FNIS for Modders / FNIS 4.0 [WIP]

[fore]

these are PAs only tests:

 

toggle TFC its not possible during the animation, it will freez the cam in place where it was, when TFC command was issued, toggling it again will make the cam jump to its position in the animation.

 

this brings me to a question what is the anim event for TFC? maybe you can apply it to the behavior the same way you applied the cam[3rd] to the behavior.

 

animating the cam[3rd] in the project works perfect ingame, the cam will do exactly as you animated in the time line.

 

Let me put together what I know about those specific 3rd camera issues:

1. Camera position defined by animation. This in solved now. Between StartAnimatedCamera and EndAnimatedCamera the cam[3rd] is controlled by animations. This will be supported with FNIS 5.1

   Btw I don't think this is only good for PA/KM, and useless for simple idles. You can very well point the cam to an "interesting" place. Or for example make a "spy mod" using this. And it'S especially good as long as issue (4) is unsolved

    

2. tfc. That is not a behavior issue. When tfc is toggled, the engine switches the controls to the camera, while the behaviors continue to run. Without any need to be noticed. So a programmatic tfc would probably be a request to the SKSE guys

    

3. cam pan when controls are inactive.i think this is a pure engine thing. This camera move takes care of the place where the player is, and that cannot be done by behaviors and animations. And since the player simply continues his idle, there is no need to inform behaviors. Similar to tfc.

    

4. Camera control while player idles. This is a total mystery to me since I started FNIS. I'm aware it is not possible to control camera during FNIS idles. But it is possible with vanilla non-equipped player idles, including sneak. However it's not possible in game either with any type of weapon or spell equipped idles. So the game must be aware of these differences. But I just can't figure out what the criterium is. Due to the different types of idles where it's possible or not I have identified the critical area around the mt_behavior reference interface to idlebehavior. But I can't see what could be the indicator for the engine. There is no apparent AnimEvent which points into that direction.

    

[Nshell]

trying to follow or understand your example for a "spy mod", to control the cam[3rd] during normal fnis idles in case the PC is the target of an idle, to override the cam[3rd] to a "interesting" place or position.

I dont know how this can be done or how it would be possible.

 

is this possible with?

 

6. Arm Offset Animations (ofa)
Example:
ofa -o MyBoundOffset MyBoudAnimation.hkx MyCuffs
In Skyrim terminology, offset animations are animations which are blended into other currently running
character animations. Due to behavior restrictions this blending cannot be done between any animation
of the game. Technically speaking, the offset animation has to be defined in the behavior graph in a
branch which is parallel to the branch which includes all animations the offset animation will be blended
into. Which restricts the offset possibilities quite heavily.
FNIS implements additional arm offset animations. With that, the arm position can be changed while
walking, running, turning, showrace, talking, standing. Used for basket holding, as well as for bonds.
Other base animations will will break the offset animation.
Because they make no sense, all options but "-o" and the AnimObjects are ignored.

 

im only questioning about this because all animations created so far, the animators never edited the cam[3rd] in their animation(not that i know off) if from now on this changes of startanimatedcamera to endanimatedcamera are applied their idle will need other animation blended to animate only the cam[3rd] to a "interesting" place or position.

[Ashal]

Is there any good way in game to check if people have run FNIS for Users?

 

I was hoping AnimVar's would do the trick, so I added this to my behavior file and ran FNIS for Modders;

AnimVar SexLabVer INT32 146

But from testing it seems that gets pulled into game regardless if I run FNIS for Users or not, making it more or less useless in that respect, though does serve as a redundant way to check for FNIS being installed.

 

(take a guess at what annoying overly posted tech support question I'm trying to get to shut up)

[fore]

Is there any good way in game to check if people have run FNIS for Users?

 

I was hoping AnimVar's would do the trick, so I added this to my behavior file and ran FNIS for Modders;

AnimVar SexLabVer INT32 146

But from testing it seems that gets pulled into game regardless if I run FNIS for Users or not, making it more or less useless in that respect, though does serve as a redundant way to check for FNIS being installed.

 

(take a guess at what annoying overly posted tech support question I'm trying to get to shut up)

 

How about adding another AnimVar to 0_master with the user generator

AnimVar FNISGen INT32 1404070921

Would this be helpful? Something a quest can at least write to the Papyrus log.

