Jump to content
The Dark Mod Forums

Sotha

Development Role
  • Posts

    5664
  • Joined

  • Days Won

    134

Posts posted by Sotha

  1. Thanks for sharing.

     

    That really was an interesting read and formalizes some of my experiences I've had with New Things in the past. It was fun to even recognize that I was a 'mop' in that New Thing back in the past and a 'geek / fanatic' in that other. Even recognized some 'sociopaths' from the past.

     

    As for the end, well, I think it is natural that all things come to an end eventually and delaying the end is the only thing we can really try to do.

     

    And when the end comes, it is not too bad either if the Ending Thing brough joy while it was still active. That joy and experience will increase in value, because memories have the habit of getting gilded. Perhaps 20 years from now TDM is long gone and we have a bunch of us old farts hanging around here remembering in how exciting circumstances TDM got Greenlit, or how frickin' unfair that teleporting flaming revenant in Transaction was (what was I thinking!) ;):D

    • Like 2
  2. Peer review won't help, because like grayman said, it is too late to change them.

     

    *If* someone makes new modules to TDM core, perhaps then it would be a good idea to do some team peer reviewing.

     

    But my personal experience is that whatever you do, and no matter how well you do it, and no matter how much others helped you to perfect it, someone else *will* find some component of your hard work, which they can (rightly or not) criticize.

     

    And that is of course irritating, and it burns out unpaid voluntary workers ultra-fast.

    • Like 4
  3. Part 7 about lights is up.

     

    @Judith,

     

    The benefit with nightvision approach is that the mission will not look so washed out, which is why I wanted to use this here. It has not been experimented with much yet.

     

    I do agree that the nightvision is probably not the newbie mapper's default choice, which is why I talked about the ambient world lights in the video. Should have probably talked more about it though. It probably isn't a terrible mistake on my part, though, because most likely the newbie mapper will be referred to the A-Z guide anyway, and that covers the ambient world lights.

     

    I am thinking that if I calibrate myself properly in the training mission gamma room and set the nightvision lamp brightness to decent level, it will work for the majority of the people. If it doesn't, the player can always tune their ingame gamma settings to the brightness level they please.

  4. Keeping everything a standard 90 degrees looks like extreme caution to me - and a waste, because you are missing out on a lot of cool shapes.

     

    What Springheel said. You can still do any kind of geometry with the func_statics. Just keep the sealing brushes blocky-90-degrees. You won't be losing any interesting shapes... The only thing you will lose is the bug hunts for weird problems. I am happy if you have not encountered those. Weird sealing problems are easy to fix, but pain in the butt to locate.

  5. It's located in the sound settings in the system panel under windows. If you are using linux there should be something similar either. But it has to be set before recording so this is just a hint if you are doing future videos. I'm not sure whether it is possible to remove the noise from your videos afterwards.

     

    It is not extremely worse, you are good to understand. It is just a bit annoying. :)

     

    Ok. I googled around and found out how to activate the noise suppression, but for some reason my system does not have the tab which has this microphone option.

     

    I don't know how to get that enhancements-tab. Sorry about the inconvenience. I hope it isn't too bad. :(

     

    I understand that the purpose of your videos is to show beginners a quick and easy way to make a map. I probably should have mentioned this beforehand but I would like to learn about "best practices" techniques in making maps. Not necessarily just building quick and small maps but with an eye on the future so I can eventually, properly make large and highly detailed maps.

     

     

    Would this "tris" therefore become more exaggerated in larger and more complex maps and ultimately have too much of a negative impact on a large map? Is there a tipping stone point?

     

     

    Is brushwork the "traditional way" of structure building? Because I've come from the Unreal community I'm having a difficult time understanding some of these new terms.

     

    The modular way is the best way. That is what all the modern games use nowadays. The difference here is that I make the modules in DR, because I am not so good with real 3d-modeling software like Blender.

     

     

    Modules do not necessarely mean a higher tris count. It is just that they are often more detailed then what mappers build by worldspawn. However, as we have LOD support using LOD modules will actually decrease tris count and therefore improve performance. If you are planning on creating big maps with open scenes, you are actually requiered to use LOD modules to gain a reasonable performance. This is actually what all modern games do (like Skyrim, Dishonored, etc...)

     

    The "traditional way" is a term that is used here in reference to the DarkEngine (the engine for the first two thief games). Beside some basic models used for furniture, decoratives etc.. the level geometry itself is build using worldspawn (so brushes and patches). So models are only used for detailing or interactive objects.

