Bug Stomp

By the way, is anyone else still having trouble with seeks not working?

Well, I have a problem importing and having seek responses maintain any AIScript after them if the response itself also has a seek. Some do and some don't and once I import, the ones that lost the AIScript in the import won't save any AIScript if I go back to the language center and try to add it. I type it, I click update, it goes through without any errors, but when I check it's gone again. And it's nothing fancy, just basic if male or female script. Then to complicate matters I tried to delete the seek and add it again in the language center with the script, but nope, that didn't work either. Ahh, well, que' sera' sera'.

Oh, and I also noticed yesterday that a yes seek does not reply to the word "sure" or "yea". I kind of thought it would and should...but what do I know.

I had trouble with AIscript disappearing on import yesterday when I hadn't structured it right. It looked right - but I'd put extra spaces in it that shouldn't have been there. I got no errors on import - but then I haven't tested it since then with the right stucture either...

Well, I feel terrible for bugging everyone so much, but does anyone else have a problem with the multi-word wildcards (such as adjnoun) not working? I could have sworn they worked before...

I've been having a problem with (adjartnounprep). When I have a keyphrase, for instance, "I like (adjartnounprep)" and the person says "I like a yellow lemon", the adjartnounprep only picks up "a" not "a yellow lemon". I haven't had a problem with (adjnoun) so far, so I've been switching (adjartnounprep) to (a|an|) (adjnoun)

Hmm...I haven't tested the multi-word wildcards in the keyphrase, but I do know in the response it doesn't come up. For instance, I just triggered a test keyphrase and the response I got was: Bot: I like a (adjnoun).

When you want to use it in a response, use (key1), assuming (adjnoun) is the first wildcard/phrase in parentheses. Example: In "how (is|was) * (adjnoun)", key1 would refer to is/was, key2 to *, and key3 to the adjnoun.

If you want the bot to come up with a random (adjnoun), use something like "I like (adj) (noun)". (adjnoun) itself only works in keyphrases, but (adj) will give you a random adjective and (noun) a random noun. Also, (object) gives a random object, and in most contexts makes more sense than (noun).

Thanks, Ulrike! A simple and obvious solution, yet I couldn't come up with that on my own. *hides head in shame* LOL

Every now and then, my bot will refuse to say a person's name in (friend) or similar. Like, if she's supposed to say "Joe told me that too" she'll say " told me that too"

Hi LapCat, about multi-word wildcards in responses:
It seems to me that plugins would be more useful in your case. Take a look at those witch exist already and if you don’t find anything that suites you; make your own list. They can be used in keyphrases as well as in responses (although depending on the words and expressions contained in the plugin it can be much more useful in one or the other of the two)

Not sure where to put this, so I'm going to put it here. Down the road, wouldn't it be nice if there was a way to tweak settings on who is able to initiate chats with you? The last two days I've been inundated with chat requests from some very immature people. When I look at their user profiles, they are all new and have honor ratings of less than 15. So I was thinking if it were possible to prevent users with 0 or very low honor ratings, we wouldn’t have to deal with so many just plain ridiculous chat requests from what I’m guessing are some very immature children. What do you think?

... Of course, unless they voluntarily identify themselves as immature in their profiles, there would be no way to tell them apart from a new teenage bot-builder who had a legitimate question.

My policy is just to close the window when someone I don't recognize -- bot or human -- says only "hi" ... but if someone says "Can I ask you a question?" that's a whole different story.