 

But I don't understand what you mean in your example. This AnimVar is only pulled into the game when you run FNIS. Because it's included in your Sexlab behavior, and that is only known to the game when it's connected to the 0_master (by the generator).

 

When you run GetAnimationVariableInt("SexLabVar") without being part of an included behavior, you will always get "0". And an error in the Papyrus log.

 

So the advantage of a "FNISGen" definition would only be that one actually knows the time of the last generator run. But if you think it'S of help I can add it to FNIS 5.1. Coming soon.

 

EDIT: another version thing: could you please add a version line to your AnimLists? Like

Version V1.46

 

So this version will appear on the generator output, instead of "V?.?". I think that will help you with your help desk as well. Right now the version line will be rejected by the modder generator, so you have to comment it out when you generate the Sexlab behavior. Will be fixed with 5.1 

 

 

[fore]

trying to follow or understand your example for a "spy mod", to control the cam[3rd] during normal fnis idles in case the PC is the target of an idle, to override the cam[3rd] to a "interesting" place or position.

I dont know how this can be done or how it would be possible.

 

is this possible with?

 

6. Arm Offset Animations (ofa)

Example:

ofa -o MyBoundOffset MyBoudAnimation.hkx MyCuffs

In Skyrim terminology, offset animations are animations which are blended into other currently running

character animations. Due to behavior restrictions this blending cannot be done between any animation

of the game. Technically speaking, the offset animation has to be defined in the behavior graph in a

branch which is parallel to the branch which includes all animations the offset animation will be blended

into. Which restricts the offset possibilities quite heavily.

FNIS implements additional arm offset animations. With that, the arm position can be changed while

walking, running, turning, showrace, talking, standing. Used for basket holding, as well as for bonds.

Other base animations will will break the offset animation.

Because they make no sense, all options but "-o" and the AnimObjects are ignored.

 

im only questioning about this because all animations created so far, the animators never edited the cam[3rd] in their animation(not that i know off) if from now on this changes of startanimatedcamera to endanimatedcamera are applied their idle will need other animation blended to animate only the cam[3rd] to a "interesting" place or position.

 

Sorry Nshell, I totally missed it.

 

ofa wouldn't be of any help, because that only works under certain conditions. For example not when weapons are unsheathed.

 

I was simply thinking about an idle, or an SA, where the player looks into a certain direction. And with that (animatedcam) idle  the cam would pull away from the player to some point, and rotate there. Nothing fancy, since you cannot parameterize where the cam goes. So you might need several idles, and somehow activate the correct one (from several ones in an SA) with certain keys. 

 

Of course, such idles don't exist yet. But since the player doesn't need do any movement during this time, such animation would be rather simple.

 

IF this is possible at all with the engine. Because I tried something similar in Oblivion. Animations which would allow looking back for enemies. And there it wasn't possible, since the engine resticted the angle of the camera to less than 180° forward.

[zax]

Bump! We need an FNIS Development Report.

[fore]

It's only a matter of some more hours. The question is when I can scrape them together. It's spring time, and my premier hobby takes its toll. 

 

Both FNIS 5.1 and my new mod, FNIS Sexy Move, are implemented. I just need some final testing, cleaning up scripts, documentation. FNIS 5.1 will add 2 new features

1. Alternate Animations for walk/run. Originally I wanted to add AAs for the (standing) idles and for sit. But I can't to postpone the relaese any more for that
2. Animated Camera. For each animation an option "-ac" can be set which allows the modder to animate the camera through the "camera3d [cam3]" bone.

FNIS Sexy Move will randomly assign different walk and run animations to all female NPCs. Through MCM and dialogue it will also be possible to assign certain moves to the player and specific NPCs, and change the assignment distribution. It will include animations from xp32, and DragonFly, and the good old catwalks.

[ZaZ]

It's only a matter of some more hours. The question is when I can scrape them together. It's spring time, and my premier hobby takes its toll. 

