[Nienna garcia]

Imho there's no need to go back. Of course it's possible that someone will find OpenMW too buggy and wants to continue their character in vanilla Morrowind till problems get resolved or whatever, but some means to export character and quest states would probably be well enough. [/my2cents]

[Motionsharp]

If you weren't the one tasked with actually implementing what I am about to ask - how would you allow esps to be backwards compatible between OpenMW and vanilla while still allowing for increased functionality? Create a hyprid esp that contains information that vanilla wouldn't use? Or design a secondary "espx" file that would contain all the non-vanilla content? This question also extends to saved game files. This should be taken into consideration if you'd ever like to see things such as different mesh and image formats or texture/bump/parallax mapping, or any other newfangled technique or script that OpenMW as an opensource project could inject into this ancient game structure.

Supplying stripped vanilla engine ESPs and enchanced OpenMW editions is probably the best idea.

CS-grade editor from scratch. Having no idea what would be involved in that, though, I really have no idea.

Seeing as the render window will have already been built (the most complicated part) all you'd have to do is implement loading the records into the list boxes, which should be pretty simple (simple repetitive load into a struct/class for that type). The only thing left would be placing objects and editing dialogue.

[Elle H]

To answer Mr. th3undon3on3's question: That's an awesome formula... that you will be free to implement in a mod as soon as is possible. But no mods, or at least no mods that change/add content, will be included in the official release. The only thing being considered is the excellent unofficial morrowind patch.

Alrighty. How friendly will OpenMW be to mod, because I'm pretty sure stuff like the enchanting formula are hard-coded so can't be changed with a simple mod in the CS. I'm happy to do it myself if I can though, because it'll likely take a lot of tweaking to get just right. Not to mention that "just right" is entirely subjective anyway.

I didn't even know the old formula, so thanks for that. I'm open in principle for using a using a new formula if the old one really svcks, but until I've tested it more I'll have to agree with tomb' and say that this would work best as a mod. Also, my immediate math intuition says that a chance/probability formula should be mostly multiplications and not use addition or subtraction at all, but that's just me.

That formula probably isn't close to the best I could come up with, but the additions and subtractions gave me a quick and easy way to fiddle with the weightings for each stat. Like I said, it's got a lot of tweaking to go. The old formula really did svck though; forcing the player to down scores of fortify intelligence potions to get any chance of success at all for strong enchantments.

That's weird. Unless you're running the exe from another working directory, that shouldn't be possible. Could you try setting the working directory manually just to be sure? (by making a link to the exe and setting properties, I think)

It was all out the same directory, but I'll give that a shot.

EDIT: No joy. I'll just call shenanigans and try again when the next version comes out.

[jessica robson]

Still, I think it could maybe be easier than creating a CS-grade editor from scratch. Having no idea what would be involved in that, though, I really have no idea.

That was the original aim of http://mwedit.sf.net

[Stephy Beck]

I really don't like the idea of MGE being merged into the engine. I (And many others) don't have a graphics card capable of running it. If I did, I'd be playing Oblivion.

[Jade Muggeridge]

Well, I assume it will be configurable..so why not just disable the features you cannot run?

[Sweets Sweets]

If you weren't the one tasked with actually implementing what I am about to ask - how would you allow esps to be backwards compatible between OpenMW and vanilla while still allowing for increased functionality? Create a hyprid esp that contains information that vanilla wouldn't use? Or design a secondary "espx" file that would contain all the non-vanilla content? This question also extends to saved game files. This should be taken into consideration if you'd ever like to see things such as different mesh and image formats or texture/bump/parallax mapping, or any other newfangled technique or script that OpenMW as an opensource project could inject into this ancient game structure.

I think Morrowind would simply ignore records that it doesn't recognize.

[gary lee]

I think Morrowind would simply ignore records that it doesn't recognize.

That is very possible, and would be convenient.

[Rachael]

The presence of other, invalid, corrupted, or new records in most cases will cause Morrowind and the CS to stop reading the datafile and ignore the remaining data. Similar to unknown animation notes.

[Samantha hulme]

The presence of other, invalid, corrupted, or new records in most cases will cause Morrowind and the CS to stop reading the datafile and ignore the remaining data. Similar to unknown animation notes.

Hmm, well that's cool.

[Queen Bitch]

That means that all new records must be kept at the end of the file, then.

[Ridhwan Hemsome]

That means that all new records must be kept at the end of the file, then.

So the free cake is out of the question.





[/nonsense]

[Taylah Illies]

bump? hows it going guys? anything new and exciting?

[maddison]

@Yacoby - I asked a while back but never got a response. So how's the external cell stuff coming? Any new screenies?

[Flutterby]

@Yacoby - I asked a while back but never got a response. So how's the external cell stuff coming? Any new screenies?

