Tease AI Bug Report Thread

[sexybastardo]

Hey there,
I noticed a problem with @CallReturn / @MiniScript:

Spoiler: show

----------

MiniScripts:

Not Stroking (@MiniScript from the Start Script):
Single @MiniScript: Works fine
@MiniScript within a MiniScript works, but leads to a dead end at the last line of the first MiniScript.
@CallReturn within a MiniScript works, tested with a maximum of three level deep nested @CallReturns. But leads to a dead end - it jumps instantly at the first @End to the last line of the MiniScript and stops.

Stroking (@MiniScript from Response File):
Single @Miniscript: Works fine
@Miniscript within a Miniscript works, but leads from the end in the last file instantly back to the first call. (That's maybe correct?)
@CallReturn within a Miniscript works, but leads from the end in the last file instantly back to the first call. (In this case its wrong, I guess - it should go back all steps.)


@CallReturn():

Not Stroking (Start Script):
Single @CallReturn: Works fine
@Miniscript within a CallReturn works fine. @CallReturn Script -> @MiniScript Script -> Rest of @CallReturn Script -> Back
@CallReturn within a CallReturn works fine, tested with a maximum of three level deep nested @CallReturns.

Stroking (Response File):
Single @CallReturn: Works fine
@Miniscript within a CallReturn works different than before. It jumps back to Stroking after the MiniScript.
@CallReturn within a CallReturn works, but leads from the @End in the last file instantly back to Stroking.
@Call after a @CallReturn: Seems to be working fine. Leads to a Module at the end of the last file.

Stroking & @CallReturn from a Module:
Single @CallReturn: don't work - it throws you back after the end of the @CallReturn and runs a random module
Sometimes I still receive some Stroke Taunts after the @CallReturn before the new module is started.

NOT STROKING & @CallReturn from a Module/Link:
@CallReturn: works fine.
@MiniScript: works fine.
@Call stops everything.


@Call():

Not Stroking (@Call from a Start Script):
@Miniscript: Seems to be working fine.
@CallReturn after a @Call works, tested with a maximum of three level deep nested @CallReturns, but the session will end at the last line of the Call script if no further command like @StartStroking is given.
@Call after a @Call works fine.

Stroking & @Call from a Module:
@Miniscript after a @Call it throws you back and runs the script file again that contains the original call (or maybe just a random module). (In some scripts this wasn't noticeable instantly ... it leads in most cases to an endless cycle)
@CallReturn after a @Call it throws you back after the last nested CallReturn and runs the script file again that contains the original call (or maybe just a random module). (In some scripts this wasn't noticeable instantly ... it leads in most cases to an endless cycle)
@Call after a @Call: Seems to be working fine. Leads to a Link at the end of the last file.

Stroking & @Call from response file:
Response files now don't let me use a @Call anymore, TeaseAI does ignore it without any message. Don't know whats happening there.

-------

Please notice that I didn't test everything, and I noticed in some cases it worked in others not. Mostly it depends if the user is stroking or not. But I don't have the time to figure it out more currently. I marked the lines red that (at least in my view) needs to be fixed.

I posted this before here: viewtopic.php?f=26&t=19545#p231212

[1885]

Just wanted to give a quick update on the status of all the reported bugs so far:

I noticed a problem with @CallReturn / @MiniScript:

Thank you for the detailed info on this. I'm going to set aside some time in the next day or two to take a deep look at it as you suggested in the other thread

I like to use the Maximize Media Window interface, with the side bar chat. Sometimes, during a session, the side chat locks up and I have to revert to the Default Window and continue the chat underneath the media window.

I have an issue where a line in side chat gets eaten every now and then (you can still see it if you highlight it), but I've never had it lock up completely. I'll ask around the team and see if anyone has any ideas

If a taunt with 2 or 3 lines have some image call on the first line, the image will change back to the domme on the second line.

Even without calling a image, now everyline change the domme pic in each line, what did not happens in the old versions. That its not a big problem, but when is a specific image (lesbian, online, etc), its strange.

Fixed in the next patch

In situations where Tease-AI asks you to input something.
Right after you make the input. NO MATTER WHAT
The domme writes a line even though no one exists

Fixed in the next patch

So when I call images I usually use --> @ShowImage[\Path\Path\Name.jpg]
This used to work even though the file ending was PNG or Gif.
I guess the command is more strict now?