 

Both FNIS 5.1 and my new mod, FNIS Sexy Move, are implemented. I just need some final testing, cleaning up scripts, documentation. FNIS 5.1 will add 2 new features

1. Alternate Animations for walk/run. Originally I wanted to add AAs for the (standing) idles and for sit. But I can't to postpone the relaese any more for that
2. Animated Camera. For each animation an option "-ac" can be set which allows the modder to animate the camera through the "camera3d [cam3]" bone.

FNIS Sexy Move will randomly assign different walk and run animations to all female NPCs. Through MCM and dialogue it will also be possible to assign certain moves to the player and specific NPCs, and change the assignment distribution. It will include animations from xp32, and DragonFly, and the good old catwalks.

 

hehe been waiting for this one , Keep up the good work . So it will be a 'ToggleSpecialAnim' thingy     but just for skyrim. ?

[zax]

Will AA's be controlled by graph variables? This feature sounds like the next big thing. Can't wait.

[fore]

Yes. I'm using conditional Transitions. And there you can use hardly anything but AnimVars. 

 

THis would be a neat thing, but has one big disadvantage: AnimVar changes survive neither save/load nor a character's unload 3d. So it's always necessary to have a work-around for that, like the good old coins, and reload the animvar when necessary.

[xaz]

I just tried the versioning system in FNIS, and discovered that my reported version is still stuck at the version I created it with. (reported 4.1.1.3 in my case)

 

The animation vars in the FNIS_FNISBase_list.txt file says it should be 5.0.2.3.

 

Is anyone using this feature, and is it working for everyone? Also, how come the fairly complex system instead of just keeping a single integer (say) 5023 and update it to whatever is suitable for that version on each release?

[fore]

The reported version of what: FNIS base?  And reported at run-time by SL?

If you see 4.1.1.3, then you definitely have this version. Because AnimVars are not stored in save files.

 

And keep in mind: there are 2 version systems.

One at generation time, so you can see the version in the generator. Define by the following AnimList line:

Version V5.0.2

 

And one for run-time, so that it can be interpreted by SL and others:

AnimVar FNISMajor INT32 5

AnimVar FNISMinor1 INT32 0

AnimVar FNISMinor2 INT32 2

AnimVar FNISFlags INT32 3

 

And yes, it's a little complicated. But the reason is that papyrus has no string AnimVars. And dealing with integer version numbers like 500023 had been very clumsy as well. Because you don't want to restrict parts of the version number to just one digit. What are Sexleb's and Zaz' version numbers, for example??? 

 

Btw. I didn't write the versioning part. It was jbezorg. But I think he did it right.

 

 

[xaz]

The reported version of what: FNIS base?  And reported at run-time by SL?

If you see 4.1.1.3, then you definitely have this version. Because AnimVars are not stored in save files.

 

And keep in mind: there are 2 version systems.

One at generation time, so you can see the version in the generator. Define by the following AnimList line:

Version V5.0.2

 

And one for run-time, so that it can be interpreted by SL and others:

AnimVar FNISMajor INT32 5

AnimVar FNISMinor1 INT32 0

AnimVar FNISMinor2 INT32 2

AnimVar FNISFlags INT32 3

 

And yes, it's a little complicated. But the reason is that papyrus has no string AnimVars. And dealing with integer version numbers like 500023 had been very clumsy as well. Because you don't want to restrict parts of the version number to just one digit. What are Sexleb's and Zaz' version numbers, for example??? 

 

Btw. I didn't write the versioning part. It was jbezorg. But I think he did it right.

First five lines of my FNIS_FNISBase_List.txt. I most definitely does not have the 4.1.1.3 version, because it crashes when multiple mods define offset animations.

 

Version V5.0.2
AnimVar FNISMajor INT32 5
AnimVar FNISMinor1 INT32 0
AnimVar FNISMinor2 INT32 2
AnimVar FNISFlags INT32 3

When I run the function to get the string number, it returns 4.1.1.3.

 

Have you used the function and tried that it works? If those numbers are not stored in the save, then are those numbers compiled into other files or similar? Or is it read from that one file at run time? If I run GenerateFNISforModders, can I overwrite settings that way? Anything else you could think of that could affect the reported version?

 

I'm quite sure that the function does not work (on my setup), and I think it could be important to know why. After all, if you have a version system, it has to work.

 

Edit: Further investigation,... loading a later (as in saved at a later date, after upgrading FNIS) save, with no other change to my setup, returns 4.0.2.3.

