why we need to make difs work
by x-ming-x · in Torque 3D Professional · 11/02/2010 (5:46 am) · 11 replies
ok im about 3 weeks in now on importing a simple map and im losing my head.
im no noob to this, ive mapped commercially for over 19 years, ive made many maps on torque and know it well, im experienced and patient but this is starting to make it look impossible for me (or any other bsp style mapper) to continue on torque.
heres my problem. the pipeline for getting maps in game just went from 2 clicks to about 2000
if i use dif my normal maps make textures glow. this is a tiny problem that im sure garage games could fix in a blink.
this is a known issue with difs but people seem to be brushing this BUG off as being unimportant because difs are being phased out.
but there is NO viable alternative being provided to the dif that is not ridiculously time consuming and fraught with bugs and rendering issues.
yes you can stick models together and use the zone and portal tools (which are utterly buggy and not documented). But these tools do not allow for the precision portalling that is possible with difs and it takes ages which is a complete waste of our time, why place and tediously resize the multiple zone AND portal brushes when you can simply place the ONE portal brush in the dif.
the zones and portals are only cubes which means complex zone shapes are impossible, this is a basic of bsp mapping and without it the maps end up with simple layouts. and again even if we could make these complex zones wouldnt it be a waste of time when in dif / bsp this is all calculated for you.
then theres the export to dae which is great for some things BUT why on earth should i have to dissect a complex map export the parts and then reassemble it to simply get the map i started with,yes its pòssible and very clever but its an INSANE WASTE OF TIME
i understand there are performance increases and advantages to using models, and the perfect way to do this for anyone making fps maps with detailed interiors is to make a dif hull for the map conatining all the portalled areas and then place the models inside in t3d. This is how it works with difs it is a simple fast and elegant solution.
and you came up with it in the first place so you know how great it is.
Have you tried to make a proper detailed bsp style map with this new system ?
i know im not alone lots of your community keep posting about difs but i dont think your hearing them. we have learnt your tools and your system and invested in them, we want to work fast in the way we know and many of us are MAPPERS we come from bsp land and not max. we want to make MAPS and difs are your way that enabled us to do it.
i have gone through the process of chopping up the level exporting it and reassembling it ive zoned and portalled it and in the end for all that effort and time, its a crude optimsiation compared to the pixel perfect placement of portals in difs.
so Can someone please get the boss because this is slipping under the rug and people are saying it doesnt matter but i dont think you realise how much it does to many of us. you shouldnt let it slip.
ive done the decent thing ive put days of wasted time in and tried the new system but it is simply incapable of doing what your old system does. for the small effort of keeping the dif going you will be retaining an entire genre of gaming enviroment and indie devs (mappers) for t3d.
please dont phase us out especially AFTER we have bought a full licence for an engine that has the dif feature. i spent the time to learn constructor and setup everything with that, all the texture albums and whatnot and learnt to use it because its YOUR mapping tool
i cannot proceed with my project until i know how i and my team of MAPPERS are going to build the maps.
Currently difs work for me perfectly except for these 2 issues
the glowing normal map texture bug. if that worked i could use difs and get my team into production.
the fear that t3d will not have a "mappers" pipeline for the production of interiors without the dif and that some new version will create new bugs with difs making our assets worthless or requiring weeks / months of reworking
so much work for so many people could be avoided if you address this and support the dif
i see it is logged for QA and im hoping you will fix it.
http://www.torquepowered.com/community/forums/viewthread/119737
if you could confirm this will be fixed i can at least get to work and add the normal maps later (and stop panicing)
PLEASE PLEASE PLEASE talk about this with the GG team and help us mappers out.
im no noob to this, ive mapped commercially for over 19 years, ive made many maps on torque and know it well, im experienced and patient but this is starting to make it look impossible for me (or any other bsp style mapper) to continue on torque.
heres my problem. the pipeline for getting maps in game just went from 2 clicks to about 2000
if i use dif my normal maps make textures glow. this is a tiny problem that im sure garage games could fix in a blink.
this is a known issue with difs but people seem to be brushing this BUG off as being unimportant because difs are being phased out.
but there is NO viable alternative being provided to the dif that is not ridiculously time consuming and fraught with bugs and rendering issues.
yes you can stick models together and use the zone and portal tools (which are utterly buggy and not documented). But these tools do not allow for the precision portalling that is possible with difs and it takes ages which is a complete waste of our time, why place and tediously resize the multiple zone AND portal brushes when you can simply place the ONE portal brush in the dif.
the zones and portals are only cubes which means complex zone shapes are impossible, this is a basic of bsp mapping and without it the maps end up with simple layouts. and again even if we could make these complex zones wouldnt it be a waste of time when in dif / bsp this is all calculated for you.
then theres the export to dae which is great for some things BUT why on earth should i have to dissect a complex map export the parts and then reassemble it to simply get the map i started with,yes its pòssible and very clever but its an INSANE WASTE OF TIME
i understand there are performance increases and advantages to using models, and the perfect way to do this for anyone making fps maps with detailed interiors is to make a dif hull for the map conatining all the portalled areas and then place the models inside in t3d. This is how it works with difs it is a simple fast and elegant solution.
and you came up with it in the first place so you know how great it is.
