IWD2: ARE version 9.1 info?

[FredSRichardson]

I've started working with the iwg2 code and I noticed that there's no info in IESDP for 9.1 version ARE files. I'd be happy to fill in a template with some of this fields fleshed out if needed (using iwg2's sources and Near Infinity), or if this info is available somewhere else could someone send me a pointer?

[Avenger]

Take a look at the dltcep source files

There is a complete list of fileformats from iesdp and improved by me.

 

Btw, there is not much of a difference, only the header is a bit larger (16 bytes).

And it is either unknown, or unused (all known instances are filled with zero).

 

The 16 bytes are inserted before the offsets part of the header (0x54).

[FredSRichardson]

Thanks Avenger, I'll take a look

 

Is IESDP going to be updated with that stuff? Or does it need to go through more validation beforehand?

[SimDing0]

DLTCEP tends to be vastly more reliable than IESDP.

[Avenger]

DLTCEP tends to be vastly more reliable than IESDP.

<{POST_SNAPBACK}>

Yeah, the animation height info just now made its way into ielister/dltcep

[FredSRichardson]

DLTCEP tends to be vastly more reliable than IESDP.

<{POST_SNAPBACK}>

Yeah, the animation height info just now made its way into ielister/dltcep

<{POST_SNAPBACK}>

 

Do either of you guys know of a good comman-line text dumping version of this utilities? I've got something that work pretty well, but I can't easily update the source code. Basically what it does is dump text versions of ARE/CRE/ITM files.

[FredSRichardson]

Okay, I just found the source for item lister

 

Wow, gemrb! I'll have to try this out

[Avenger]

ielister is the text dumper version works either as total commander or midnight commander plugin.

[FredSRichardson]

Thanks Avenger, I got the CVS sources and built ielister, but I also figured out how to hack the soures to the NI lister hack I put together. The only advantages that has is that it can deal with resoures in biffs and it fills in things like string refs.

[FredSRichardson]

Okay, in AREA v9.1 I'm looking at the Actor structure offset 0x28 "flags" field which has little documentation. IESDP calls it a "flags" field with these values:

 

bit 0: CRE attached (0=yes,1=no)

bit 1: Unknown

bit 2: Unknown

bit 3: Actor Name as Death Variable

 

NI calls it an "Is visible?" field, and DLTCEP doesn't pretend to know what it is (just calls it "Fields").

 

I noticed a few things in IWD1&2:

 

bit0 is always set.

bit1 and/or bit2 are often set for certain NPC's

In IWD2, bit3 is set for inanimate objects like Kegs and Beams.

[Avenger]

dltcep mentions those two bits in its tooltip, and iesdp's info came from me

 

The reason you never see attached creatures in iwd2 areas is because you never check a saved area! If you uncompress a .sav file, you will see that bit0 isn't set in any of the creatures. And dltcep will show those creatures as already 'embedded'. Btw, dltcep can attach the creature by pressing the 'embed' button. This lets you create an area without additional .cre files

 

The actor name as death variable bit works only in bg2, because in iwd/iwd2 the actor's name (in the area) is always overriding the death variable (of the creature).

 

I haven't managed to find the meaning of the other two bits.

[FredSRichardson]

Oops, I missed the tooltips. So are you saying that setting bit0 means that CRE's will be embedded in the save game and nothing more? This is good to know.

 

For bit1/bit2, it looks like they are set for certain "important" NPC's (ones you interact with for something). I believe in IWD1 I only saw bits 0, 1 and 2 set (bit0 always on). IWD2 is much the same with the exception that keg and beam CRE's get bit3 set.

 

Thanks for the info on death variables. I was wondering about that. (It makes the IWD1->2 conversion much easier )

[Avenger]

bit0 set means the actor must be embedded, in other words it isn't yet embedded. (Well that's probably the same as 'will be embedded').

Though, in the save game ALL actors are embedded

 

I call 0x28 'fields' because of the two bits i found, they override fields in the actor (or in other words enable fields of the area).

 

bit3 could mean something completely different in iwd2 than in bg2.