IWTE - World editing - (General Discussion)

[makanyane]

I am guessing the TB blocks also affect texture quality distance as well eh?

I think so, if you do textures with extreme normals then you can see it switching in and out at the complex edges. i.e. the ones within zone have 'shadows' cast from the normal, and the ones further out don't, so if you have a really bumpy deep normal for the texture the change is quite abrupt and noticeably horrible....

Edit... I can't figure out what highest, upper middle, lower middle, and lowest mean in relation to the TB blocks. The default seems to be lower middle so I am using that for the ruined buildings and lowest for the debris nearby.

ah, should have explained that but didn't expect you to go and reset everything...

the TB blocks are 3 dimensional, so the nominal locations for the complexes are actually in a 4 x 4 x 4 cubic structure, you can view the plan view in the 2d screen, the height dimension is set by the 'upper middle' etc. as the tool doesn't show 3d for that. for interest here's Argantonio's explanation about 'octree's which is how this is set up http://www.twcenter.net/forums/showt...ee#post4273886

for your purposes if you've already set most things to be in 'lower middle' then use the 'resize Table 1 blocks' button to set the middle 'y' value for the height of the centre of the whole thing to something just above your maximum structure height in ms3d, then that should be fine to include all of those.


(side note; the complexes don't have to fit exactly within the cubes shown, the cubic system is actually about the relative position of the centre of the cube representing the relative location of the centre of the bounding box for the complex, so long as you don't have huge overlaps or get things completely the wrong way around then it should be fine)

[Red_Devil]

Hi! back again and working on those default buildings again, not the most interresting to do with the IWTE, most of it with milkshape...
but I'm having a problem. when trying to add a new structure in a complex, there is an reeor message popping up telling me

Traceback (most recent call last):
  File "<string>", line 11780, in ClickedbuttonBoxaddstruct
  File "<string>", line 11876, in addstruct
  File "IWTE_v6_4_1\build\pyi.win32\IWTE_v6_4_1\outPYZ1.pyz/WorldwdObject", line
 921, in addstructure
  File "IWTE_v6_4_1\build\pyi.win32\IWTE_v6_4_1\outPYZ1.pyz/WDWorldLibrary", lin
e 19285, in addstructure
  File "IWTE_v6_4_1\build\pyi.win32\IWTE_v6_4_1\outPYZ1.pyz/WDWorldLibrary", lin
e 6477, in addstructure
  File "IWTE_v6_4_1\build\pyi.win32\IWTE_v6_4_1\outPYZ1.pyz/WDWorldLibrary", lin
e 4471, in addstructure
  File "IWTE_v6_4_1\build\pyi.win32\IWTE_v6_4_1\outPYZ1.pyz/WDWorldLibrary", lin
e 5419, in readms3dfile
ValueError: could not convert string to float: XXXXXXXXXXXXXXXX

I've managed to add other structures without any issue, but it only does that for a few of them, not adding them in the world and corrupting the files.
does it have something to do with the bounding sphere, the dimension of the structure, the TB? I tried a few things but no results...

[makanyane]

if it's happening just with certain ms3d files, then it's probably something in the ms3d file
(huge numbers in the group 'names'? or group/model comments that shouldn't be there?)

or can you add those models in some complexes but not others?

If it looks like the ms3d could you send the file?

Otherwise I either need the files or more details about what's different between the times it works and when it doesn't...

[Red_Devil]

seems to have worked...
I just gave it a generic name before: i let the regroup01 or duplicate01 eventhough i took the habit of regrouping every blocks together...
now i did as you said (naming it Eo0_g1_c1s0 for example) and it worked. thanks!

[wilddog]

Finally made some progress on the deployment process... just some final tests to try and recheck I didn't break anything in the process.

[wilddog]

Version 12_09_A released.
Please report any bugs.
I reset some parameters particularly around deployment blocks.