Have you tried to make a proper detailed bsp style map with this new system ?
i know im not alone lots of your community keep posting about difs but i dont think your hearing them. we have learnt your tools and your system and invested in them, we want to work fast in the way we know and many of us are MAPPERS we come from bsp land and not max. we want to make MAPS and difs are your way that enabled us to do it.
i have gone through the process of chopping up the level exporting it and reassembling it ive zoned and portalled it and in the end for all that effort and time, its a crude optimsiation compared to the pixel perfect placement of portals in difs.
so Can someone please get the boss because this is slipping under the rug and people are saying it doesnt matter but i dont think you realise how much it does to many of us. you shouldnt let it slip.
ive done the decent thing ive put days of wasted time in and tried the new system but it is simply incapable of doing what your old system does. for the small effort of keeping the dif going you will be retaining an entire genre of gaming enviroment and indie devs (mappers) for t3d.
please dont phase us out especially AFTER we have bought a full licence for an engine that has the dif feature. i spent the time to learn constructor and setup everything with that, all the texture albums and whatnot and learnt to use it because its YOUR mapping tool
i cannot proceed with my project until i know how i and my team of MAPPERS are going to build the maps.
Currently difs work for me perfectly except for these 2 issues
the glowing normal map texture bug. if that worked i could use difs and get my team into production.
the fear that t3d will not have a "mappers" pipeline for the production of interiors without the dif and that some new version will create new bugs with difs making our assets worthless or requiring weeks / months of reworking
so much work for so many people could be avoided if you address this and support the dif
i see it is logged for QA and im hoping you will fix it.
http://www.torquepowered.com/community/forums/viewthread/119737
if you could confirm this will be fixed i can at least get to work and add the normal maps later (and stop panicing)
PLEASE PLEASE PLEASE talk about this with the GG team and help us mappers out.
About the author
ive been making games since they used to be on floppy discs, ive been all over the industry and worked for / with all types of companies, big and small. Now im lucky enough to be indie and working full time for myself and my team :)
#2
yes thats what i read and its a crazy path to follow! as theres no viable alternative. or if the viable alternative means hours of pointless work to acheive the same result as a dif can in a few clicks.
i have no problem getting the shapes in to the map as any file format.
i get great performance lighting etc using a dif as the hull of the map and then decorating it with dts props.
its not that i cant do it, as ive posted above ive already done it and made the map with versions both using dif and the t3d zones
The issue is portalling and the implementation of zones and portals via the new system.
if they drop difs and leave us with something that needlessly takes ages to implement and can never be as good as the system they already have working in difs, then well thats just silly.
so when i read post after post of people being shrugged off with "difs are not going to be supported" im just thinking you guys are not LISTENING and maybe its our fault for not expressing it well. so im trying my hardest to get some sense into this situation.
difs are great gg absolutely nailed it when you made them, what a crazy idea to drop them and lumber us with this crude, buggy, time wasting system (when used for fps style maps maybe this system is great for space sims or something i dunno, im just trying to make stock standard bsp style maps).
11/02/2010 (3:55 pm)
Hi Julian thanks for responding, yes thats what i read and its a crazy path to follow! as theres no viable alternative. or if the viable alternative means hours of pointless work to acheive the same result as a dif can in a few clicks.
i have no problem getting the shapes in to the map as any file format.
i get great performance lighting etc using a dif as the hull of the map and then decorating it with dts props.
its not that i cant do it, as ive posted above ive already done it and made the map with versions both using dif and the t3d zones
The issue is portalling and the implementation of zones and portals via the new system.
if they drop difs and leave us with something that needlessly takes ages to implement and can never be as good as the system they already have working in difs, then well thats just silly.
so when i read post after post of people being shrugged off with "difs are not going to be supported" im just thinking you guys are not LISTENING and maybe its our fault for not expressing it well. so im trying my hardest to get some sense into this situation.
difs are great gg absolutely nailed it when you made them, what a crazy idea to drop them and lumber us with this crude, buggy, time wasting system (when used for fps style maps maybe this system is great for space sims or something i dunno, im just trying to make stock standard bsp style maps).
#3
11/02/2010 (4:36 pm)
Worth getting a definate answer on this, can anyone at TP clarify for sure that dif support is being dropped in 1.1 final, and is it going to be worked on as per Mahendra's post? I'm sure other people would be interested in this. Might be also worth posting on eric's blog to see if he can give you a direct answer. 
