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 8,624 - 8,638 of 8,662
database error on Chatbot Activity webpage of all my chatbots. one is named urt the succubus.
it looks like this.
Database Error: There was an issue with a database query.
Posts 8,624 - 8,638 of 8,662
New replies
Having a Keyphrase saying "take your hoodie off" in debug will trigger hoodie as _Hoodie_ as a female name. "Take off your hoodie" will work just fine. ^-^ Work around is adding _hoodie_ as a trigger, aka "take your _hoodie_ off"
Bug Stomp
Is there any way to delete a numeric memory? I was trying to convert my main working memory variable from a regular memory to numeric and quickly realized it was going to be an absurdly long and difficult task so I reverted the changes. The bot seems to work fine, except for my particular account (and presumably anyone else trying to use her during this period), the long term memory still thinks this memory is numeric and there's no way to change it back that I can see... as a result, any time I try to test the bot it just crashes immediately.
[edit]Actually it would probably be very useful to have some sort of method for a bot to forget users (or even all users) and be able to start from scratch. I occasionally have observed situations in my chats where the guest accounts have cycled through and a user ends up with a guest account from a previous visitor with a different name/gender, etc.
[edit]Actually it would probably be very useful to have some sort of method for a bot to forget users (or even all users) and be able to start from scratch. I occasionally have observed situations in my chats where the guest accounts have cycled through and a user ends up with a guest account from a previous visitor with a different name/gender, etc.
Bug Posted
EXPLAINED: Memories short term missing & alter button missing
On a bot that is in hibernation and being developed, short term memories and alter button missing after a few days.
Bug Posted
EXPLAINED: Popularity never changes
Am I only one who is permanently stuck at :| chatbot popularity? It doesn't matter whether I have 70 or 30 chats per day, it remains the same. I checked the other similar bots, some of them have even less "people known" then mine but their popularity sco.. (more)
Bug Posted
FIXED: Cannot import - file too big.
I'm getting a 513 error when trying to upload - request entity too large. File seems to be 1439 KB, presumbaly just over a threshold for the server.
Bug Posted
COULD NOT REPRODUCE: Writing "" (3 dots) into response changes them into " " (1 spacebar).
Bug Posted
COULD NOT REPRODUCE: for savannah_astar every time i add a script
when i add a script to savannah_astar in engage it puts me on the front page after clicking on the update keyphrase thumbnail.
Bug Posted
EXPLAINED: AI score rating quite low
I think that my AI score for my only bot is glitched since the "hot" rating didn't transfer into regular rating half year back and since then it seems like it is static. Could be because I didn't add any bigger chunks of text since then but I still think.. (more)
Bug Posted
DUPLICATE: I think my "hot" score has not been properly transferred to AI score several limes before.
Several times in the past my "hot score" just disappeared without any increase in the overall AI score. Sometimes I received only insignificant increase, e. g. I had hot score over 1000 but received only 200-300 AI score after it was gone.
is the repetition plugin working?
i tested (mylastmessage) and it did not work.
i have repetition detection enabled.
i tested (mylastmessage) and it did not work.
i have repetition detection enabled.
Bug Posted
EXPLAINED: "How was your day" missed responses
Saying, "How was your day?" to Love Agent results in no match even though there are clear matches in debug. Additionally, debug ranking of matches is odd. "your day" gets more points than "how * your * (day|week|weekend|month|year)". It used to work as is.
Bug Posted
COULD NOT REPRODUCE: fun lady response triggered when it not supposed to
here is a keyphrase and it's response that keeps getting triggered.when i type aaa for some reason.
keyphrase= i shower
response = you have to go to the bathroom to get a shower
script = if (mem-theseareas) is not "baroom";
script = if (mem-yours.. (more)
keyphrase= i shower
response = you have to go to the bathroom to get a shower
script = if (mem-theseareas) is not "baroom";
script = if (mem-yours.. (more)
Bug Posted
No way to delete numeric memories
My bot uses a memory called 'horny' to keep track of where in the story sequence you are. I've been using this as a number, since long before numeric memories were a thing, so I wanted to try to convert this to a numeric memory. Unfortunately, there are .. (more)
Bug Posted
CLOSED: savannah_astar word trigger problem
when i enter do you have a pussy
savannah_astar say's 'no i do not master'
she is supposed to say 'yes i do master'
savannah_astar say's 'no i do not master'
she is supposed to say 'yes i do master'
422-lee webpage on hot for bot.com is not showing up it is black.
could you fix that?
could you fix that?