Enjoy

[Red_Devil]

wow, tried it out and i'm just amazed by the speed... a few stuff though i'm not used too. such as the view option, when viewing in the 3D window, the models showed previously will still be showed if the view is not cleared in the 2D window. is this intentional? in any case, really nice job. tried it a few times and haven't found any bugs.

Nothing related to the new version: i'm having issues with the pathfinding on a settlement... I modified the tga with the right blue and green colours and saved it back to a .worldpath but in game it acts as if it was entirely green. i tried amending in the 2D window directly, it shows the same apthfinding as in TGA and still won't work. any advice?

edit: i don't know how exactly, i managed to fix it... i'm still open to any suggestions in case it happens again

[makanyane]

Nothing related to the new version: i'm having issues with the pathfinding on a settlement... I modified the tga with the right blue and green colours and saved it back to a .worldpath but in game it acts as if it was entirely green. i tried amending in the 2D window directly, it shows the same apthfinding as in TGA and still won't work. any advice?

edit: i don't know how exactly, i managed to fix it... i'm still open to any suggestions in case it happens again

If you'd only used the tga method I'd assume you had a slightly wrong blue (I'm not sure which way 'off' colours default) but if it looks right in the 2d window then that can't be it.

Either something weird we haven't come across yet - or maybe you temporarily weren't loading the correct file
(path name / location issues)

[Red_Devil]

still totally impressed by the speed. the problem hasn't occured agin, so i'll let it drop.

I was wondering. would it be possible for you to add a "test settlement" option? I don't know if with your script, you can call the M2TW exe, and launch immediately a singleplayer battle with the modded settlement (select the settlementName folder with a few options:
player is defender/attacker
player has x units AI has x units
and launch the battle with lower tier units (peasants)?

don't have enough knowledge in java (i think that's what you're coding in right?) to know if it's possible.
it would make the tests a lot faster ^^
or perhaps it's not worth the time...

[makanyane]

still totally impressed by the speed. the problem hasn't occured agin, so i'll let it drop.

I was wondering. would it be possible for you to add a "test settlement" option? I don't know if with your script, you can call the M2TW exe, and launch immediately a singleplayer battle with the modded settlement (select the settlementName folder with a few options:
player is defender/attacker
player has x units AI has x units
and launch the battle with lower tier units (peasants)?

don't have enough knowledge in java (i think that's what you're coding in right?) to know if it's possible.
it would make the tests a lot faster ^^
or perhaps it's not worth the time...

Hiya, WD says NO...


longer explanation;
selecting settlementName folder is not actually something the M2.exe is set up to do directly - you'd need the launcher to copy/replace/rename your test files into one of the default settlement types that you can load from custom battle
(i.e. put them in place of the North European City and launch a custom battle with NE City added)

you could if you really wanted to design a .bat file to copy/replace/rename your test files into a selected default settlement file location but as a lot of the file paths are actually inside the binary files (i.e. from .world to .animinstances, or .worldterrain to the masks) that would complicate things (the masks don't work properly from a different folder)

it's really just easiest to do your development with the custom settlement using the NE City names (or whatever level settlement has the right height of wall), so you can go directly in to the standard custom battle menu and do the defend/attack set up etc - you can even make a 'save game' of the custom battle set up so you can reload the battle with same men and equipment. Re-loading the custom battle save game will load the revised version of your .world if you've changed it, so it's a really easy way to test if it's changed troop behaviour. (also if you add the coordinates of the square your settlement will actually be used on to custom_locations.txt, then you can choose that location for the battle and it comes out looking the same as the campaign map battle will - which is nice of M2, RTW used to rotate the map by 90 degrees )

When you're happy it's all finished, then change the file/path names if you don't want it to stay as the default for the culture and sort out your .worldpkgdesc




Wilddog is using Python for programming

[Red_Devil]

haha, thought so. that wouldn't be worth the time...
I do use the save custom battle thing and i must admit it satisfies. but my laptop is a bit slow between the menus but not in battle sometimes, which frustrates me.
didn't know about the .txt change though. that is interresting. thanks!