As far as I know, @ShowImage[] always required the correct extension, I don't know why it would have worked the other way before

If one Domme image is tagged with more than tree tags all the tags get ignored.
Even worse: If you tick the tag GarmentCovering it gets tagged twice with it if you don't rename the garment in the text field on the bottom.

Can give me a specific example? I've tried using an ImageTags.txt where there were more than three tags for each picture, but it didn't affect anything. Do you mean more than three tags in the @DommeTag() Command itself? (For example, @DommeTag(Smiling, Boobs, Pussy, HalfDressed) ) @DommeTag() has only ever supported three tags, but the code for handling all Tags has been changed extensively and way above my knowledge level during my last hiatus. To the point where I can't even make sense of it really, so getting Tags untangled is going to be an undertaking. As far as I can tell though, all @DommeTag() tests I've run have worked, so any further information you can give me will be helpful

Is it normal that TAI doesn't save either chat logs, error logs, session images, I don't even see saved up variables etc... Any idea why and how to fix it? after puch update I hade to create fake files because during start up TAI wanted some gif file in session images folder and txt file ...

I had that session images thing happen to me one time, but I was able to click through the error and everything worked fine. I keep wanting to take a deeper look at it, but it hasn't happened to me again since. I'll take a deeper look at all the startup code when I get the chance

As far as nothing saving, are the folders read-only by any chance? Is the first question off the top of my head I can think of to ask

First of all, it's great that this program is still being worked on and thank you to anyone who is putting their time and energy into this.
In playing with the new functionally to use a douple comma ",," in @RT()-commands I think I found a pretty significant bug. When you start TeaseAI it automatically changes any ",," inside a @RT()-command into a simple "," thereby ruining the lines. It doesn't just show it wrong, it actually changes the scripting in the txt-file. I assume this is because of the "Auditing Scripts"-part whenever you start TeaseAI. It seems to think ",," inside a command must be a mistake and changes it to ",".

Fixed in the next patch, it was Audit Scripts causing it as you suggested. It's been changed to leave commas alone

I've removed the brackets, slash and exlamation mark from the file now, to see if that's the cause

pepsifreak fixed the timeout issue for the next patch

I'm using version 0.54.7. Just tested with only one picture for the Domme (named 01.jpg), with a tag:
01.jpg TagFullyDressed

As I mentioned above, the code for ImageTags has been changed so drastically I don't even know where to start. Please bear with me, they'll be high on the list but I'm going to have to seek help for it. I also want to isolate any problems with @DommeTags() and get those fixed as well, but Daragorn helped me narrow a couple things down on that front, so progress is being made

Great job so far, just two things I'd like to bring to your attention. One might be more to do with the scripts and the other is more of a feature request.
1) I've been noticing that after tits/ass slideshows, the screen isn't staying on the selected image but going back to the Domme pix. Is that a script thing, or a script engine thing?
2) Since updating, I've notice that there are a lot of broken images that do not display any information. If the information would show even when the image didn't; you could figure out what is broken on your machine and fix it. Because I think more likely than not it's my settings here.... But which ones!?

Even though I did the rebuild, there are still hundreds of links to the offline servers in my URL files. I'm not sure if this is a temporary Tumblr issue, or something wrong with how TeaseAI builds its URL files.

1. Fixed in the next patch
2. The next patch has a new feature that lets you fix all broken images across all URL Files quickly and easily so that they display correctly

Also, last night I tried using custom lines for Lazy Sub. (Apparently this option is only available at higher vertical resolutions. It might be helpful to toggle with Avatar window to allow for more room for apps?) I had a lot of difficulty with this. My fourth custom line kept getting over-written by the other lines. For instance, if I clicked the first line, the fourth one would change to a duplicate of the first. If I clicked the 2nd line, its text would copy the 4th. And so on. So I couldn't get the 4th line to stay the way I wanted it. A few times, after editing the customs, I was unable to get the focus back to the chat input and could no longer carry on the conversation with the domme no matter what I tried.

What resolution are you using?

However, I'm still only seeing the default Microsoft Anna. Does anyone else see their other installed voices as an option with the current patch?

Fixed in next patch

Another really small thing. Sometimes, I copy a text to put in a script, and the text use ` instead of ' .
Like this: Don`t and Don't
TeaseAI dont recognizes this and put a ? in the place. Not sure if its something just here.