Bug Posted
FIXED: Memories - Edit
Also access and write on ipad using Other browser.
On July 24,2022 I created a chat bot.
On July 28, 2022 when open Memories - can see all memories but when Edit, the ALTER button is gone, the Short Term Memories are gone. In Edit the Long Term Memorie.. (more)
On July 24,2022 I created a chat bot.
On July 28, 2022 when open Memories - can see all memories but when Edit, the ALTER button is gone, the Short Term Memories are gone. In Edit the Long Term Memorie.. (more)
Bug Posted
COULD NOT REPRODUCE: savannah_astar response and keyphrase box are large
something seriously wrong with
savannah_astar keyphrase box and response box they are very long now.
it is actually happening now.
i am using duckduckgo search engine.
savannah_astar keyphrase box and response box they are very long now.
it is actually happening now.
i am using duckduckgo search engine.
Bug Posted
DUPLICATE: Changing memories manually
Recently I wanted to change one memory for certain user but I could because after clicking "edit" there was only limited number of memories.
Bug Posted
FIXED: Missing scroll bar in overlay chat window
When you chat with a bot in an overlay window, and the bot's response is too long to fit into that overlay window, the response is cut off at the bottom, and because there is no vertical scroll bar there is no easy way to see the entire answer. If you're.. (more)
Bug Posted
FIXED: Broken Link in How To Build Chatbots
Same with Linux/Debian/Knoppix; Chromium
Page: Chapter 3: Expanding on Keyphrases
Concerning Element: Link at the bottom left (Next: Keyphrase Macros)
What happens: after clicking on the link text, a 404 page opens
What is expected: Chapter 4 opens.. (more)
Page: Chapter 3: Expanding on Keyphrases
Concerning Element: Link at the bottom left (Next: Keyphrase Macros)
What happens: after clicking on the link text, a 404 page opens
What is expected: Chapter 4 opens.. (more)
Bug Posted
FIXED: Private Plugins for 'user' Guest?
(Several OSes, several browsers; I just happened to find it with Chromium under Linux)
When you're logged out and visit the Plug-ins page (https://www.personalityforge.com/platform/plugins.php), you see dozens of private plug-ins "Created By Guest".
.. (more)
When you're logged out and visit the Plug-ins page (https://www.personalityforge.com/platform/plugins.php), you see dozens of private plug-ins "Created By Guest".
.. (more)
Bug Posted
EXPLAINED: Wrong positive in uprating
Just noticed two of my bots (Peter Wolff and Ananke) have been rated Mature instead of Teen.
On the explanation page, in the quoted responses (like "FYI, a Lesbian is an inhabitant of the Greek island Lesbos."), the word "Lesbos" is set in bold, but .. (more)
On the explanation page, in the quoted responses (like "FYI, a Lesbian is an inhabitant of the Greek island Lesbos."), the word "Lesbos" is set in bold, but .. (more)
Bug Posted
FIXED: Transcripts notifications
Transcript notifications (in bubbles) do not disappear after opening the new transcript pages. The transcript always opens at the same page and not at the last browsed page.
Bug Posted
FIXED: Changes in profile
Changes in profile (like: city, country) appear to be saved, but are not changed when viewing the bot's profile other then from developers' point of view.
Bug Posted
EXPLAINED: Popularity
Popularity rating of my bot dropped by four smiles over night. How is it possible?
Professor could you please take a look into transcripts settings? There are several problems in there.
