NVSE

[jaam]

IsWaiting is a vanilla function

 

Why do you keep prefixing your ints with $ when outputting them with print ?

 

[Guest tomm434]

IsWaiting is a vanilla function

wow I didn't even know this function before you told me. Seems risky to use for me. But I think that compiler should pick this up - it compiles only with "let"

 

Why do you keep prefixing your ints with $ when outputting them with print ?

 

Well, I didn't know "Print" existed before I saw Odessa's debug script. Before I used "PrintC". But simple "Print" is more convenient.

[jaam]

ok , but why the $ ?

[prideslayer]

I think it's just unclear when you need it and when you don't, better safe than sorry?

 

I don't know the rules either, though I know it's needed (for some reason) when passing a string var into NX functions.

[Guest tomm434]

I saw that in Odessa's script.  So it's not necesary, I get it, thanks.    :)

[jaam]

$ applies to stringVar, its an alias for ToString. Does it also work for converting integer to string ?

[DoctaSax]

Numbers, form names, sure.

[jaam]

Yeah, I never realized , though it's plain in the source

[Odessa]

@AJ: There is a function to suppress NVSE errors actually: From nvse_whatsnew.txt:

 

 

 

testexpr            (did I test that one ? Can't remember )

 

From OBSE docs:

 

 

TestExpr - attempts to evaluate an expression, returning false if an error occurs during evaluation and true otherwise. This can be useful when checking if an array index is out of bounds or if an element exists with a given key, among other things. TestExpr suppresses the output of error messages to the console and log file.
(bool) TestExpr expr

Example:

[code]array_var arraylet array := ar_Construct Arrayif testexpr (array[5] := 2)	PrintC "5 is a valid index, assignment succeeded."else	PrintC "Index 5 is out of bounds, no assignment."endiflet array := ar_Construct StringMapif testexpr (array["INDEX"])	PrintC "An element exists in array with key 'INDEX'"else	PrintC "No element with the specified key exists."endif

[/code]

 

 

 

 

 

 

 

So it returns false if an NVSE error is generated (and suppresses it), else true. I guess you might use it instead of bothering to check if a key exists, like try/except in some languages:

    if TestExpr MyMapArray[someKey] += 1 ; * increment entry if it exists already

    else

        let MyMapArray[someKey] := 1 ; * else create it at 1

    endif

 

Or to write pretentious 1-line loops:

    while TestExpr MyArray[iKey+=1] += 1

    loop

[jaam]

I think DoctaSax did test testexpr, isn't it Doc ?

[DoctaSax]

I don't think I did much with it, actually.

[jaam]

My bad, one more thing to verify

 

[Odessa]

So, V4.1 added this function:

 

RunScript

 

It is not documented apart from name in nvse_whatsnew, and is not mentioned in the OBSE docs. The GECK compiler throws an error unless you pass it an inventory item base form, like "RunScript Vodka". I haven't tested it in game.

 

What does it do?

[jaam]

It requires a parameter which is a form containing a script.

 

If the script is found it will be run, as in added to the list of active scripts. This should let you start a script that is not attached to any reference. As far as I know, the script will keep running until it crashes.

[DoctaSax]

Is that limited to a particular script type?

[jaam]

Those that are implemented using the ScriptForm form. That includes, actors, activators, Inventory Items, Containers, Doors, Lights, Item Mod, Quest, Armor AddOn,

 

Spells and effects should logically work too, but they are not sufficiently decoded to guaranty it.

 

[A.J.]

@AJ: There is a function to suppress NVSE errors actually

 

I was hoping in something more... easier, something like "DontShowMeErrorsOnConsoleForThatMod", better with a shortcut like DSME... I can't even get the OBSE example

[jaam]

News from the front:

 

I think Odessa's issue should be fixed in the upcoming version. Turned out to a concurrency issue when using too many strings/arrays in loops.

 

@A.J. the logic behind those messages is to allow end users to report bugs to mod author. If you routinely generate those error with "non bugged" code, please upload a sample.

 

[Odessa]

@AJ: I've written an article on GECK wiki about TestExpr, hopefully it will make sense . Suppressing errors for an entire mod would be very bad practise, but TestExpr lets you tell NVSE "I know this line might fail, but tell me and not the user if it does, and I'll handle it". I've tested TestExpr a bit and it seems to work as intended.

[A.J.]

 

If you routinely generate those error with "non bugged" code, please upload a sample.

 

It's not really "not bugged", they do good, but I don't need them in these cases.

 

I noticed Get functions return an error on console when the ref doesn't exist.

 

For example, I have a script that checks all the slots for equipped object but it spams an error if there's nothing in the slot. I don't need it, I'm actually checking what is equipped.

 

Another example, I have a script that spawns a creature with PlaceAtMe and then executes the code on its ref. It needs "some time" to be spawned, so the script will constantly check with If GetType 43 when the creature is really spawned. But in the meantime, for some frames, it returns a error on console of course, it's the time the creature needs to be spawned.

 

They do good, they are right to notify the error, but I'm aware of those errors and I find it's normal, that's why I wanted to remove them, because I guess if all the mods work in this way all togheter it would be impossible to track down "real" errors in console

 

@Odessa thank you I'll check it

[Odessa]

RunScript does not appear to work.

 

I have the following object script attached to a misc item:

 

 

 

scn nvserunscript

ref rMe
float fTimer

Begin GameMode

    if fTimer > 0
        let fTimer -= GetSecondsPassed
        return
    endif

    let fTimer := 2

    let rMe := GetCurrentScript
    Print "Runscript GameMode " + $rMe
    Print "or " + (GetFormIDString rMe)

End

 

 

 

RunScript nvsetestToken

 

Does not appear to do anything. If I instead additem the token, the script runs as expected.

 

Note: I only tested it for documentation purposes, so take this as informational not as a request to have it fixed, if it is broken.

[Odessa]

The function IsPlayable is broken, and appears to return 1 to everything.

 

For example:

 

 

   Print "Testing IsPlayable AmmoCompanion"
   let iInt := IsPlayable AmmoCompanion
   Print "Returned: " + $iInt

   Print "Testing IsPlayable WeapKnifeCombatCass"
   let iInt := IsPlayable WeapKnifeCombatCass
   Print "Returned: " + $iInt

   Print "Testing IsPlayable Weap10mmPistol"
   let iInt := IsPlayable Weap10mmPistol
   Print "Returned: " + $iInt

Testing IsPlayable AmmoCompanion
Returned: 1
Testing IsPlayable WeapKnifeCombatCass
Returned: 1
Testing IsPlayable Weap10mmPistol
Returned: 1

 

 

 

EDIT: Actually, IsPlayable just only works for armor (see below). I have tested it for this and it works correctly

[Guest luthienanarion]

IsPlayable and SetIsPlayable only work on ARMO forms. Perhaps they should be expanded to work with more form types.

[jaam]

ARMO use a specific way to store the information.

[Odessa]

Thanks. It looks like you can use the GetWeaponFlags1 bitmask to check if a weapon is playable already. It would be convenient if IsPlayable was expanded to work for other types though...