<!--quoteo(post=1853368:date=Jun 16 2011, 03:06 AM:name=kabab)--><div class='quotetop'>QUOTE (kabab @ Jun 16 2011, 03:06 AM) <a href="index.php?act=findpost&pid=1853368"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Cory a very easy fix to this problem in the meantime is doing a tri-planner projection instead of planner for the decals.. It shouldn't be to hard to do.. It will at least get rid of the yucky stretching...<!--QuoteEnd--></div><!--QuoteEEnd--> I don't think it's quite as simple as that. I've talked to Max about fixing the stretching to improve the temp DI effect, but it sounds like it is actually more problematic and time consuming then it is worth.
The tri-planar projection is technically already in game, just only applied for gorges. The only problem is that the stretching will happen no matter what, as the current texture application method for the infestation uses a circle, which can go through the floor/walls, which makes that stretching look. A solution may be to change the infestation to a square texture and to lock it so it doesn't go through tangent planes, but having square infestation would look a lot worse to see than the circular kind. Detecting and applying exclusions according to an alternate plane though would probably be quite complicated though.
Actually, this reminds me of how Crytek does textures for almost vertical cliffs, even though they use heightmaps. They make a 3D texture, which contains three separate texture faces that align along the three axes (x,y,z), then with the use of wizardly magicks applies these textures according to a certain angular threshold. I actually wouldn't be surprised if the actual dynamic infestation uses some kind of voxel displacement system with 3D textures applied. Then again, I also heard Cryengine2 had like 70,000 lines of shader code, so I don't know how complicated 3D textures would be.
PlasmaJoin Date: 2003-04-26Member: 15855Members, Constellation, Squad Five Blue
I think their (limited) time is better spent on doing the real implementation when the time comes, rather than spending hours making the current test implementation look nicer just for a few weeks.
Someone in the NS2 mods forum though has posted an improved texture, maybe that's "good enough" for now?
<!--quoteo(post=1853058:date=Jun 15 2011, 02:52 PM:name=Jason ######)--><div class='quotetop'>QUOTE (Jason ###### @ Jun 15 2011, 02:52 PM) <a href="index.php?act=findpost&pid=1853058"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Optimization then EXO,onos,jetpack and stuff (cause they have a real function and you can test gameplay with them AND look pretty) then "real" Di... (and i dunno.. the current one is kinda memorable <3)<!--QuoteEnd--></div><!--QuoteEEnd-->
Why are the poor Shift and Shade always so ignored? They are the Charlie Brown of NS2 features.
<!--quoteo(post=1853435:date=Jun 16 2011, 01:59 AM:name=Squidget)--><div class='quotetop'>QUOTE (Squidget @ Jun 16 2011, 01:59 AM) <a href="index.php?act=findpost&pid=1853435"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Why are the poor Shift and Shade always so ignored? They are the Charlie Brown of NS2 features.<!--QuoteEnd--></div><!--QuoteEEnd-->
Probably because the aliens are major-league OP right now and don't really need the help.
<!--quoteo(post=1853437:date=Jun 16 2011, 07:02 PM:name=jkflipflop)--><div class='quotetop'>QUOTE (jkflipflop @ Jun 16 2011, 07:02 PM) <a href="index.php?act=findpost&pid=1853437"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Probably because the aliens are major-league OP right now and don't really need the help.<!--QuoteEnd--></div><!--QuoteEEnd-->You should try the new build. :P
<!--quoteo(post=1853359:date=Jun 15 2011, 10:40 PM:name=ScardyBob)--><div class='quotetop'>QUOTE (ScardyBob @ Jun 15 2011, 10:40 PM) <a href="index.php?act=findpost&pid=1853359"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->1. Poor scalability DI doesn't scale very well with map size. In particular, it spreads too fast for rockdown and too slow for summit. Tram seems pretty close to the right speed (or maybe a bit on the fast side). Not sure this is the best method, but having the DI spread speed and radius as a ratio to map size might work (though it'd be a bit confusing for new comms).<!--QuoteEnd--></div><!--QuoteEEnd-->
It's also a pretty static mechanic, with the Hive using Energy regeneration that is affected by nothing currently. Either switch it to P.Res, or give Overlords some other dynamic method to acquire more Energy for a Hive (Infestation absorbing corpses?)
One of the few managed variables that scales along with map size is the number of res nodes and tech points. Tech Points already provide the Aliens with another Hive and therefore another Energy pool; so how about res nodes? Switching Infest to P.Res--or as some have been discussing, a separate Commander resource that's tied to the CC instead of each 'potential Commander', to prevent Commander-rotations--would cause Infestation to scale with the number of Harvester's producing res, which not only indirectly rewards spreading Infestation, but also scales with map size.
As long as I can tell "is there Croach on that wall or not?" when I look at it, I don't care what the beta placeholder looks like. Making it look sexy isn't worth anybody's time if it's a completely different implementation than is intended for the final version.
re: is it croach or infestation? I think people will use Infestation, Croach, and Creep interchangeably, and everybody will know (or figure out quickly over voicecomms) what everybody is talking about. It's not a big deal.
Just a long shot but.... If a cinematic effect played to mask a transition where props and textures changed while the so called "Croaching" is taking place would it have any impact on anything other than overall game file size (performance?). Basically have a decal cinematic cover the area while the props are replaced with the same prop with alterations of damage and deformation. One would think this would be an option to at the very least explore and/or test.
