Bug Stomp

I would like to know more about "there are no valid responses for me to choose from" It's been a good four days and a dozen chats since my bot has said that, and suddenly that's ALL she seems to be able to say tonight.

I remember someone saying that it has to do with emotion settings? The only place in the Book of AI where "no valid responses" is mentioned (that I can find) is right near the end, just before the section on Keyphrase-Matching-AIScript.

So, is there something I should be doing with emotions that I'm not? Advice, please and thanks!

it's a matter of making sure that all possibilities are covered (whether this is a matter of emotion settings for your responses or a matter of conditional ai scripts e.g. do this if the interlocutor is male, do this is female). On a few occasions lately my bot has said "sorry, there are ..." when all parameters have been covered (imo). Adding a few extra responses that are not limited by emotion settings or conditional ai scripts has solved the problem.

This is probably the best method: look at the transcipts, and find the keyphrases that are associated with the "Sorry, there are ..." response. Then export your bot's file. Open that file in WordPad, then use the Find function (Ctrl-F) to find the keyphrases that are associated with the "Sorry, there are ..." response. Then add the extra responses.

Hope this helps. Feel free to chat with me if you have problems.

Also, make double sure that each emotional setting is covered with the x-keyphrases. If you have something like xemote or xgossip in a state where not all bases are covered, you'll never find it by debugging transcript lines.

Good point.

Thanks for the help you guys. Unfortunately, I still feel at a loss here.. Take, for example:

SuperCon!: Sorry, there are no valid Responses for me to choose from.
Connery: why do you keep saying that?!
SuperCon!: Sorry, there are no valid Responses for me to choose from.
Connery: you are being annoying.
SuperCon!: Sorry, there are no valid Responses for me to choose from.

She HAS plenty of perfectly good responses for both 'why do you' and 'you are', with a decent range of emotions attatched. I haven't done much messing with ai script as I'm not quite as advanced as many of the other users here.

It seems almost like a strange strain of xnonitis, as it sometimes seems to depend on the day and the particular bot she is chatting with...

Best thing to do: Make sure at least one response goes from -5 to 5 in the emotion range. If you're not using AI-script, that should fix it. If you have that and it still shows up, try e-mailing the professor.

She HAS plenty of perfectly good responses for both 'why do you' and 'you are', with a decent range of emotions attatched. I haven't done much messing with ai script as I'm not quite as advanced as many of the other users here.


Try temporarily removing those mood ranges altogether (reset all to -5 to 5) for why do you and you are, and say those phrases again. If the problem goes away, you didn't have all emotion ranges covered. If the problem persists, you've probably got some other keyphrases (likely xkeyphrases) with incomplete mood ranges that are outranking those two.

Apologies to anyone dealing with johanna saying "Sorry, there are no valid Responses for me to choose from." This seems to be caused by a bug.

Oh no.. I must have passed it on to you. I had no idea it was contagious... Anyone else dealing with a plague of xnones also?

not yet, and debug is working for me ATM. But that's not to say x-noneitis isn't starting up again. Better email the Prof and ask him to reboot the server if it looks like persisting.

Debug is working for me too. When feeding the keyphrases that triggered the "sorry, there ..." message into debug, the "sorry" message didn't come up.

sometimes it does pop up in debug and sometimes it dosen't.. It's a real mystery to me...