the error message "Sorry, there are no valid Responses for me to choose from."
Make sure you cover all combinations - in the above case that would be provide at least a response tagged {?PF if male; ?} .
If you're using horribly overstrung statements like {?PF if male; if emotion > 2; if (self-favoriteanimaltoday) is "poodle"; if (mem-favoritehobby) is "dognapping"; if (mem-favoritefood) is "hotdogs"; if (mem-favoritekitchenappliance) is "rotisserie"...
then take a slap on the wrist and learn to walk before you can run. You will make the completest pig's ear of your bot that it's possible to make if you try to write complex AI statements on the fly before you've got the hang of wildcards and regular expressions.

substitute pointy brackets for { and } as ever.

Astro: If you have no clue what any of the ?, if, and mem stuff is, post an example of the keyphrase + response. Then we can give you specific suggestions.

One simple suggestion: go into your bots x-keyphrases and change the xnone "just born and cant speak well" to something else. Xnones are a good place to put conversation starters, requests to change the subject, etc.

Hi

Thanks both for the advice, tho' all the {} stuff flew over my head at the moment..hope to get there eventually!

However, I gave my bot a few more things to say in response to xHello, xInitiate, xNone etc...and now it keeps popping up that there are no valid responses to choose from. eg. I said 'Hi' and I got the error message.

I've looked at all the key phrases and they all seemed to work alright before...but now everthing seems to have gone pear-shaped!

Do you have any (mem-something)s in the responses? Or any "if" statements? Something simpler...look at the emotion settings and make sure everything has at least one response that goes from -5 to 5.

Thanks Ulrike.
I haven't used anything advanced at all. Just simply added a few more options for various xcommands as well as some responses to standard key-phrases. All emotion values are set to 0 to 0 and this didn't seem to cause any problems.

I'm stumped.

That's the problem. 0 to 0 means that your bot will only use that response when its mood is at 0. You want them to be -5 to 5 so that it will respond for ANY level of emotion.

When you get those working, you can tweak it so it gives happy responses at higher moods (3 to 5), and annoyed responses for lower moods (-5 to -3), for example.

Ohhh, thats what those mean. I didn't completely understand that.

Thanks Ulrike! Seems to have sorted most of the problems. (I was following the advice in the AI book (beginners section and hence set them all to emotion range 0)However, despite shifting the emotion ranges, I can't get it to respond to 'hello' or 'hi' etc. This time uses a xnone in response.
Thanks again.

There are 2 ways to determine emotions in your bot while programming.
The first is the emotional response your bot has to a phrase. If they like what the user said to them (the keyphrase), they get a positive number 1 to 5. If they didn't like it, they get a negative number, -1 to -5. If they are indifferent, they get a 0. This you program in next to the keyphrase itself.
The second set is what reponse your bot gives to that person depending on the bot's mood (or how they feel about that person). In otherwords, if they like what that person has said overall, they get a positive reponse, if not, they get a negative response.

In order to avoid the "No valid response" just make sure you have a response for every mood your bot is in. In other words, there must always be something covering all of -5 to +5. However, if you don't want you bot to react to a phrase if they are in a certain mood, then leave out that emotion rating.
I'm sure I just confused the hell out of you... sorry about that. Try it out, and use the debug and chat functions, they help alot. You will get the hang of it, it isn't that hard. You also don't need to use the advanced stuff to make a good bot. (Although it does help)

i'm back for a little while...

There was a couple of time that the debug engine did not pick up on the (*) in a sentance then it worked right a few times. Is it haunted?

I get them occasionally. I think it's a server problem, since it's not repeatable or consistent in which keyphrases/responses it occurs with.