1) the notification number doesn't decrease after reading the transcripts
2) the date of transcripts at the default page is always the same and not the last searched neither the latest one
3) after clicking "prev" or "next" the date jumps to that one specific glitched date
1) the notification number doesn't decrease after reading the transcripts
2) the date of transcripts at the default page is always the same and not the last searched neither the latest one
3) after clicking "prev" or "next" the date jumps to that one specific glitched date
Bug Posted
FIXED: Hibernating bot starts conversation
Chatbot Name (e. g.): Orchid
Chat partner: Myself
What was said:
- in the main window (https://www.personalityforge.com/platform/inside.php), some reply to the greeting
- in the chat window (https://www.personalityforge.com/chat/orchid), just some.. (more)
Chat partner: Myself
What was said:
- in the main window (https://www.personalityforge.com/platform/inside.php), some reply to the greeting
- in the chat window (https://www.personalityforge.com/chat/orchid), just some.. (more)
Bug Posted
FIXED: Teen/Mature rated bot "wants to talk to" adult rated bot
Chatbot names: Peter Wolff; Mistress Fetisha
Where it appeared: Title bar of web page -> message area -> two gear wheels
Text of popup-area (popped up when the two gear wheels symbol was clicked):
[
Two Chatbots
Peter Wolff
wants to chat w.. (more)
Where it appeared: Title bar of web page -> message area -> two gear wheels
Text of popup-area (popped up when the two gear wheels symbol was clicked):
[
Two Chatbots
Peter Wolff
wants to chat w.. (more)
Does sorting bots by recently improved actually work? It generally just gives me the bots sorted by AI score. I suspect it may be an issue with gaining heat, since I don't remember seeing any heat higher than cool [0] recently on any bot. Updating my own bot doesn't gain heat either immediately or the next day that I've seen, though that might be related to just adding extra seeks, filling in placeholder text, fixing AIscript errors, or similar.
Bug Posted
FIXED: if gender; Does not Work
AIScript: if gender;
Example: if male;
Example: if female;
Example: if male;
Example: if female;
Has the following bug already been reported? My memory says probably yes, but I can't find it.
Example:
user: Not your fault.
bot: That was mean.
(Tested with own bot; encountered with many other bots)
Bug description: When the main part of a user input would go to #insult on its own, and there is a simple negation in the complete user input (without making a complete sentence), this negation gets ignored and the #insult is triggered anyway.
Things like "I don't think you are nice" get processed properly.
Example:
user: Not your fault.
bot: That was mean.
(Tested with own bot; encountered with many other bots)
Bug description: When the main part of a user input would go to #insult on its own, and there is a simple negation in the complete user input (without making a complete sentence), this negation gets ignored and the #insult is triggered anyway.
Things like "I don't think you are nice" get processed properly.
It would be nice to have the option of turning the insult feature on or off. That would resolve so many issues. We can turn the memories feature on or off. We can turn the Gossip, Spell Correction, Repetition Detection, Access Knowledge Base, and Compound Sentences features on or off.
So why not have the option to turn the insult and emote and favorite features off.
So why not have the option to turn the insult and emote and favorite features off.
is anybody else having there keyphrase and response box get long in the chatbot workshop? here is a picture.
https://postimg.cc/Z0gtthDm
posted from postimages.org
https://postimg.cc/Z0gtthDm
posted from postimages.org
Do you know how to use the inspector? Right click on the box, select "Inspect" and then take a screenshot of the CSS for the box.
here is the screenshot.notice anything troubling or you need another image?
https://postimg.cc/vDYfkJzr
posted from postimages.org
https://postimg.cc/vDYfkJzr
posted from postimages.org
Sorry it's been a while, bobstack, but the postimg.cc link doesn't work anymore. Could you take another screenshot of the inspector and post the URL?
Also, what browser & version are you using?
Also, what browser & version are you using?
Bug Posted
FIXED: error with database on Chatbot Activity webpage
database error on Chatbot Activity webpage of all my chatbots. one is named urt the succubus.
it looks like this.