I think whatever font you're using doesn't have a character for `. I tried it with Cambria and it displayed okay

Error I had today:
Date/Time: 2017-03-01 20:22:12
Version: 0.54.7.0
Title: Exception occured finding GotoLabel "(AV_NoPosi)" in file "[TEASE AI DIR]\Scripts\House of Tease\Stroke\Link\LinksRamdom11.txt"
Message: Index was out of range. Must be non-negative and less than the size of the collection.

Afther the error, all interaction stopped. Asking to go faster/slower/stop had the AI tell me I wasn't supposed to be doing anything, everything else just got no response.

It's probably something in the script, but I hope it helps everyone involved.

This has been one of the most persistent reasons for crashes in Tease AI, but @Goto() and @Interrupt() handling in these cases have been improved in the next patch, and hopefully should bring an end to these kind of crashes altogether

[Trusfrated]

Also, last night I tried using custom lines for Lazy Sub. (Apparently this option is only available at higher vertical resolutions. It might be helpful to toggle with Avatar window to allow for more room for apps?) I had a lot of difficulty with this. My fourth custom line kept getting over-written by the other lines. For instance, if I clicked the first line, the fourth one would change to a duplicate of the first. If I clicked the 2nd line, its text would copy the 4th. And so on. So I couldn't get the 4th line to stay the way I wanted it. A few times, after editing the customs, I was unable to get the focus back to the chat input and could no longer carry on the conversation with the domme no matter what I tried.

What resolution are you using?

My primary laptop display is 1366x768. I have secondary monitor (my TV) which is 1920x1080. With the laptop, I don't see anything below the "Let Me Cum!" button. (I'm speaking exclusively about the standard Lazy Sub app that has custom options, not the Lazy Sub AV option.) It kind of seems like the Shortcut fields take up a lot of extra vertical space, but I was just saying maybe as another way to allow for the whole app would be to toggle the Avatar. Not a huge deal, but I doubt I'm the only one using this resolution.

[thatwilldoit2]

Can give me a specific example? I've tried using an ImageTags.txt where there were more than three tags for each picture, but it didn't affect anything. Do you mean more than three tags in the @DommeTag() Command itself? (For example, @DommeTag(Smiling, Boobs, Pussy, HalfDressed) ) @DommeTag() has only ever supported three tags, but the code for handling all Tags has been changed extensively and way above my knowledge level during my last hiatus. To the point where I can't even make sense of it really, so getting Tags untangled is going to be an undertaking. As far as I can tell though, all @DommeTag() tests I've run have worked, so any further information you can give me will be helpful

I was talking about the ImageTags.txt
eg:
01.jpg TagAss TagFullyDressed TagCloseUp TagGlaring TagPussy TagFace
02.png TagAss TagHalfDressed TagSideView TagFace TagPussy TagPussy TagFeet
03.png TagAss TagHalfDressed TagSideView

The Images didn't get shown at all if one single image was tagged with more than three tags. So 01 and 02 in this example were never shown, but 03 would have been fine. I used to be able to reproduce this but in my test setup it worked now, I'm not sure why. Maybe the bug just shows up mid tease? But I'm positive it used to not work in my test script as well.
The situation: House of Tease didn't show me tagged images at all if there was a tagged taunt. I asked around and Genome said this:

Notice that images cant have more than 3 tags.

So I reworked my ImageTags.txt so that every image had no more than three tags (like 03 in the example) and it worked again.
Genome, were you talking about no more than 3 tags in the ImageTags.txt or the @DommeTag()? I think you need to clarify, otherwise I consider it fixed since I can't reproduce it anymore for some reason.

[sexybastardo]

Hi again.
I noticed two very small problems:

1. While running a video that has a script with the same name in /Video/Scripts that contains a @Contact1,@Contact2 or @Contact3 line, this line will be visible despite the fact the contact isn't in the room.

2. Avoid the Edge: The @EdgingHold command within the Avoid the Edge scripts don't work all the time. Mostly they just get skipped.

[ze_nordic]

Hey, I encountered a few bugs with the newest version of the software.

First, I ran into a problem where if I right-clicked the chat screen at the wrong time when the mistress was typing, a generic error popup was thrown by the application and it the script seemed to stop completely. This happened during setup of the Spicy, I will try it again later on to see if it happens in other situations as well.