<!--quoteo(post=1855953:date=Jun 24 2011, 01:30 AM:name=SN.Wolf)--><div class='quotetop'>QUOTE (SN.Wolf @ Jun 24 2011, 01:30 AM) <a href="index.php?act=findpost&pid=1855953"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->while the props are replaced with the same prop with alterations of damage and deformation<!--QuoteEnd--></div><!--QuoteEEnd--> Parallax occlusion mapping could do the job
<!--quoteo(post=1855953:date=Jun 24 2011, 01:30 AM:name=SN.Wolf)--><div class='quotetop'>QUOTE (SN.Wolf @ Jun 24 2011, 01:30 AM) <a href="index.php?act=findpost&pid=1855953"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Just a long shot but.... If a cinematic effect played to mask a transition where props and textures changed while the so called "Croaching" is taking place would it have any impact on anything other than overall game file size (performance?). Basically have a decal cinematic cover the area while the props are replaced with the same prop with alterations of damage and deformation. One would think this would be an option to at the very least explore and/or test.<!--QuoteEnd--></div><!--QuoteEEnd--> I figure, if they're doing a dynamic mesh, they could easily edit the meshes of props in real-time rather than pre-sets. Pre-sets are faster to process but limited and result in bigger file sizes.
I can't believe how people argue over Croach, it's a derivation of encroach, example: "Oh my god these aliens are encroaching our worlds" turns into, over a period of time, "Croach spotted on Sol III!" as it (Alien infestation) becomes less generic and more specific.
DI is a technical, meta-term that has no foundation in the NS universe at all.
Infestation is horribly basic, "I have an infestation of bugs..." "Yeah? Well I have an infestation of ALIENS!"
Creep is a good word, but when you realize that SC's alien infestation creeps around worlds at a steady rate and NS' alien infestation only creeps around rooms at the same rate (SC=1 map per 30 or so minutes, NS=1 building in that map in 30 or so minutes) there is a huge difference.
Comments
I don't think it's quite as simple as that. I've talked to Max about fixing the stretching to improve the temp DI effect, but it sounds like it is actually more problematic and time consuming then it is worth.
--Cory
Actually, this reminds me of how Crytek does textures for almost vertical cliffs, even though they use heightmaps. They make a 3D texture, which contains three separate texture faces that align along the three axes (x,y,z), then with the use of wizardly magicks applies these textures according to a certain angular threshold. I actually wouldn't be surprised if the actual dynamic infestation uses some kind of voxel displacement system with 3D textures applied. Then again, I also heard Cryengine2 had like 70,000 lines of shader code, so I don't know how complicated 3D textures would be.
<a href="http://blog.wolfire.com/2009/06/how-to-project-decals/" target="_blank">http://blog.wolfire.com/2009/06/how-to-project-decals/</a>
Or is that a ###### load of work?? None the less I'd imagine a more robust decal system in the future will open up lots more possibilities :D
Someone in the NS2 mods forum though has posted an improved texture, maybe that's "good enough" for now?
then EXO,onos,jetpack and stuff (cause they have a real function and you can test gameplay with them AND look pretty)
then "real" Di... (and i dunno.. the current one is kinda memorable <3)<!--QuoteEnd--></div><!--QuoteEEnd-->
Why are the poor Shift and Shade always so ignored? They are the Charlie Brown of NS2 features.
Probably because the aliens are major-league OP right now and don't really need the help.
DI doesn't scale very well with map size. In particular, it spreads too fast for rockdown and too slow for summit. Tram seems pretty close to the right speed (or maybe a bit on the fast side). Not sure this is the best method, but having the DI spread speed and radius as a ratio to map size might work (though it'd be a bit confusing for new comms).<!--QuoteEnd--></div><!--QuoteEEnd-->
It's also a pretty static mechanic, with the Hive using Energy regeneration that is affected by nothing currently. Either switch it to P.Res, or give Overlords some other dynamic method to acquire more Energy for a Hive (Infestation absorbing corpses?)
One of the few managed variables that scales along with map size is the number of res nodes and tech points. Tech Points already provide the Aliens with another Hive and therefore another Energy pool; so how about res nodes? Switching Infest to P.Res--or as some have been discussing, a separate Commander resource that's tied to the CC instead of each 'potential Commander', to prevent Commander-rotations--would cause Infestation to scale with the number of Harvester's producing res, which not only indirectly rewards spreading Infestation, but also scales with map size.
I
re: is it croach or infestation? I think people will use Infestation, Croach, and Creep interchangeably, and everybody will know (or figure out quickly over voicecomms) what everybody is talking about. It's not a big deal.
Parallax occlusion mapping could do the job
I figure, if they're doing a dynamic mesh, they could easily edit the meshes of props in real-time rather than pre-sets. Pre-sets are faster to process but limited and result in bigger file sizes.
DI is a technical, meta-term that has no foundation in the NS universe at all.
Infestation is horribly basic, "I have an infestation of bugs..." "Yeah? Well I have an infestation of ALIENS!"
Creep is a good word, but when you realize that SC's alien infestation creeps around worlds at a steady rate and NS' alien infestation only creeps around rooms at the same rate (SC=1 map per 30 or so minutes, NS=1 building in that map in 30 or so minutes) there is a huge difference.
Which is what you're doing now, lets stick with Infestation and stop arguing.
Personally I'm looking forward to the final implementation of DI and how it'll work.