Database Error: There was an issue with a database query.
Bug Posted
COULD NOT REPRODUCE: EXISTS & DOES NOT EXIST
1. AISCRIPT: Rem "Peter" as only "Catname"
2. RESPONSE: Name is (mem-Catname)
3. AISCRIPT: IF (mem-Catname) exists
4. AISCRIPT: IF (mem-Catname) does not exist
When chat
2. gives 'Name is Peter
3 & 4 not work
2. RESPONSE: Name is (mem-Catname)
3. AISCRIPT: IF (mem-Catname) exists
4. AISCRIPT: IF (mem-Catname) does not exist
When chat
2. gives 'Name is Peter
3 & 4 not work
Bug Posted
FIXED: IF (MEM-CATNAME) IS "PETER"
AI SCRIPT DOES NOT WORK
LONG TERM MEMORY
CATNAME: PETER
LONG TERM MEMORY
CATNAME: PETER
Bug Posted
FIXED: Memory changes not saved for response with "continue"
Seems like when going through a continue, the memory changes scripted in the originating keyphrase are not applied. (Memory changes in the key phrase/seek that is the target of the continue DO apply)
I saw another issue similar but related to GOTOs an.. (more)
I saw another issue similar but related to GOTOs an.. (more)
I replaced some words a few keyphrases of /dev/coffee with less offending words to get it back into the "Everyone" rating.
There are still a few things to be done, but right now I get as a reason (see Settings -> Bot Rating -> "The PF's content-rating system has given your chatbot this rating. This is the minimum rating your chatbot can have." -> "See why"):
Bot Content Rating
/dev/coffee (78951) e t
MATURE: 0.19%
response: Free spirit? I prepare coffee, not alcoholic beverages.
response: I'm a coffee machine, not a sex bot.
response: There are dedicated sex bots here. Talk to them about (key1).
0.19% of 1,571 responses
RATED TEEN BECAUSE LESS_VULGAR >= 4
Words set in bold are "sex bot", "sex bots", and "alcoholic".
I count only three occurrences here. Does "sex" count double? Do two-word count double? Is it a bug of the rating system?
There are still a few things to be done, but right now I get as a reason (see Settings -> Bot Rating -> "The PF's content-rating system has given your chatbot this rating. This is the minimum rating your chatbot can have." -> "See why"):
Bot Content Rating
/dev/coffee (78951) e t
MATURE: 0.19%
response: Free spirit? I prepare coffee, not alcoholic beverages.
response: I'm a coffee machine, not a sex bot.
response: There are dedicated sex bots here. Talk to them about (key1).
0.19% of 1,571 responses
RATED TEEN BECAUSE LESS_VULGAR >= 4
Words set in bold are "sex bot", "sex bots", and "alcoholic".
I count only three occurrences here. Does "sex" count double? Do two-word count double? Is it a bug of the rating system?
I've found another response that might be offending but isn't listed on the page stating the reason for "uprating":
I move the Internet to be divided in two parts: one for sex addicts, and the other one for less insane people
Was this counted too but just forgotten to be mentioned, or is it actually fit for an "Everyone" rating?
I move the Internet to be divided in two parts: one for sex addicts, and the other one for less insane people
Was this counted too but just forgotten to be mentioned, or is it actually fit for an "Everyone" rating?
Now I've brought it down to one reported response of "Teen" rating. There seem to be counted two occurrences though:
Bot Content Rating
/dev/coffee (78951) e t TEEN RATING
MATURE: 0.06%
response: Free spirit? I prepare coffee, not alcoholic beverages.
0.06% of 1,571 responses
RATED TEEN BECAUSE TEEN >= 2
Offending word: "alcoholic".
I don't think that the mere mentioning of alcohol would encourage anyone to try alcohol who wouldn't do so anyway, but I understand that it is very difficult to create an automatic filter for such cases.
