paramecij

The goal would be to write first and foremost a tech demo to show off engine features and for performance testing, and later to serve as an example implementation of different engine features (code samples) .. Josh and others made some videos to show off LE2, and they probably will too for LE3, Josh (and others) also made a lot of tutorials showing how to do different things in the engine, they also answered a ton of forums posts on how to do things over and over again, and Metatron made his gamelib... all of this required some assets, code and a lot of time to make/write.. there's some duplicated effort as a result of everyone working on their own. So if everyone willing is planning to do this again for LE3, all i'm saying is, there's a better more optimized way to do this if we work together on the same framework and share the same asset pool.. Let's say we start small (LEBuilder style) and set-up a small application framework and person A is in charge of maintaining it, every other person uses the same framework and adheres to same rules, for example every person uses the same class scene and has to implement sceme::load() scene::start() scene::unload() so person A can easily integrate it to the demo app ... - person B joins in and writes a basic physics example using only primitives - person C joins in and donates some of his already made assets ( soldier, test level ) - person D joins in and writes a character controller example using the assets available ... depending on how things turn out, we could then think about doing a simple game or whatever (again, with already available assets), the idea is to save on time while giving out more to the community.. I have a lot of stuff i'm willing to put to a good use.. You couldn't do all too much in the evaluation kit anyway, with this you'd get right to evaluating the engine instead of trying to figure out the editor and write some scripts.. and for the developers (they get the source) it would be a good starting point or a reference in the future (like some of the tutorials and docs are now)..

You can make islands regardless (you start from the bottom of the sea .. but sure, just flatten the terrain at some middle height, then you can lower and raise it .. you can also fill your heightmap file with some solid gray color in gimp or photoshop

If you are using arbuz 3dmax exporter, you have to be careful with export material id's option.. this can lead to unnecessary surfaces (check the manual) as Josh pointed out

I'm afraid the newton vehicle controller won't be up to the task if you're aiming for a real realistic driving simulator. You can however implement your own vehicle controller using newton or your own or 3rd party physics. You'd have to do this in any other engine anyway (except if there's an engine out there aimed at driving simulators). But you can get decent driving using the provided controller if you fine tune some parameters around and don't care for full realism. Regarding deforming, not out of the box, but you could easily cheat this a little in different ways... I think you wont go wrong if you decide for LE, even if you had cryengine you'd probably want to write your own driving model as theirs is pretty arcade like. It's easy to make a racing game in LE but to make a realistic simulator in LE is as hard as in every other engine - only LE has unmatched ease of use (for me at least) ... I've made a small app play around with different vehicle controllers and fine tune them, if you decide for this engine I can give you the source to help you get started.. (it's in C though)

It would be good to have a official LE3 tech demo, a collection of scenes or maybe a simple game to show off all the engine features and let users evaluate performance at the same time. This could be a community made thing (and let Josh do what he does best), but it must be officially supported. It would also be good if it was open source (to developers), in that way it can serve as a good reference on how to do different things, an invaluable help at learning and even later.. VeTaL already expressed the need for more art and coding guides, this could be something like that! If we set things up right and worked closely together it could drastically help Josh bug solve the engine, at all stages,.. early on we'be beta and stress testing for him, later on users would have a working reference to check if it's not a problem with their code, model, material... there's even more reasons why we should do this, but it's a thing we all can profit from .. There's some talented people here, ..and i'd be interested in a collaboration like this, if there is enough interest (Metatron is already doing something in the way of this with his gamelib, and others too with support tools and assets).. what do you think? ..Josh?

since you got fragmotion already, which reads animated .x files just fine, from there you can export to .smd (half life) format and use an older user made converter found somewhere here on the site, there are also scripts/plugins for major 3d packages (3dmax, blender..), and tools like milkshape, uu3d, fragmotion already have good support for .smd .. I've had less trouble with unreal and halflife formats then with fbx as an exchange format for animated game models, but uu3d is good investment (it's like fragmotion with every 3d format imaginable, and more tools)

I think for starters you should do some passive fauna - like birds, rabbits, rats, cats, etc. .. I think there's a shortage of these, and it would be interesting for a more broader user base than medieval or sci-fi themes. I think it's better to start small and iron out the dev process first then to start with the main hero type projects right away. Plus I think that main hero chars aren't worth the trouble making - everyone has their own vision on how their main protagonist will look and animate like, so I think it's better if you work on 'supporting cast' characters or monsters, and do main chars upon request.. And I second Rick's comments on bones, sure It can work fine if you just parent the helmet on head, gun on right hand, ..but what if you char has no right or hands at all? It's more flexible this way and I find it's easier to ignore/remove than to add a bone if need be.. also since you're thinking modular, swapping things and such, I hope this also includes animation, so they could at least be shared across different characters, or used on previous chars if you make new/different ones.. I know it's impossible to try and do every animation people would need in their games, so I think a way to easily extend the animation set of a character in the future would be golden.. ..just my thoughts, and good luck with whatever you decide to do!