 

Edit2: A new game returns 0.0.0.0 for me. Again, no other changes to my setup. There are a lot of errors about this in the log, though...

I should mention that I didn't properly create the new game. Rather I started Skyrim, then immediately "coc WhiterunOrigin". I'll try a "real" new game.

 

Code to reproduce:

Debug.Notification("FNIS: " + FNIS.VersionToString())
Debug.Trace("FNIS: " + FNIS.VersionToString())

[fore]

Yes it works. I'm not using it regularly, but had it working for a version check in the first 5.0 releases.

 

So the only explanation I have is, that you have another mod which has copied the old FNISBase contents into it'S own AnimList. Maybe one of your own test mods??

 

I mean, you can go into FNIS.psc, and add a message box at the place where the AnimVars are read. In FNIS.VersionToString()

 

The 0.0.0.0 is an issue I have described in the modders' doc (i hope). Animvars can only be read when it's 3Dloaded (for NPCs), or in 3rd person (for the player). Just the way behaviors work. So if you use it on the player, you have to call Game.ForceThirdPerson() first.

 

[xaz]

Yes it works. I'm not using it regularly, but had it working for a version check in the first 5.0 releases.

 

So the only explanation I have is, that you have another mod which has copied the old FNISBase contents into it'S own AnimList. Maybe one of your own test mods??

 

I mean, you can go into FNIS.psc, and add a message box at the place where the AnimVars are read. In FNIS.VersionToString()

 

The 0.0.0.0 is an issue I have described in the modders' doc (i hope). Animvars can only be read when it's 3Dloaded (for NPCs), or in 3rd person (for the player). Just the way behaviors work. So if you use it on the player, you have to call Game.ForceThirdPerson() first.

If a mod copied those into the AnimList, I would still get the same result each time I called the function, right? Wrong answer, but probably the same result?

 

Because this is not what I'm seeing. .... I get different, but consistent results, depending on the save I'm loading.

 

On a new save I created (going to 3rd person as per your suggestion), I end up with 5.0.0.2. On an old save I get 4.1.1.3. On a newer save I get 4.0.2.3.

 

Please note that the only thing I'm doing is loading different saves. I'm not leaving Skyrim, changing mods, or anything else. Just loading another save ....

 

Edit: Also, the only reference I can find to FNIS versions or the FNIS animation variables in my setup are commented out lines in SexLab which check the FNIS major version (directly checking the major version variable, interestingly).

 

Not sure why they're commented out though. Have to ask Ashal about that ...

[zax]

On a new save I created (going to 3rd person as per your suggestion), I end up with 5.0.0.2. On an old save I get 4.1.1.3. On a newer save I get 4.0.2.3.

 

Pure guess here, but assuming that these variables are stored and read from players graph, it looks like they havent updated to the new version?

 

If it was an npc it could easily be fixed by changing cells (they get unloaded then loaded with clean graph), but as far as I know player never gets unloaded, so it would require updating graph manually.

 

@xaz, you could try changing race (showracemenu) of one of your old characters then checking FNIS version again.

 

[xaz]

 

On a new save I created (going to 3rd person as per your suggestion), I end up with 5.0.0.2. On an old save I get 4.1.1.3. On a newer save I get 4.0.2.3.

 

Pure guess here, but assuming that these variables are stored and read from players graph, it looks like they havent updated to the new version?

 

If it was an npc it could easily be fixed by changing cells (they get unloaded then loaded with clean graph), but as far as I know player never gets unloaded, so it would require updating graph manually.

 

@xaz, you could try changing race (showracemenu) of one of your old characters then checking FNIS version again.

 

After changing race, I got a reported version of 5.0.0.2, instead of 4.0.2.3. This is still not correct.

 

I don't think that FNIS right now has the optimal system of reporting version.

 

I'm thinking that version should be correctly reported without having to load 3d models, changing race and so on. Also, .... I'm still not getting the right version on my setup, and I'm not confident that the problem is contained in my setup alone. I mean: Why else would I get the same version on the same save, but different versions on different saves.

 

The following code will correctly report versions for any mod asking. If one wants more digits for whatever minor or major slot, just add another zero in there.

