Project 'Kadath'

[sinbad]

I've been being coy about my 'private project' it's been known that I've been working on recently because a) I don't want to raise false hopes, b) I don't really want to be pestered with questions about it and c) I'd rather someone didn't beat me to it but I'm going to lay some of my cards out on the table here now. This has been triggered by various discussions about level formats and challenges like the one in this thread: http://www.ogre3d.org/phpBB2/viewtopic.php?t=13528 , and I feel keeping too quiet may result in people making long-term decisions which might change had they known what I was up to.

The question of indoor level formats has been bandied about for ages. Q3A BSP was until recently still covered by restrictive licensing, and in any case is too fragmented for efficient use on modern GPUs. Octrees work ok, except where the levels are very large and highly occluded - hardware occlusion can deal with that but it has relatively high hardware requirements. Portal rendering is a nice solution, but the majority of implementations require the level to be specifically constructed in terms of (possibly convex) closed regions and connecting portals, which is somewhat cumbersome for artists who would much rather use their everyday toolset in a more natural way.

I'm working on a solution to this problem right now, with a project I've codenamed 'Kadath' (I have a real name in mind, but we'll see). I had been continually hoping over the last couple of years that someone in the community would step up to the challenge of coming up with a good scene manager which managed indoor scenes but was not constrained by Q3A limitations, but I'm still waiting. With 1.0.0 out the way I finally decided 'if you want something done, you have to do it yourself' - although my workload with the XSI exporter, general maintenance etc haven't let me really start it until very recently. It's still in the early stages - I'm aiming for early 2006 as the first release, and plan to build on it further throughout next year if the initial reception is good. I'm not going to talk about the intricate details of exactly how the project is going to work at this stage, since it's too early - just that it the results will be GPU- and shader-friendly, and won't require specialised level creation tools, and will be intended to play nice with non-indoor scenes later. I'm fairly confident I can pull this off well, but time will tell. This is also why you're not seeing me around on IRC or the forums quite as much as I used to be - OGRE itself is now only a part of my workload.

I also have to say that this product will not be free. 'Real life' changes all the time, and my situation is such that in the next 12 months my OGRE work has to graduate from a part-time hobby to a full-time job, or I have to scale back the amount of time I put into it. Needless to say, I'm trying the former first, but that means I cannot continue to give all my work away for free. I haven't decided on the exact pricing yet (I thought I should concentrate on writing it first ), but bear in mind I specifically want to aim this at the indie market - amateurs, students, and small companies - so you can count on it being priced affordably for each group.

Ok - so there you have it. I hope I can pull this off, and that people will like it - time will tell I guess. Please don't pester me with lots of questions about the detail - I'll reveal that when I'm ready. I just wanted to announce my high-level intentions so as people are informed about what may happen over the next few months.

Sinbad out.

[Kencho]

Good luck!

[Tyn]

That definatly sounds interesting to me, I've seen various people saying of how BSP is too old and inefficient on modern GPUs and yet not really giving much alternative. If this could be that alternative then you could really be onto something hot. I'm a little wary by having to pay for it for a project that's developed for free itself, then again, I always am, worth it or not . I'm sure that in mid 2006 I would be in a position to pay money towards this and if it helps you to keep working on what you want to do, well, a happy dev can only be a good thing for Ogre IMO. I'm sure others will feel the same.

Good luck!

[tuan kuranes]

Good Luck.
It's a Great Idea, we all agree there's a real need here, as seen many times in those forums. Bsp is really getting old and inadequate, whilst still widely used.

[edit]removed lots of question that I wanted to ask... but shouldn't [/edit]

[iq]

Interesting the bit about 'play nice with outdoor scenes'. Does this mean you can have a (optimized rendering of) surrounding viewable through windows or does it optimize rendering of interiors if seen from outside? (ie allows highly detailed interiors in a rpg without any scene switching inside/outside).

Good luck with that project - hopefully it works out nice.

[Fredz]

I also have to say that this product will not be free.

I perfectly understand you point of view, but I was just wondering if you did consider releasing you work under a double license (commercial and GNU GPL), just like what Trolltech does for QT ? This way you could still get money from people using it in commercial games, and in the same time allow free projects (under the GNU GPL) to use it.

[sunetos]

Sounds great; from what I've learned so far it seems your project (maybe alongside qsilver's) could fill in a currently much-needy hole in ogre. If it helps your motivation any, if you had already done this by a week ago, my team would've bought it without blinking and gone through much less debate/deliberation before deciding on ogre (as long as the plugin wasn't going to cost more than our new graphics workstation).

[Chris Jones]

good luck with this project!

i am very interested in this, but the only thing that puts me off is it costing. im a student, and nearly always have no money, what type of range of pricing would it be? ~£20? ~£50? £100? more?

[Timme]

Will there be a "free" version of the engine? Whats about the source code of the engine? Will I be able to download it for "free"? Will the engine cost money or only the tools?

[haffax]

I haven't decided on the exact pricing yet.

[spookyboo]

Sinbad, I hope you can realise your plans. If someone can pull this one off, its you. Good luck.
And for everyone who still wants everything for free; someone who has offered so much for free is entitled to make some money out of it. If the price is right and the product good why not buy it?

Just one question?

and won't require specialised level creation tools

Does this imply some kind of high-level scene definition language?

Good luck

[sinbad]

I also have to say that this product will not be free.

I perfectly understand you point of view, but I was just wondering if you did consider releasing you work under a double license (commercial and GNU GPL), just like what Trolltech does for QT ? This way you could still get money from people using it in commercial games, and in the same time allow free projects (under the GNU GPL) to use it.