     

    The "modern way" of building is to build the level geometry (walls, floors etc...) out of models (or modules, it is basically the same), too. This is mostly due to gain a high level of detail even in large games, which would be much more time consuming if it would be done the traditional way, and to gain better performance. Worldspawn is only used to partition the map (for performance and sound propagation).

     

    <- This exactly. Thanks Obst!

     

    ---

     

    In other news:

     

    Interior designers flee in panic. New part is uploaded and ready for viewing:

    Part 5: Interior Details

    • Like 2
  6. Asking for trouble how? :huh:

     

    If you keep your sealing geometry in grid 8 and 90-degree blocky and as simple and straightforward as possible, the risk of getting leaks is greatly reduced. Maintaining the mission is much easier.

     

    Using this blocky-simple-sealing-geometry -approach, I've never had much issues with leaks and weird results. For example, when I made Glenham Tower, I lost a lot of time bugfixing a weird sound leak. The sounds from the outside was leaking inside the tower, even though the tower was properly sealed and visportalled. It turned out that sealing worldspawn must be 8 units thick. I had slanted 8 units thick, so it was thinner than 8 units. This caused the sound to leak. When I widened the slanted sealing walls to slanted 16 units thick, the sound leak went away. Had I used simple, blocky approach then, I would have not lost time to seek out the issue.

     

    Also, as veteran mappers know, slanted geometry may start deteriorating due to the floating point rounding issues. At least at some point it did, it might have been fixed already.

     

    In short, hours of mapping experience indicates that keeping sealing geometry as simple as possible (8 units thick, 90-degree blocky, non-slanted) is generally predictable and safe. Using slanted sealing geometry may result in strange and hard-to-troubleshoot issues, i.e. "asking for trouble."

  7. 1. When you were building your walls you used your pre made modules. Would it not have been better to make the walls to size instead of using a series of modules? I would think there would be a performance drag using so many individual modules (especially in larger maps).

     

     

    Ah, good question! I should have talked more about it in the video.

     

    Here I had a simple walls without any features and I still used the modules. It is true that I could have gotten exactly the same result by simply making the walls the traditional way by dragging new brushes between the window/door modules. Sure, using the modules will result in little bit more tris than using the traditional way, but the small extra amount of tris is so little, it does not matter.

     

    But imagine if I had spent more time in the module building and made some really fancy looking wall modules. Now I would immediately benefit from the modular approach over the traditional way, because making a single high detail module and cloning it around gets me more details very quickly rather than building each detailed wall section the traditional way. Also the modules give be these benefits: 1) I can control each modules piece shadow casting (a performance improvement tool) and 2) it is very quick to re-arrange stuff with the models (if I build the traditional way, I cannot just reshuffle geometry pieces to get the resut I want.) If the modules were of very high detail, I could also make each module to have LOD detail models (but that's a lot of work). Oh, and also, the wall modules had already textures that were perfectly aligned with all the other modules (I used the FIT operation in the surface inspector). If I would build the traditional way, I would be facing the extra work of aligning every single wall piece textures by hand (lots of boring work).

     

    2. I know you wrote in your previous reply that there was no incentive to avoid overlapping your modules. Is that solely because they are "func_static"?

     

    Yes, the engine treats func_statics differently than worldspawn. Someone more proficient with the engine, may correct me, but I think that dmap does not optimize func_statics in any way. This way they can have any kind of weird arrangement of brushes and the engine will just treat it as a model. Worldspawn gets optimized during dmap. This may result is unexpected results if you have made weird brushwork. Thus, IMHO, worldspawn should be just used for simple geometry of sealing brushes (and floors.) Everything else is good to be func_statics. This makes the work less error-prone.

     

    The floors must be worldspawn so that the AI can walk on it. The AI does not see or feel func_statics! The AAS nagivation mesh will be applied on worldspawn.

     

    When talking about func_statics, I must remind that the AI does not 0see of feel it. All func_statics must be monsterclipped. This will have it's own part in the mapping blog later.

     

    3. Also, you mentioned that if you were using "worldspawn brushes", you would be more careful, is that the reason you are making your structures from "func_static"?

     

     

    Yes. The engine treats func_static and worldspawn differently. The main benefit is that the func_statics are grouped (they are quick and easy to clone around), and I can control the shadows of each individual module, and the func_static can be anything and the dmap will not try to optimize/change it. I always try to build so that floors, monsterclip and sealing walls are worldspawn. Everything else if func_static. It makes life so much easier.