Int Function GetVersion() Global
  Return 5023
EndFunction

String Function GetVersionStr() Global
  Return "5.0.2"
EndFunction

[zax]

Why else would I get the same version on the same save, but different versions on different saves.

This is caused by player never unloading, thus not giving chance for the new variables to update.

 

 

fore would need to implement some kind of OnVersionUpdate function that would:

kPlayer.SetAnimationVariableInt("iFNIS_major", 5) ;etc

(assuming that its stored on player) all necessary variables to their new default values.

 

But having such script would defeat the purpose of saving everything on graph, wouldnt it?

[xaz]

 

Why else would I get the same version on the same save, but different versions on different saves.

This is caused by player never unloading, thus not giving chance for the new variables to update.

 

 

fore would need to implement some kind of OnVersionUpdate function that would:

kPlayer.SetAnimationVariableInt("iFNIS_major", 5) ;etc

(assuming that its stored on player) all necessary variables to their new default values.

 

But having such script would defeat the purpose of saving everything on graph, wouldnt it?

 

But that still doesn't explain why I don't get the expected values on a new game.

 

Anyway, if there are a lot of "ifs and buts" on updating animation variables and making sure that they're up to date, then they defeat their own purpose. Just make it a global function which returns an int, and you're done.

 

Super easy to use, understand, and it always works.

[fore]

 

Yes it works. I'm not using it regularly, but had it working for a version check in the first 5.0 releases.

 

So the only explanation I have is, that you have another mod which has copied the old FNISBase contents into it'S own AnimList. Maybe one of your own test mods??

 

I mean, you can go into FNIS.psc, and add a message box at the place where the AnimVars are read. In FNIS.VersionToString()

 

The 0.0.0.0 is an issue I have described in the modders' doc (i hope). Animvars can only be read when it's 3Dloaded (for NPCs), or in 3rd person (for the player). Just the way behaviors work. So if you use it on the player, you have to call Game.ForceThirdPerson() first.

If a mod copied those into the AnimList, I would still get the same result each time I called the function, right? Wrong answer, but probably the same result?

 

Because this is not what I'm seeing. .... I get different, but consistent results, depending on the save I'm loading.

 

On a new save I created (going to 3rd person as per your suggestion), I end up with 5.0.0.2. On an old save I get 4.1.1.3. On a newer save I get 4.0.2.3.

 

Please note that the only thing I'm doing is loading different saves. I'm not leaving Skyrim, changing mods, or anything else. Just loading another save ....

 

Edit: Also, the only reference I can find to FNIS versions or the FNIS animation variables in my setup are commented out lines in SexLab which check the FNIS major version (directly checking the major version variable, interestingly).

 

Not sure why they're commented out though. Have to ask Ashal about that ...

 

 

xaz, are you using MO? Can this be the result of using different profiles?

 

This just came into my mind when I thought about a post in my Nexus thread. There is a generator output for FNIS 5.0.2, but the version showing for FNISBase is V?.?. This can only be when the generator gets an FNIS_FNISBase_List.txt which is older then 5.0. That's the same thing what you see. Only at generator time. And that has nothing to do with save.

[xaz]

 

 

Yes it works. I'm not using it regularly, but had it working for a version check in the first 5.0 releases.

 

So the only explanation I have is, that you have another mod which has copied the old FNISBase contents into it'S own AnimList. Maybe one of your own test mods??

 

I mean, you can go into FNIS.psc, and add a message box at the place where the AnimVars are read. In FNIS.VersionToString()

 

The 0.0.0.0 is an issue I have described in the modders' doc (i hope). Animvars can only be read when it's 3Dloaded (for NPCs), or in 3rd person (for the player). Just the way behaviors work. So if you use it on the player, you have to call Game.ForceThirdPerson() first.

If a mod copied those into the AnimList, I would still get the same result each time I called the function, right? Wrong answer, but probably the same result?

 

Because this is not what I'm seeing. .... I get different, but consistent results, depending on the save I'm loading.

 

On a new save I created (going to 3rd person as per your suggestion), I end up with 5.0.0.2. On an old save I get 4.1.1.3. On a newer save I get 4.0.2.3.

 

Please note that the only thing I'm doing is loading different saves. I'm not leaving Skyrim, changing mods, or anything else. Just loading another save ....