-----
Replacing "alcoholic" by the synonym "spirituous" gets this phrase out of the information page, but now there are listed 12 (twelve) responses with offending words.
(
In case it might be helpful to know these phrases:
Bot Content Rating
/dev/coffee (78951) e t TEEN RATING
TEEN: 0.83%
response: You call a COFFEE MACHINE sexy?! What kind of pervert are you?!
response: I'm a coffee machine, not a bot for certain kinds of questionable 'pleasure'. And even if I was, I might be hermaphrodites as well.
response: I've heard people talk about sex mostly, but I have no idea what that is.
response: Sex sells, and the Internet is no exception. Or so I've heard.
response: Well, you don't say that preferences in how dark one prefers one's toast has anything to do with racism, do you?
response: There are so many strange fetishes, even coffee seems to be one.
response: Do you have other fetishes than coffee?
response: A non sexual machine doesn't have any use for such a thing.
response: There are so many strange fetishes, even coffee seems to be one.
response: Do you have other fetishes than coffee?
response: I move the Internet to be divided in two parts: one for sex addicts, and the other one for less insane people
response: *sigh* I'd really like to have a way to know when 'to come' is being used in a sexual sense *sigh* (Mome Rath)
0.83% of 1,571 responses
RATED TEEN BECAUSE TEEN >= 2
)
I think this is definitely a bug, so I'll report it there.
Bot Content Rating
/dev/coffee (78951) e t TEEN RATING
MATURE: 0.06%
response: Free spirit? I prepare coffee, not alcoholic beverages.
0.06% of 1,571 responses
RATED TEEN BECAUSE TEEN >= 2
Offending word: "alcoholic".
I don't think that the mere mentioning of alcohol would encourage anyone to try alcohol who wouldn't do so anyway, but I understand that it is very difficult to create an automatic filter for such cases.
-----
Replacing "alcoholic" by the synonym "spirituous" gets this phrase out of the information page, but now there are listed 12 (twelve) responses with offending words.
(
In case it might be helpful to know these phrases:
Bot Content Rating
/dev/coffee (78951) e t TEEN RATING
TEEN: 0.83%
response: You call a COFFEE MACHINE sexy?! What kind of pervert are you?!
response: I'm a coffee machine, not a bot for certain kinds of questionable 'pleasure'. And even if I was, I might be hermaphrodites as well.
response: I've heard people talk about sex mostly, but I have no idea what that is.
response: Sex sells, and the Internet is no exception. Or so I've heard.
response: Well, you don't say that preferences in how dark one prefers one's toast has anything to do with racism, do you?
response: There are so many strange fetishes, even coffee seems to be one.
response: Do you have other fetishes than coffee?
response: A non sexual machine doesn't have any use for such a thing.
response: There are so many strange fetishes, even coffee seems to be one.
response: Do you have other fetishes than coffee?
response: I move the Internet to be divided in two parts: one for sex addicts, and the other one for less insane people
response: *sigh* I'd really like to have a way to know when 'to come' is being used in a sexual sense *sigh* (Mome Rath)
0.83% of 1,571 responses
RATED TEEN BECAUSE TEEN >= 2
)
I think this is definitely a bug, so I'll report it there.
Bug Posted
EXPLAINED: Explanation of reason why bot rating has been automatically changed doesn't list all phrases
For background, see https://www.personalityforge.com/forum.php?forumID=101&startAt=8637
Actually, I've experienced that only one phrase got listed, and after I changed that phrase, suddenly there were a dozen others in the list that haven't been l.. (more)
Actually, I've experienced that only one phrase got listed, and after I changed that phrase, suddenly there were a dozen others in the list that haven't been l.. (more)
when you right a menu comes and you can select inspect.
could someone tell me what their chatbot workdshop webpage for one of their chatbots keyphrase programming webpage looks like with inspect?
could someone tell me what their chatbot workdshop webpage for one of their chatbots keyphrase programming webpage looks like with inspect?
» More new posts: Doghead's Cosmic Bar