Hm that's a good idea, .. couldn't you use the 'v' from UVs directly to drive the intensity ? you would mix some gray color multiplied by the v coordinate to the texel lookup ? I used something like this for a post-process vignette effect, isn't that basically what you're looking for? I'll take a look at the shaders later and get back to you ..

He could have at least replied ... but anyway i'm not from any high risk country (i'm from slovenia for christ sake, we're small and harmless..), and even if i was, it would still be discrimination. Everyone has the same options, regardless where they are from, they can pirate software or they can buy it, I choose to buy it, again and again, I won't choose to pirate it even if it's the only option I have... I agree it's his decision, but I think it's a poor one to make (not from financial view, but from moral..) .. Thank you for your effort to try and help me, you're a good guy..

Now here's a novel game idea!! .. I'm sure there's a group of people desperately waiting for something like this

Ok i'm sorry and I appologize for my selection of words, didn't want to sound so harsh.. but it's a thing of frustration for me, and people here are really rubbing it to my face I have tried to buy UU3D numerous times (over the span 12+ months) , the payment went through - my funds were withdrawn (he was using paypal, why wouldn't it accept orders?), and I received a refund a couple of days later .. now the buy page isn't even working (maybe he blocked whole IP ranges?).. I emailed Brad many times (10+), got no answer whatsoever... tried to ask somebody in the US to buy it for me, but people are too untrusting and i'm not even sure if this is legal or allowed by Brad..

This is rather simple to do even now ( I just re-create the convex-hulls from mesh surfaces on load time based on scale parameter ), but it doesn't allow for true dynamic scaling of physical objects, at best you can switch to the nearest pre-scaled body in chain, which doesn't play nice with an ongoing physical simulation

..so why do we have the Terminate() function again? .. I think every serious or pro library should have a clean exit method and the ability to initialize it again .. else it hints of bad memory management etc..

You either have to make the models bigger (easy), or modify the shaders to draw them bigger (harder) ... The max height and slope in the editor refers to terrain elevation the vegetation layers starts and stops showing (so you don't have tall trees on top of mountains and below sea level)

you can buy uu3d only from specific countries, if you're lucky enough that's your best option.. but i've found fragmotion to work well animated .x files (static also), .dae doesn't work all that great in many different apps from my experience, fbx is much better now, but for animated models I use .smd format (half-life2) because I found it to best most reliable with different apps for me ..

Why haven't we found a good alternative to tree-view? You can't make a good looking UI with it, and it's too much clicking for my liking - always collapsing and expanding stuff which is getting crammed by too many sub-levels and too long strings... other than that I like the way you're going with this (and LE3 overall)!! P.S.: I think google's picassa has a good UI, not stock windows, but not distracting either, best of all, it simply works and fast at that, it never took me more than 2 clicks or a couple of keystrokes to find the right picture (amongst 15.000+)

While converting some of my animated assets to GMF via the fbx2gmf, I noticed some of the leaf bones get pruned.. one of those pruned bones controls bullet shell ejecting (while the other two didn't get pruned -what gives?(they are also without child bones and have the same parent) ), other missing bones are mainly position placeholders (firespot, gunsmoke, etc..) and aren't as important ... so ...what exactly are leaf bones? why do they get pruned? how can I get around this?

..because of the seams?

Very nice! You made the whole model in max? Or did you also sculpt on it with mudbox/zbrush or similar app? .. I'm not sure how I'd start on a model like this - a brief overview of how you did it would be most welcome @Aily: I agree that it's too dark, but most people DO have badly set-up monitors (maybe Ywa as well)..

Ah, I see.. I was considering buying the source and was about to ask for your opinion on memory mapped files etc. how hard would it be to implement and such.. no matter, I still appreciate the effort and looking forward to what you'll bring us

Yeah, that's the biggest problem I had when I switched to VS2010.. had to re-compile most of the libs, which sometimes proved to be a bigger challenge than I thought (that's why I still have the previous VS installed) .. but I had some show-stopper errors, this one is just for the missing debug info (i think) so I don't think it makes a difference for the release build..

Yeah that's the biggest problem for me right now, getting functional animated models to GMF.. Here are some things you can check from the top of my head: - exporter supports only skin and not physique (and no morphing) - select ONLY the skinned object and not the bones (while using the export selected option) - make sure skin is the topmost modifier - if you're in editable mesh put edit poly modifier on top (and below skin) - you must have a proper material to view the skinned model correctly (in your .mat files you must use a shader that has '_skin' in it) - try to reset your transforms - sometimes you can't see nothing in modelviewer or maybe just the hierarchy but no mesh, because of the big scale, but if you scale it down in game it works fine - when exporting to .fbx use version 2010 binary and turn off curve filters, resample, etc, but leave animation and skins on... - always import the exported .fbx back into max, to check if the files has been properly exported and sometimes you'll be able to figure out what's causing the problem this way Hope some of this helps, but first i'd try exporting with just the mesh selected and checked your .mat files to use skin shaders (judging from your pic)

UI looks nice, and the title and graphic look very interesting! Congrats on nearing a release with your game!

this is so AWESOME! ..I always wondered what you guys are doing with the source..

I can donate some of my crappy trees and foliage if you don't find anything else to use