Bug Stomp

Detailed Search appears to be broken. It won't leave the first 16 bots when the "Next" link is clicked. Keeps looping to the first sixteen found by the search criteria.

I've been having a lot of difficulty with "Language Center Import file". Seems I'd import something, and the Language center would not replace the old Language center with the new one, but would selectively replace key phrases and seeks and such based on differences between the old language center and the incoming file. Unfortunately, my Language Center got messed up in this process, and it wound up with statements I could not get rid of. Some were duplicates, some were near duplicates. I could export the bad file, fix it, re-import it, and the mistakes would still be there. I tried fixing it in the on-line Language Center, but that didn't work either.

I think I have found a work around. I created a stripped down import file: nothing but xkeyphrases, with minimal responses. I imported that, then imported my corrected file without duplicate lines, messed up indents and so on. Seems to work.

I wonder if the repetition detector is broken or switched off intentionally.

sometimes when I put a key phrase for example like left 4 dead sometimes the bot does not recognize the word unless it is capitalized can this be fixed because sometimes I asked my bot like mortal kombat and I think he says im a good person unless I capitalize it.

Use raw mode, because such expressions are spell corrected otherwise, if it is recognised if capitalized that must be because that the spelling correction recognizes that as a name.

Btw, with capitalized, do you mean, "Mortal Kombat", or "MORTAL KOMBAT"?

Mortal Kombat I believe.

The AI engine seems to be stripping AI script out of xkeyphrase responses AND regular keyphrase responses. I've got several xnone responses that I want to use AI script with, such as:

I'm sorry, I didn't catch your name... {PF? if (mem-name) is "sweetie"; once; ?}

(I'm using curly bracket because the forum doesn't like the gt and lt characters.) Maybe that's the problem? How do I fix it?

The language center seems to be stripping the AI script entered after the responses as well.

I am also losing the "once only" designator on the first response only. The rest of them seem to be staying put.

Or is there some rule against using AI script in x-keyphrases? Or is my AI script incorrectly formatted?

Unless I'm doing something radically wrong, this is a bug, and a bad one.

OK, so the AI engine is importing *some* of my files correctly, and *some* files get AI script stripped out. The only difference I can see in the files is that the files that get stripped do not have keyphrases in alphabetical order, while the files that keep their keyphrases are modifications of exported files, thus are in alphabetical order.

I tested the file in alphabetical order, it still got all the AI script stripped from the file when I imported it.

Does anyone have any suggestions? As far as I can tell, the files headers are identical, the AI syntax is correct, I can't see any problems. Help!!

I am completely at a loss to proceed. I cannot get the Forge to accept AI Script in my bots. I switched operating systems, word processors (using Word Pad which is supposed to work). Whether I try to import a language file with script in it, or enter the script after a response in the on-line Language Center, the script is stripped from the response.

I really am at my wits end. Any suggestions would be greatly appreciated. I don't know how you guys do it. I would like to find out. Please, please, please help.

Sam Little,
I just uploaded a file and looked at xkeywords and found all scripts in place. You need to click on "edit" to see the scripts and seeks.

Also, your regex ("^[A-Za-z]$") (re) has wrong syntax. Lose the "quotes" for starters and maybe ( ) are in the wrong place.
try this: ^([A-Za-z]+)$ (re)
BTW, this will not match "James Bond" but this would - ([\\w\\s]+) - but the last line will match almost anything... might as well use (*)

Ronnie075, will ^([A-Za-z]+)$ (re) match the 'James' part of 'James Bond'? If so, that's fine. I don't have a problem with my bot responding: "I will call you James, then" when she asks: "What is your name?" and gets the response: "James Bond"

If he responds "My name is James Bond" then the "my name is" keyphrase will pick up "James Bond" as (postkey) so that part is covered already. But I have problems getting my bot to recognize one word answers with no actual words to work with for a key phrase.

Sam Little

No, that RegEx won't match a *part* of an input line. The ^ means that the match must begin at the beginning of the line, and the $ that it must end at the end of the line.

^([A-Za-z]+) (re)

would match the 'James' part. It wouldn't match names with diacriticals like 'Rüdiger' or 'Søren'. You might try

^(\\\\w+) (re)