 

Edit: Also, the only reference I can find to FNIS versions or the FNIS animation variables in my setup are commented out lines in SexLab which check the FNIS major version (directly checking the major version variable, interestingly).

 

Not sure why they're commented out though. Have to ask Ashal about that ...

 

 

xaz, are you using MO? Can this be the result of using different profiles?

 

This just came into my mind when I thought about a post in my Nexus thread. There is a generator output for FNIS 5.0.2, but the version showing for FNISBase is V?.?. This can only be when the generator gets an FNIS_FNISBase_List.txt which is older then 5.0. That's the same thing what you see. Only at generator time. And that has nothing to do with save.

 

No, I'm not using MO. I use Wrye for Skyrim.

 

When I run the generator I see the correct version numbers as expected.

 

The problem I'm talking about is inside Skyrim though.

None of the output lines from the script lines above have yet matched "5.0.2.3" which is the expected output from my configuration.

 

I do get consistent results, that is my quicksave always returns "4.0.2.3", an older save always returns "4.1.1.3", and if I create a new character (or change race with "showracemenu") I get "5.0.0.2".

 

It's clear that the version system you use doesn't work on my setup. Possibly not on any setup. And none of us have any idea why that is.

 

Edit: My point is that if the system is brittle, unreliable or plain doesn't work, I can't use it as a modder. Imagine if I put in a check in Zaz to make sure the right FNIS version is used, and even 10% of all users had dialogues popping up when loading a save. .... and I had no idea how to fix it, or why it happened.

[fore]

Stop and calm down, xaz. 

 

The versioning is reliable. Isn't it used by Sexlab since long. In fact, this system was implemented by jbezorg about one year ago, and Ashal was involved at one point.

 

Purpose of the system is NOT to present the version of the generator, but the version of the FNIS data (AnimList, behavior file). If I would store the version in a script, like suggested before, it would not get the intended information.

 

Now look at all these version numbers you get. They are not just ANY version numbers, these are version numbers that were stored into the FNIS_FNISBase_Behavior.hkx at one time. 4.0.2.3 was the very first FNIS release with version, 4.1.1.3 the next. 5.0.0.2 was the first beta. So why will the system show you these old values?

 

It is DEFINITELY NOT because saves include AnimVars. You can believe I know what I'm saying. I have to implement my new FNSI Sexy Move around this stupid effect. AnimVars are gone when you load a save. So where do you get your (apparently meaningful) version numbers?

 

The only possible explanation: from an old FNIS_FNISBase_Behavior.hkx.

 

So with MO, if you use an old save, based on a profile with an old FNIS_FNISBase_Behavior.hkx, it's explainable. In fact, it's a meaningful result. You don't want to know what your current FNIS installation is, but what it was when you ran the generator for that save. 

 

Now you are the first one I hear that uses Wrye. I have now idea how Wrye works. But I know Wrye will reset the old configuration when you uninstall a mod. Is this similar to a profile? I think Wrye is the key to your issue.

 

One more question: you said you got 5.0.0.2 for a "new game". You don't start all new from the beginning in Helgen, did you? If so, are you sure that FNIS_FNISBase_Behavior.hkx is the most recent one? Re-install that.

 

EDIT: Re-considering all this it's more plausible not to use the installation information from FNISBase for a version, but instead store the same data as GENERATION information into 0_master, and then read it from there. 

But I don't think it will change anything with your problem, xaz. It's just seems more plausible to know the version (maybe also the time?) when the behaviors where generated. And maybe this time data can shed light into your problem.

[zax]

Just did a quick test, graph variables for player are stored in save.

[xaz]

Stop and calm down, xaz. 

 

The versioning is reliable. Isn't it used by Sexlab since long. In fact, this system was implemented by jbezorg about one year ago, and Ashal was involved at one point.

 

Purpose of the system is NOT to present the version of the generator, but the version of the FNIS data (AnimList, behavior file). If I would store the version in a script, like suggested before, it would not get the intended information.

 

Now look at all these version numbers you get. They are not just ANY version numbers, these are version numbers that were stored into the FNIS_FNISBase_Behavior.hkx at one time. 4.0.2.3 was the very first FNIS release with version, 4.1.1.3 the next. 5.0.0.2 was the first beta. So why will the system show you these old values?