[Red_Devil]

I've got a problem here.
I'm using the north_european palisade model with doors and stuff, for now only doing the walls. but when defining for example a door's animation, the orientation i choose with the IWTE 90° puts the object siege docking on the opposite side (ram supposed to arrive from left to right, the docking point is on the right side of the door...) so i reversed the orientation (270°)
But when defining another door with a 315° orientation, I finally found out that the right orientation was 45° so tangent to the door. strange. I restarted the whole settlement but the problem still occured. any error from me or is it some kind of bug?
thanks.

edit : loaded the files with the former iwte version and it works just fine.

[wilddog]

@Red_Devil - Thanks for spotting that. It is a bug. I should correct. Its the collision that is being set to the wrong angle and thats doing it.

[makanyane]

Idiot who tested new version (me), tested with a gate at 0 degrees and a gate at 180 degrees... those appear to be the only angles that work! Idiot didn't try 45 etc... ugghh, oops, sorry

[Red_Devil]

ah that explains it ^^

well, glad to be of service. I'll carry on with the basic stuff of the settlement and wait for some patch. I'll be patient.

[wilddog]

@Red_Devil - version 12_10_A is uploaded - it should fix the problem.

[Red_Devil]

thanks a lot, didn't test the settlement yet, but it appears everything works fine when setting the definitions.

edit : worked fine for 90°, 315° and 180° and as soon as i got out of the 45° multiples, it rounded to the nearest one in my case 80.2° -> 90°...
also, xhen adding the second animation (town wooden palisade animation) i have to pick new animation each time... i don't know if i did something wrong or if it's corelated.

[evenmov]

Hi , I notice this
4)Allow assignment of non Breach able walls for AI purposes (ie walls where you scale with ladders but don't create a usable breach).
em ,how can i do this
I have tried to use IWTE_6_4 to do this and it works , I create a wall that can not be damaged and can use ladder ,the AI will attack it
but if AI have cannon it will attack the wall until death ( even the wall can not be damaged )
Can the new IWTE do it better ?

[makanyane]

thanks a lot, didn't test the settlement yet, but it appears everything works fine when setting the definitions.

edit : worked fine for 90°, 315° and 180° and as soon as i got out of the 45° multiples, it rounded to the nearest one in my case 80.2° -> 90°...
also, xhen adding the second animation (town wooden palisade animation) i have to pick new animation each time... i don't know if i did something wrong or if it's corelated.

The in between angles does seem to be messed up again - not so sure it's defaulting to 45, looks more like it's getting confused about clockwise/counter-clockwise adjustments - Wilddog's looking at it.

the picking thing seems fine for me - do you mean you have to pick it from 3 external binaries & it's not letting you select one you already added ???

Hi , I notice this
4)Allow assignment of non Breach able walls for AI purposes (ie walls where you scale with ladders but don't create a usable breach).
em ,how can i do this
I have tried to use IWTE_6_4 to do this and it works , I create a wall that can not be damaged and can use ladder ,the AI will attack it
but if AI have cannon it will attack the wall until death ( even the wall can not be damaged )
Can the new IWTE do it better ?

there's an indicator in the objects table of the .world files that tells the ai something has a breach, there's also an alternative indicator we found (used in one of the breaches in stone_fort_C) that tells the ai "although this thing is a 'wall' the breach is blocked" - Latest version of IWTE allows you to set that indicator when you make or change the breach - there's a drop box where you select either 'normal' or 'blocked' (and you need to edit the pathfinding for the breach so it is blocked - using the pink squares)

so yes that should now work - at least for the ai - human player might still target the wall if they don't realise it's not an access point.

you should actually give a 'wall' object damage state groups, even if the damaged version is just a texture change, then if it is hit it will switch to 'destroyed' version, giving it only one group means it still looks like a target but is impervious

[evenmov]

wow , it sounds great , I think i can use this new functions to do some new wall !
I often use m2tw 's wall because the ai and the damage video
now if I want a wall that can not be destory maybe I could do something like sand wall