  8. I will undoubtedly have more questions. Is it okay that I continue to ask so many detailed questions?

     

     

    Of course it is okay to ask questions. I hope I can provide an answer, though...

     

    Question: In your second video, you made a support beam with a 45-degree bracket. I noticed that you didn't deintersect the bracket so that the brush wouldn't overlap onto the other brushes. Is this not an important factor? Does DR have an intersect/deintersect option? Are BSP holes not a concern? Do BSP holes even occur in DR?

     

     

    I made the support beam 45-degree piece the way I did because it was a quick way to do it. Since the beam is a func_static, it is sort of a model, so it does not matter if the 45-degree piece penetrates into the horizontal and vertical beams. I could have cut the 45-degree piece with the clipper so that it was flush against the beams, but there is no incentive to do so.

     

    When working with worldspawn brushes, then the mapper needs to be more careful. I would make sure all worldspawn brushes are blocky, 90-degree and in grid 8. Slanted sealing worldspawn brushes is like asking for trouble.

  9. Uploading parts 3 and 4 now.

    @Jetrell,

    thanks for your interest.

     

    I joined this community around 2007 (wow a decade ago next september! Time truly flies when you're having fun!) and started learning DR. Before that I had fiddled just a little with mapping for Doom 1, Half life 1 and Farcry, but those works were never released. I was just playing around and testing stuff.

     

    So yeah, one could say that I have over a decade of hobbyist expertise with game editors. But you do not need years of experience to get basic or advanced level of proficiency. I would argue that if you watch this video series, do the A-Z mapping tutorial here ( http://wiki.thedarkmod.com/index.php?title=A_-_Z_Beginner_Full_Guide_Start_Here! ) and make your own little 1-5 room mission, then you have learned everything it takes to efficiently create missions.

     

    Each mapper might have a different mapping technique (or workflow.) The one you see in this video series is the one I have developed for myself over the years: it is fast, efficient and fun. I think, learning the DR editor is quite fast: basically you just repeat simple steps over again (create brush, texture, clone, place). But the mapping technique is the thing that takes longer to develop. Perhaps this video series provides a shortcut for new mappers?

     

    Why to map? For me it a creative exhaust. I like to create new stuff. Some people paint or draw, but I'm not very good in those, so I map. One especially fun part in mapping that it has many components: creation of a place out of nothing, telling a story, making fun gameplay, etc.

    • Like 2
  10. Thanks!

     

    I do not know how many parts it will be, but I am guessing it will be 5-10 parts.

     

    Time completion: I think the video editing, exporting and uploading is the slowest part of the process. I am more familiar with mapping than video stuff. The mapping is currently one third to half way ready, but in the end, it depends how much time I can allocate on this. Maybe, depending how busy I am RL, I'll get it ready before summer.

     

    The player starts in the back yard, as indicated by the plan.

  11. Thanks for playing, I am glad you liked them!

     

    U3 is in the works and actually it is pretty long along the path, but I do not have time to work on it due to busy Real Life . Also I'm having a sort of artistic block with it as I feel I'm not bringing forth anything new. It is basically U2 in a new environment, which I suppose would suffice to many, but I really, really, really would like to put something EXTRA in it... but do not have the time.

     

    Let's see. Hopefully, at some point, I have time to finalize it. Either quickly completed as-is.. Or with the EXTRA's.

    • Like 3
  12.  

    I'm not ignoring anything. It sucks seeing bugs in your mission after it's been released and there's nothing you can do about it. Everything that has been reported will be addressed the next time the mission is updated, whenever and however that happens.

     

    Don't whip yourself too hard, man. It happens to everyone of us. Remember the time when I released Phrase Book. There were some unpickable doors, which broke the mission for many.

     

    Let's just fix it!

     

    Howsabout quickly updating the most critical bugs, and putting the updated mission in the internet. Then add the TDM update thread a line

    "For those interested in the saint lucia prequel, the version that comes with the mod has some bugs. You can download an updated ultra-hot-fixed version of the mission conveniently from this link. Just overwrite the prologue9.pk4 in the fms folder and you are good to go."

    • Like 2
  13. @grayman,

    1) the desk settings are okay and not to blame.

    2) the book bounced in the desk when I opened it and the frobbing was impeded. Thus the desk door ceetainly had an interaction with the book and with the problem.

    3) the only visible issue is the book clipping.

    4) if the book moved on problematic opening of the desk door and there is confirmed interaction between the book and the door, the book clipping with the door is the most likely cause.

×
×
  • Create New...