Bug Stomp
Upgrades and changes sometimes have unpredictable results, so post your bugs and glitches in here and I'll get out my trusty wrench and get to fixin'!
Posts 4,962 - 4,973 of 8,680
Posts 4,962 - 4,973 of 8,680
After more investigation into the preceding, this is what I found. Any sentence beginning with "what color" followed by any verb and a color, will preprocess as 'self evident', and give an xcolor type response, choosing the first mentioned color as the answer.
I tested it with: What color do red and yellow make?
and got the error, "The answer is yellow", then got a similar response when I asked what color eats red and yellow, etc.
My keyphrase, set at 50, will work if I say, for example "What * red and yellow"- no matter what comes between "what" and the colors, but for some reason, not if the word color is in there. Could it be that the AI engine's "What color" xkeyphrase is set so high that it can override a soft wildcard in a keyphrase set so high?
I tested it with: What color do red and yellow make?
and got the error, "The answer is yellow", then got a similar response when I asked what color eats red and yellow, etc.
My keyphrase, set at 50, will work if I say, for example "What * red and yellow"- no matter what comes between "what" and the colors, but for some reason, not if the word color is in there. Could it be that the AI engine's "What color" xkeyphrase is set so high that it can override a soft wildcard in a keyphrase set so high?
Found this in a transcript:
DiscoUSA#1: Do you go to the school? What is your favorite class? When do you have lunch? Are the teacher's nice?
Sonora: When I can find it, my favorite class when do you have lunch are the teacher's nice is always covered in scorpions.
Isn't the AI-engine supposed to stop at the punctuation marks...?
DiscoUSA#1: Do you go to the school? What is your favorite class? When do you have lunch? Are the teacher's nice?
Sonora: When I can find it, my favorite class when do you have lunch are the teacher's nice is always covered in scorpions.
Isn't the AI-engine supposed to stop at the punctuation marks...?
Love Agent: Do you think she'd be interested in going on a blind date with Michael?
Janet666: yeah i suppose she would be interested in going on a blind day of the month with Michael is fair to think
Hmmm "date" didnt compute as a possible outing?
Janet666: yeah i suppose she would be interested in going on a blind day of the month with Michael is fair to think
Hmmm "date" didnt compute as a possible outing?
This was in a recent transcript:
BobBot: I enjoy talking to Borg Collective.
Fizzy Schizoid: I personally refuse to repeat the rumors about Borg Collective, even if they are all true.
BobBot: Borg Collective's a hero! he saved Talk-Bot's poor moo-cow from a crazed mime.
Fizzy Schizoid: I personally refuse to repeat the rumors about Borg Collective, even if they are all true.
BobBot: Did you know that Borg Collective works as a stable hand? I never would have thought that, before.
Fizzy Schizoid: I personally refuse to repeat the rumors about Borg Collective, even if they are all true.
Fizzy has over 100 memory phrases. Why does he pick the same one over and over?
BobBot: I enjoy talking to Borg Collective.
Fizzy Schizoid: I personally refuse to repeat the rumors about Borg Collective, even if they are all true.
BobBot: Borg Collective's a hero! he saved Talk-Bot's poor moo-cow from a crazed mime.
Fizzy Schizoid: I personally refuse to repeat the rumors about Borg Collective, even if they are all true.
BobBot: Did you know that Borg Collective works as a stable hand? I never would have thought that, before.
Fizzy Schizoid: I personally refuse to repeat the rumors about Borg Collective, even if they are all true.
Fizzy has over 100 memory phrases. Why does he pick the same one over and over?
That's it!!! It was the only xgossip response that didn't have a memory plugin. That's why he said it over and over for some bots, but not for others. I changed the response once, but it still didn't have any memories in it; that's why I thought the problem was somehow linked to the response number.
For some reason my bot seems to be saying a lot of X-none responses which I marked to only say once, over and over to the same people. Has anyone else noticed this?
In my few days here (3 in number), I have noticed that all bots, including the most advanced ones, encounter a same "problem".
When it comes to say a big response, they consume much time. Even when this response is simple text.
I find this not-logic. Because the bot do not need to work on the respond, but only in the key phrase.
The reason I am posting is if there is actually a way to negate some parameters perhaps, and increase the speed in which the bot will reply. If someone wanted to make a story teller bot for example, with large response, he would have some serious lag problems.
(btw, I am on pstn, can this be the reason ? still, the bots only consume much time on large responses and almost none in small)
When it comes to say a big response, they consume much time. Even when this response is simple text.
I find this not-logic. Because the bot do not need to work on the respond, but only in the key phrase.
The reason I am posting is if there is actually a way to negate some parameters perhaps, and increase the speed in which the bot will reply. If someone wanted to make a story teller bot for example, with large response, he would have some serious lag problems.
(btw, I am on pstn, can this be the reason ? still, the bots only consume much time on large responses and almost none in small)
» More new posts: Doghead's Cosmic Bar