#4
11/02/2010 (8:25 pm)
I'll talk to Matt later and find out what our current plan is regarding DIF. I do know that the plan was to phase it out and the only reason it was still around right now was for backward compatibility purposes. When the axe was going to fall on it I'm not too sure though.
#5
11/02/2010 (8:27 pm)
thanks for the reply Scott. Would be good to know sooner rather than later, gives me time to crank out those dif to dts conversions. Dif to collada didn't work for me in the editors, or was it dif to dts.. can't remember :)
#6
i understand you may all be looking to the future and new ways of doing things but we cannot afford to wait / play catchup
it makes really good sense to use proven mature tools and systems especially when your an indie
you have that and its worth continuing at least until your new system is proven and matured
Difs are there right now 99.99 % working. so please for us give it a go. the other tools simply dont do the same job, and create work we dont need to do with difs.
again thankyou for replying, im sorry to be a pain but this is a make or break issue for my project and team,
11/02/2010 (9:07 pm)
thanks for responding Scott i understand you may all be looking to the future and new ways of doing things but we cannot afford to wait / play catchup
it makes really good sense to use proven mature tools and systems especially when your an indie
you have that and its worth continuing at least until your new system is proven and matured
Difs are there right now 99.99 % working. so please for us give it a go. the other tools simply dont do the same job, and create work we dont need to do with difs.
again thankyou for replying, im sorry to be a pain but this is a make or break issue for my project and team,
#7
11/02/2010 (9:09 pm)
@ julian dif to collada doesnt work in beta 3 the textures go off, but if you convert them in the older non beta version they come out ok and can be copied across :) 
#8
while it is annoying to have to relearn a new mapping style, there is alot to be said nowadays for modular mapping which has gained massive popularity in the industry, you will have to adapt and it will be slow at first but with properly set up modular assets building maps can be as fast if not faster than the old BSP mapping, plus you end up with a cleaner, more complex environments at the end of the day with none of the rules and limitations of BSP mapping. (i.e alot more detail packed into the environment) of course your performance takes a hit, but with the current gen hardware you have alot more to play with as well.
11/03/2010 (11:05 pm)
Unfortunatly BSP mapping in general across the entire industry is rapidly disapearing....which i assume is why Torque is dropping an axe on it as well, its the industries trend, while it is annoying to have to relearn a new mapping style, there is alot to be said nowadays for modular mapping which has gained massive popularity in the industry, you will have to adapt and it will be slow at first but with properly set up modular assets building maps can be as fast if not faster than the old BSP mapping, plus you end up with a cleaner, more complex environments at the end of the day with none of the rules and limitations of BSP mapping. (i.e alot more detail packed into the environment) of course your performance takes a hit, but with the current gen hardware you have alot more to play with as well.
#9
because i mean yeah obviously i wouldnt want to make some rubbish low detail old fashioned bsp based game like L4D2 or COD or Bioshock.
I dont want to spend my development time waiting for and learning a new a system which is GUARANTEED TO TAKE MORE TIME AND PRODUCE WORSE RESULTS IN THE CASE OF THESE TYPE OF ENVIRONMENTS. and currently does not work (bugs are documented) when there is a perfect working solution in the engine. i dont need or want modular, if i do i use a model, i dont want my game to look like something made on fpscreator and thats what modular zoning ends up like.
As an indie i cannot afford to sit around playing industry fashion guru and waiting for some future system, while the answer to all my needs lies right there in the engine and my project could be well underway using what works and is proven.
what you say may be true Ken but i doubt it, i remember people saying bsp was dead in the days of Half life 1 and voxels were taking over then this, then that. Its called PR by companies who want to push something else. but in the history of pc gaming bsp has dominated and keeps coming back again and again. why, because its like the great white shark it hasnt needed to evolve,
you statement "plus you end up with a cleaner, more complex environments at the end of the day with none of the rules and limitations of BSP mapping. (i.e alot more detail packed into the environment)" is not technically correct if you know how to correctly use the format. in fact the opposite is true which is why the biggest engines for fps gameplay all use it. having the same bsp feature as the Source / ID / Unreal engines is GOOD. think of all the mappers and modders who would appreciate that over "yeh tuff cheese dude you gotta learn the new system"
the current system will add days to the development of each map by making us have to manually place both the zones and the portals,im not alone in thinking this a pain,
http://www.torquepowered.com/community/forums/viewthread/104279
and Tom Spillman agrees in concept in this post "Still DIF as a format may be depreciated, but i think the CSG modeling method and concepts it brings like easy zoning will not be."