Second, if you forget to set the mistress image folder (whoops...) you get a million error messages in the chat about the path being incorrect. It seems like it would be good if the program handled this situation a bit more gracefully, the script didn't stop and the program seemed to work mostly fine, though it made reading the chat impossible since it kept jumping up when new images should have been shown. I'm not completely sure if this is related to Spicy or TeaseAI, gut feeling would be it is TeaseAI itself.

Third, there was an error during that spam about liked and disliked image text files not being found, although I'm not sure if this is related to the script trying to access those files too soon for some reason or what. The error was correct though, the files weren't in the folder since this was the first time running the script and the files weren't there yet. Will report this if I encounter it again.

Thanks for your efforts devs!

[1885]

1. Random domme is not functional. Also ther is a flag to choose random. Better to make a radio: random/fixed
2. Text to speech begins to repeat on color text (for example: glitter)
3. Debug > Timers> Start timer 1 - does not work, typing "Thank you" works instead
4. Part of side chat text disappears sometimes (when switching between windows)
5. Switching windows when in Fullscreen mode shows menu. Pressing alt after that focuses cursor on name field, not chat field as before switching.
6. Please add a method to be called when last picture was reached, so domme can tell something before starting a new slideshow

Thanks for the reports! Random Dommes should be working correctly in the build that went up today, let me know if you're still having problems. I'll put the rest of these issues to look at for the next build.

First, I ran into a problem where if I right-clicked the chat screen at the wrong time when the mistress was typing, a generic error popup was thrown by the application and it the script seemed to stop completely. This happened during setup of the Spicy, I will try it again later on to see if it happens in other situations as well.

Thanks, I'll look into this

Second, if you forget to set the mistress image folder (whoops...) you get a million error messages in the chat about the path being incorrect. It seems like it would be good if the program handled this situation a bit more gracefully, the script didn't stop and the program seemed to work mostly fine, though it made reading the chat impossible since it kept jumping up when new images should have been shown. I'm not completely sure if this is related to Spicy or TeaseAI, gut feeling would be it is TeaseAI itself.

Good point, there's been some changes to the way that slideshows are assigned, and I didn't think to test how it handled no slideshow being present. I'll look into that too

Outputting error messages to screen is a new feature to help get more information during bug reports, but if it gets to verbose you can turn this off in Settings->Misc->Output Error Messages to Chat Window for anyone who doesn't know about it

Third, there was an error during that spam about liked and disliked image text files not being found, although I'm not sure if this is related to the script trying to access those files too soon for some reason or what. The error was correct though, the files weren't in the folder since this was the first time running the script and the files weren't there yet. Will report this if I encounter it again.

Scripts definitely should try to establish Liked and Disliked image files before using them, but there's a chance this is related to a bug in Tease AI that I'm still tracking down. I'll keep at it for sure

[Daragorn]

First, I ran into a problem where if I right-clicked the chat screen at the wrong time when the mistress was typing, a generic error popup was thrown by the application and it the script seemed to stop completely. This happened during setup of the Spicy, I will try it again later on to see if it happens in other situations as well.

Mmm i will try to replicate it...never had it happening to me, though (if you manage to explain me exactly all the steps to reproduce it it will be a good thing :D

Second, if you forget to set the mistress image folder (whoops...) you get a million error messages in the chat about the path being incorrect. It seems like it would be good if the program handled this situation a bit more gracefully, the script didn't stop and the program seemed to work mostly fine, though it made reading the chat impossible since it kept jumping up when new images should have been shown. I'm not completely sure if this is related to Spicy or TeaseAI, gut feeling would be it is TeaseAI itself.

This is not a bug, it is intended...in the options there is a "output error to chat" options...if you are annoyed by all those errors, set it to off...if you leave it on it will give you those messages because it is intended to do that...it should help you understand what the error is and how to solve it. But, as i said, if you don't want to see error messages, just untick the option in the settings

Third, there was an error during that spam about liked and disliked image text files not being found, although I'm not sure if this is related to the script trying to access those files too soon for some reason or what. The error was correct though, the files weren't in the folder since this was the first time running the script and the files weren't there yet. Will report this if I encounter it again.

this is not a bug either...it is a script problem, if you check for something that isn't there, it will throw errors obviously (once again, uncheck the error to chat options and you will not see those messages).