I've certainly considered it, and I've not completely ruled it out yet although it's looking unlikely to be worthwhile. Dual licensing works fine for libraries where the viral nature of the GPL provides the incentive to pay money for an alternative license, but it works far less well for tools. Sure there will be a SceneManager to go with this which could be covered, but perversely GPL'ing all of it would encourage people to use the tool under GPL and use the output format with another engine. Would people really pay money for GIMP if you offered it under another license? I don't think so.

I love open source but the sad fact is it's incredibly difficult to make a living from it unless you are in an extremely mainstream subject area like web servers, mail servers etc so that companies will pony up for support / consulting on a regular, sustainable basis. I've tested those waters with my company TKS over the past 18 months and, sadly, the volumes of paying customers just aren't there to make that a viable business model in it's own right.

Will there be a "free" version of the engine? Whats about the source code of the engine? Will I be able to download it for "free"? Will the engine cost money or only the tools?

There will very likely be a limited free trial of some sort. Details when it's ready.

I won't be drawn on possible pricing right now except to say that the standard for this sort of target audience has been set at $100 per seat, that is my benchmark / starting point but I've yet to think detail. All in good time.

Thanks for the encouragement - I certainly hope I can make this worth your whiles.

[Kencho]

You will If I finally go on with a serious gaming (or simulation) project, count with one license for me Quality software is always welcome

[sinbad]

I'll answer these 2 questions but no more

Interesting the bit about 'play nice with outdoor scenes'. Does this mean you can have a (optimized rendering of) surrounding viewable through windows or does it optimize rendering of interiors if seen from outside? (ie allows highly detailed interiors in a rpg without any scene switching inside/outside).

Both. But that's not on the plan for the first release, that's on the plan for the second one, which will only get done if the first one is well received

and won't require specialised level creation tools

Does this imply some kind of high-level scene definition language?

An exporter framework which any modeller can plug into, either directly or via Collada. Some metadata will be required against the data that is passed to the tool - I have yet to evaluate how that will be exposed to the artist in practice, probably through widgets in the modeller or perhaps through generalised properties. My primary targets will be XSI and Blender but the concepts can be applied to anything that can spit out a polygon soup. Who knows, if it goes outstandingly well maybe one day I'd be able to afford to buy 3DS & Maya to make custom plugins for those too

[Olex]

I say it is about time sinbad got some $$$ from all of this, i.e. OGRE, I am surprised there have been so few.

Is this a first your first development of a tool for OGRE that you wish to make $ from, sinbad?

Certainly, wish good luck with full-time job on your creation.

Just noticed that you said, no more questions. Never mind then.

[sinbad]

I meant questions about the detail of the tool, that kind of question is ok

Yes, this will be the first graphics related tool I have ever sold commercially. I have a day job which pays the bills, my graphics coding has always been a hobby and I've always given away the results for free for the fun and learning of it. Come to think of it, I've been giving away graphics code (of varying quality I might add, OGRE is the crowning achievement so far ) for the past 12 years.

[Olex]

Say, Guernsey, Channel Islands is rather an exotic place (politically and economically speaking), I wonder if this means that there aren't many opportunities for graphics development jobs there. Does this mean that you are shooting for an indie-kind of full-time job with Internet being your market?

[ahmedali]

@sinbad
I think you have done a lot lot with ogre for OS community. You deserve to get something back from your investment of time & efforts. But ofcourse 'Kadath' is an independent project just like some other project/plugin and licensing will be also independent of ogre.

But can you tell me about some technical features that will make it a better indoor SM.

[sinbad]

Say, Guernsey, Channel Islands is rather an exotic place (politically and economically speaking), I wonder if this means that there aren't many opportunities for graphics development jobs there. Does this mean that you are shooting for an indie-kind of full-time job with Internet being your market?

Correct, the local employment market is dominated by finance, and although there are a few other things going on, graphics / games production is not one of them. Yet - who knows what I might start After all, Specsavers started here, Willie Johnson lives here, and Victor Hugo wrote Les Miserables while here, so anything might be possible I'm a big believer that the games industry in particular needs more outlets, not just the hit-driven retailers we have now, so my ultimate aim is to be a part of what comes next.

@ahmedali: I've said all I'm willing to say for now - I only raised this so people knew something was going on. I'll reveal more details in due course.

[Chris Jones]

Correct, the local employment market is dominated by finance, and although there are a few other things going on, graphics / games production is not one of them. Yet - who knows what I might start Wink After all, Specsavers started here, Willie Johnson lives here, and Victor Hugo wrote Les Miserables while here, so anything might be possible Smile I'm a big believer that the games industry in particular needs more outlets, not just the hit-driven retailers we have now, so my ultimate aim is to be a part of what comes next.

if u did that, and were hiring people, id apply. i really want 2 get into the games industry and id perfer to not go to somewhere like america, didnt u say theres no mcDonalds where u r? bliss

good luck with your project, i will almost certainly buy it when its released!

[dannomite]

Sinbad, I hope you consider pricing in the range that Torque offers for their engine. The more affordable the better. =)

[alphageek]

This is great news Sinbad. The best part of it is that you are trying to move into making Ogre your fulltime job.

I'm fairly confident I can pull this off well

Me too.

[machinimist]

sorry, i don't quite get it. will kadath "just" be some kind of super scene manager? or a torque competitor based on ogre and the super scene manager?

[sichuan]

OGRE's license will be changed in the future?

[haffax]

No.