Better than it was. I am now just trying to get rid of texture seams and improve performance. Geomorphing is implemented, but it doesn't generate the values to morph to correctly yet. Textures fade between levels, which removed a lot of the "jolt" of spliting and unspliting the quad tree. Basically all the changes I have done are not things that translate well into screenshots. I don't think it is video worthy yet though. I would like to be rid of the seams at least.

As mentioned, there are seams along some of the borders, and while this isn't really that difficult to remove at the lod levels where the textures are pregenerated, I am still thinking about how to do it in a good way for the lowest level, where the textures are drawn onto the landscape as defined by an alpha map, but due to the hardware we are aiming it at, it kills performance.
If a cell has 5 textures on it, the GPU has to render 10 passes, which takes expensive to a whole new level. I am going to see if I can make it any better for my current card (two texture units ftw) without shaders, and then write a better one using shaders.

I am probably going to upload a demo version of how far I have got at some point.

[stephanie eastwood]

Well I'll be buggered. It was Yacoby who was behind OpenMW. This is the OGRE3d Morrowind engine, that I heard about on the FreeGameDev/FreeGamer community, right? Funny how sometimes when you return to a community, another one shows up as well. =)

[Krista Belle Davis]

Better than it was. I am now just trying to get rid of texture seams and improve performance. Geomorphing is implemented, but it doesn't generate the values to morph to correctly yet. Textures fade between levels, which removed a lot of the "jolt" of spliting and unspliting the quad tree. Basically all the changes I have done are not things that translate well into screenshots. I don't think it is video worthy yet though. I would like to be rid of the seams at least.

As mentioned, there are seams along some of the borders, and while this isn't really that difficult to remove at the lod levels where the textures are pregenerated, I am still thinking about how to do it in a good way for the lowest level, where the textures are drawn onto the landscape as defined by an alpha map, but due to the hardware we are aiming it at, it kills performance.
If a cell has 5 textures on it, the GPU has to render 10 passes, which takes expensive to a whole new level. I am going to see if I can make it any better for my current card (two texture units ftw) without shaders, and then write a better one using shaders.

I am probably going to upload a demo version of how far I have got at some point.

Out of curiosity what kind of card do you have?

In other news... yay exterior update!

[biiibi]

Well I'll be buggered. It was Yacoby who was behind OpenMW. This is the OGRE3d Morrowind engine, that I heard about on the FreeGameDev/FreeGamer community, right? Funny how sometimes when you return to a community, another one shows up as well. =)

It isn't actually me, although I did write an engine at one point (My first go at developing a engine, so it never got anywhere). http://www.gamesas.com/bgsforums/index.php?showuser=468902 is doing most of it. I am just doing the terrain engine at the moment. Maybe some other things after that.

(The water maybe *shrugs*)

Out of curiosity what kind of card do you have?

I have a nVidia 6150LE. Supports shader model 3, but only two texture units, and is generally crap. It is integrated into the motherboard, and shares ram. I can't upgrade due to the fact that my computer is low profile.
It is also made by Dell, and not worth the effort of transferring the motherboard etc into a new case, unless you really know what you are doing hardware wise. (Which I don't)

I am not that interested in a new computer though, as the only games I play work fine.

[Jhenna lee Lizama]

What texture filters are you using to blur the seams? Is the expensiveness happening before making the textures seamless or after?

[Miragel Ginza]

What texture filters are you using to blur the seams? Is the expensiveness happening before making the textures seamless or after?

The expensiveness is that I am using http://en.wikipedia.org/wiki/Texture_splatting, and so is requiring (at present) 2 passes to render every texture, and I am rendering a cell as a single entity, so culling isn't as good.

I am going to go and ask the Ogre forum if they have any input into the two issues (seams at the edges of quads, and the performance issues)

[CHangohh BOyy]

Hey, when you ask at the Ogre forum would you post a link to the page?

[Kat Stewart]

Hey, when you ask at the Ogre forum would you post a link to the page?

http://www.ogre3d.org/forums/viewtopic.php?f=5&t=49626

[Shae Munro]

BTW, how fast does what you have render terrain minus the textures? Is it to the point that we could theoretically render an exterior cell with the buildings and rocks and stuff without the terrain textures?

[Austin England]

BTW, how fast does what you have render terrain minus the textures? Is it to the point that we could theoretically render an exterior cell with the buildings and rocks and stuff without the terrain textures?

Yeah. I can actually get around 30fps if I ditch the view distance right down so that I don't render multiple high detail cells, and a decent card will be able to render the terrain at a far greater speed that mine can.

[Madison Poo]

From what you said on the ogre forums, isn't the first problem similar to the texture seams that Slartibartfast fixed? Or are those two completely different things?