The session will not stop if those things happen, as you see, because the program can handle them....but if you have asked it to give you error messages, it will do it :D

[traedor]

This might be intended behavior, but text-to-speech only works on responses for me.

[Trusfrated]

This might be intended behavior, but text-to-speech only works on responses for me.

I'm very happy to have my full list of TTS voices back! But as tradeor said, it does seem that not *everything* the domme says is spoken. Some lines get skipped for some reason.

Also, when TTS encounters a #MISSING red vocab word it goes crazy and repeats something like HTML code ("equals fawn color red" or something like that) over and over for a minute. Then she goes to the next letter and starts over, to the point you either need to mute it for a few minutes or restart TeaseAI to stop the speaking.

All that said though, THANK you to the team! This is a great update so far!

[avatarbr]

Also, when TTS encounters a #MISSING red vocab word it goes crazy and repeats something like HTML code ("equals fawn color red" or something like that) over and over for a minute. Then she goes to the next letter and starts over, to the point you either need to mute it for a few minutes or restart TeaseAI to stop the speaking.

TTS continue working when teaseai play a sound? If not, you can create a response with some sound and type that to stop the TTS.
Or any other thing it can stop the TTS line, like a @Nullresponse?

[Oruz]

Is there any way to reset Tease AI back to original settings?

My Tease AI glitched, in a way that I don't understand, and I'd like to get it to the original settings to sort it out.
However, even if I delete the whole thing and restart, it still stays the same as before.
Help.

[holaba_be]

Is there any way to reset Tease AI back to original settings?

My Tease AI glitched, in a way that I don't understand, and I'd like to get it to the original settings to sort it out.
However, even if I delete the whole thing and restart, it still stays the same as before.
Help.

Rename the directory it's in. TAI saves settings somewhere in your %users% folder, based on the location of the executable. If you rename the installation directory, it should have the default settings.

[1885]

This might be intended behavior, but text-to-speech only works on responses for me.

I'm very happy to have my full list of TTS voices back! But as tradeor said, it does seem that not *everything* the domme says is spoken. Some lines get skipped for some reason.

Also, when TTS encounters a #MISSING red vocab word it goes crazy and repeats something like HTML code ("equals fawn color red" or something like that) over and over for a minute. Then she goes to the next letter and starts over, to the point you either need to mute it for a few minutes or restart TeaseAI to stop the speaking.

Thanks for the reports, I'll be looking into the TTS stuff over the next couple of days to try and sort it out. We're collecting the latest crop of bug reports from this build and working to get them addressed. Once they're sorted out, I'll release 54.9 as a bugfix patch (with just a couple of added features) so the wait won't be so long compared to 54.7->54.8

Is there any way to reset Tease AI back to original settings?

My Tease AI glitched, in a way that I don't understand, and I'd like to get it to the original settings to sort it out.
However, even if I delete the whole thing and restart, it still stays the same as before.

As holabe said, the actual settings are stored in the users folder. You can rename your directory, or you can go into your users folder and go to AppData->Local->Tease_AI and delete the folder(s) there

[1885]

This might be intended behavior, but text-to-speech only works on responses for me.

I'm very happy to have my full list of TTS voices back! But as tradeor said, it does seem that not *everything* the domme says is spoken. Some lines get skipped for some reason.

Are you guys encountering this with Spicy by chance? I believe that the intro sequence with Vivienne is output to the screen using System Messages in Spicy's scripts, which wouldn't get picked up by the TTS. Do you get the full expected range of TTS when using Wicked Tease, for instance?

Also, when TTS encounters a #MISSING red vocab word it goes crazy and repeats something like HTML code ("equals fawn color red" or something like that) over and over for a minute. Then she goes to the next letter and starts over, to the point you either need to mute it for a few minutes or restart TeaseAI to stop the speaking.

This is a two-fold bug. Daragorn discovered that reporting a missing or invalid #Keyword was causing the program to get stuck in a loop that took a while to sort itself it out. Meanwhile, TTS was speaking the html used code to make the error message red, which it shouldn't have been doing in the first place, but since it was it was getting repeated for every step of the loop.

Daragorn fixed the looping bug, and I've made it so TTS won't speak the html code anymore. These fixes will be in the next patch.