http://www.torquepowered.com/community/forums/viewthread/90297
im not moaning at you Ken or the hard working guys at GG honest, and i dont mean to sound rude and go ALL CAPS on ya, but im just trying to cut through the procrastination about "industry and future stuff" and focus on the simple point of how important it is for us to be able to move on into production and how thats being blocked by this issue.
and im not asking that GG make some great new solution that would surely be more time and hassle than fixing the one tiny bug thats stopping difs from being useable.
and im not even saying fix the bug now and release a patch or i cant continue. it would be fine to simply wait for your next version and then add normal maps when the fix is in. Just need someone to make a small decision and tell us so we can move ahead.
i was thinking about this and i would like to make an offer. because i feel many of your users are not aware of or are having difficulty implementing difs. i could make some tutorials and some example maps/free resources to help ensure this feature can be enjoyed and used by everyone as much as i do.
There is nothing better for me than loading up a huge dif turning on wireframe and watching that sweet perfect cullling as you walk around. and you know that means you have a good foundation for your map/interior from the start. the moment it loads all of your portal and zoning needs are done to perfection , no need for another click.
11/04/2010 (8:09 pm)
EXACTLY UNFORTUNATELY !!! so why not keep it. at least till you have a WORKING replacement. because i mean yeah obviously i wouldnt want to make some rubbish low detail old fashioned bsp based game like L4D2 or COD or Bioshock.
I dont want to spend my development time waiting for and learning a new a system which is GUARANTEED TO TAKE MORE TIME AND PRODUCE WORSE RESULTS IN THE CASE OF THESE TYPE OF ENVIRONMENTS. and currently does not work (bugs are documented) when there is a perfect working solution in the engine. i dont need or want modular, if i do i use a model, i dont want my game to look like something made on fpscreator and thats what modular zoning ends up like.
As an indie i cannot afford to sit around playing industry fashion guru and waiting for some future system, while the answer to all my needs lies right there in the engine and my project could be well underway using what works and is proven.
what you say may be true Ken but i doubt it, i remember people saying bsp was dead in the days of Half life 1 and voxels were taking over then this, then that. Its called PR by companies who want to push something else. but in the history of pc gaming bsp has dominated and keeps coming back again and again. why, because its like the great white shark it hasnt needed to evolve,
you statement "plus you end up with a cleaner, more complex environments at the end of the day with none of the rules and limitations of BSP mapping. (i.e alot more detail packed into the environment)" is not technically correct if you know how to correctly use the format. in fact the opposite is true which is why the biggest engines for fps gameplay all use it. having the same bsp feature as the Source / ID / Unreal engines is GOOD. think of all the mappers and modders who would appreciate that over "yeh tuff cheese dude you gotta learn the new system"
the current system will add days to the development of each map by making us have to manually place both the zones and the portals,im not alone in thinking this a pain,
http://www.torquepowered.com/community/forums/viewthread/104279
and Tom Spillman agrees in concept in this post "Still DIF as a format may be depreciated, but i think the CSG modeling method and concepts it brings like easy zoning will not be."
http://www.torquepowered.com/community/forums/viewthread/90297
im not moaning at you Ken or the hard working guys at GG honest, and i dont mean to sound rude and go ALL CAPS on ya, but im just trying to cut through the procrastination about "industry and future stuff" and focus on the simple point of how important it is for us to be able to move on into production and how thats being blocked by this issue.
and im not asking that GG make some great new solution that would surely be more time and hassle than fixing the one tiny bug thats stopping difs from being useable.
and im not even saying fix the bug now and release a patch or i cant continue. it would be fine to simply wait for your next version and then add normal maps when the fix is in. Just need someone to make a small decision and tell us so we can move ahead.
i was thinking about this and i would like to make an offer. because i feel many of your users are not aware of or are having difficulty implementing difs. i could make some tutorials and some example maps/free resources to help ensure this feature can be enjoyed and used by everyone as much as i do.
There is nothing better for me than loading up a huge dif turning on wireframe and watching that sweet perfect cullling as you walk around. and you know that means you have a good foundation for your map/interior from the start. the moment it loads all of your portal and zoning needs are done to perfection , no need for another click.
#10
11/04/2010 (8:42 pm)
I got a chance to talk to Matt and the word is that DIF will still be around for 1.1 Final. Beyond 1.1 though we will no longer support it.
#11
many thanks :)
ps if anyone needs help with Difs and portaling feel free to contact me
if people request ill make a resource.
11/04/2010 (9:08 pm)
ok great news i guess, and a humble thankyou to Scott and Matt, I hope things work out for your new system and ill be eager to try it. but its great to know i can now base my game on 1.1 and move into production.many thanks :)
ps if anyone needs help with Difs and portaling feel free to contact me
if people request ill make a resource.
Torque Owner Jules
Something2Play