Dawn of Light

textures/grass_[realm].csv gives those details, like amount of grass models used for each level and chance.
Important part is that those are grouped, and group number is directly presenting pixel gray level in grassmap
So basically if you going to build new zone, just check different group content from csv and paint with correct levels of gray.
As example how I extract this csv (I'm only interested about name and chance) .. first level array is group number/pixel gray level.

Code: Select all

Array
(
[0] => Array
(
[0] => Array
(
[name] => grass clump 1
[chance] => 30
)

[1] => Array
(
[name] => grass clump 2
[chance] => 30
)

[2] => Array
(
[name] => small plant
[chance] => 25
)

[3] => Array
(
[name] => seed grass 2
[chance] => 12
)

[4] => Array
(
[name] => seed grass 1
[chance] => 3
)

)

[10] => Array
(
[0] => Array
(
[name] => brown small plant
[chance] => 15
)

[1] => Array
(
[name] => brown grass clump 2
[chance] => 10
)
..
..

Also I'm using densemap to clear "roads" from grass layer, but I'm not really sure what this densemap should really do, but at least for this it works ok

About those non-terrain, I have studied a bit, but some how coordinates didn't work same way as normal zones.

Oh and pushed few screenshots from Midgard to blog.
http://mharjula.blogspot.com/

Thank you Silakka, I PM'ed you the info about DAOC dungeons, let me know if you need any other info.

Ok, I'll check this, though I'll probably try to make something for nifskope to get models more easily to Crysis

Do you mean that if you are parsing a dungeon with three instances of BDungHall1.NIF with different WORLD matrices that you would just create three copies of that NIF File and then use NifSkope to change the transform matrix on Scene Root from identity matrix to the correct WORLD matrix specified by using Dungeon.Place file? That is fairly wasteful of resources, but it should work just so you can see the dungeons. Not to mention painful and time consuming on your part.

edit: excellent tool. My last question was pointless, i had time to test it.

Do you mean that if you are parsing a dungeon with three instances of BDungHall1.NIF with different WORLD matrices that you would just create three copies of that NIF File and then use NifSkope to change the transform matrix on Scene Root from identity matrix to the correct WORLD matrix specified by using Dungeon.Place file? That is fairly wasteful of resources, but it should work just so you can see the dungeons. Not to mention painful and time consuming on your part.

I don't have currently working tool chain to get models to Crysis, so I try to get NifSkope to output collada xml format, so it's easier import directly to Crysis.(without all extra steps in other tools)

Interesting, EA/Mythic seems to have few links to my blog from their beta forum

Anyway, I already played with NifSkope code and I have xml base structure already done and some material data already correctly in place .. but sadly this going to take awhile as Collada format takes some time to build correctly.
Also I need to figure how to build model "scene" structures as it's much simpler one in obj format.

Uploaded DAoC vs Crysis2 comparison image to blog
http://mharjula.blogspot.com/

These comparisons really do highlight the difference

I have fantasies about somehow marrying the soon to be released Doom 3 engine source with the DOL server.

Some updates, now moved to free CryEngine SDK, also built java tool to convert things more easily.
I got few sshots in http://mharjula.blogspot.com/ ... though meant more as eye candy than real world scenery

:Silakka
http://mharjula.blogspot.com/

Nice work

awesome

Hello Sillaka,

Really good looking... I guess updating the graphic engine / rework on the interface would move DAoC in the run again (given the game mechanisms are somehow up-to-date)

How is the crysis SDK btw, is it quite open with access to the lower level stuff? does it allow you to use it in conjunction with c++ or c# code?