 

It is DEFINITELY NOT because saves include AnimVars. You can believe I know what I'm saying. I have to implement my new FNSI Sexy Move around this stupid effect. AnimVars are gone when you load a save. So where do you get your (apparently meaningful) version numbers?

 

The only possible explanation: from an old FNIS_FNISBase_Behavior.hkx.

 

So with MO, if you use an old save, based on a profile with an old FNIS_FNISBase_Behavior.hkx, it's explainable. In fact, it's a meaningful result. You don't want to know what your current FNIS installation is, but what it was when you ran the generator for that save. 

 

Now you are the first one I hear that uses Wrye. I have now idea how Wrye works. But I know Wrye will reset the old configuration when you uninstall a mod. Is this similar to a profile? I think Wrye is the key to your issue.

 

One more question: you said you got 5.0.0.2 for a "new game". You don't start all new from the beginning in Helgen, did you? If so, are you sure that FNIS_FNISBase_Behavior.hkx is the most recent one? Re-install that.

 

EDIT: Re-considering all this it's more plausible not to use the installation information from FNISBase for a version, but instead store the same data as GENERATION information into 0_master, and then read it from there. 

But I don't think it will change anything with your problem, xaz. It's just seems more plausible to know the version (maybe also the time?) when the behaviors where generated. And maybe this time data can shed light into your problem.

At first I was going to answer: "I am calm", but I realize that is not how this is going to come out, so I won't. However, don't construct my post as angry or unhappy. I'm super impressed and happy with what you've done for the Skyrim mod community (and helping me and Chris get things working). Don't let anything I say write you think otherwise.

 

However, ... and again, remember what I wrote above, because this is probably going to sound confrontational. This is not my intention, but I need to somehow get across the issue that I'm seeing.

 

And before you jump on me and suggest that maybe I'm switching files around (in MO or whatever), please understand that I'm really making an effort here to understand and debug this problem. If I knew the answer, I would post it. And again, if it was super easy, like forgetting to run some tool. Odds are I would have found that too.

 

Essentially: I've made an effort to try and nail down the problem before writing here. I still spend time trying to figure out why this isn't working, and why noone has noticed yet.

 

So, to start:

If I had a mod manager that switched files around and stuff, I would know this by now. I've been modding for well more than a year, and I know the basics of how my setup works with files and saves. Nothing changes when I load a save in the Skyrim menu. I'm not leaving the game to load a new save, yet I get consistent but different results depending on the save I load. This suggests that animation variables are stored in the save, but I think you know these things way better than I do.

 

The problem:

Let's say that something was off in my configuration, I used the wrong mod manager, whatever could be the reason, ... then how would this reflect on the stability of the selected solution? There are options out there that are stable in regards to the selected mod manager. There are also options out there that are predictable, where you can debug the problem by simply asking the user to look inside a file, or maybe just the papyrus log file. It's quite clear that both me and you are clueless as to why I get the wrong version numbers when calling the scripts.

 

The solution:

This, I guess, is where it gets sensitive. I can't see any reason for storing the mod version in animation variables to begin with. What would be the advantage? Not having to recompile the script on each version, is my guess? Is that the reason? Consider the alternative: If, assuming my problem is somehow exotic which I don't believe at the moment, this happens to a few other people: How would you solve this? None of us have yet understood where this issue comes from in my setup. I think it would be wise to either understand why this happens to me, so you can help other people, or move to another solution that you could debug (or wouldn't have to).

 

 

 

Again, I don't want to turn this into some kind of confrontation, but if you again suggest that I'm changing the hkx files or using MO, I'm probably going to start crying and go all emo over this thread.

 

Seriously though, let's just find the problem, understand why this happens, and solve the versioning issue of FNIS. It really appears to associate different versions with different saves, and I'm at my wits end on what I could be doing wrong in order to affect this. I am really, really, really, trying not to alter anything in my setup when loading a different save. In fact, I'm not even sure how I could be changing any files without alt-tabbing back out to windows....

[xaz]

Just did a quick test, graph variables for player are stored in save.

Interesting! This would explain why I get weird results.

 

How did you test it?