| Last Modified On : | December 16, 2009 9:22 AM PST |
Rate |
|
| What If Home | Product Overview | Intel® TM ABI specification | Technical Requirements FAQ | Primary Technology Contacts | Discussion Forum | Blog |
|
Smoke is a tech demo that showcases a framework to support n-way threading of game technologies. By properly threading a game it can have more accurate physics, smarter AI, more particles, and/or a faster frame-rate. Smoke demonstrates one way to achieve better games. All of the source code for Smoke is available for download. The demo includes:
|
What's new with Release 1.2
|
Video: Smoke Overview Video & Tech Overview Video
Article: An Overview of Procedural Fire & Designing a Parallel Game Engine
Installation: You can download the installer for the demo.
You can download all the source code here. There are build instructions included in the source code zip file under smoke\docs
You can also watch the other video done to Take an in-depth look at Intel’s Smoke Demo
Q: Where can I get the source?
A: All the source is available.
Q: How do I build the source code?
A: There are build instructions included in the source code zip file under Smoke\docs.
Q: How do I see the advantages of threading?
A: While the demo is running you can toggle the number of active threads.
Q: What is the license?
A: You can read the license here. The goal is to allow developers to use the Smoke code written by Intel for any purpose (including derivative works) with no limitations or obligations.
Orion Granatir is a senior engineer in the Visual Computing Software Division. He is the Tech Lead on the Smoke project. Prior to joining Intel in 2007, Orion worked on several PlayStation 3 titles as a senior programmer with Insomniac Games. His most recent published titles are Resistance: Fall of Man and Ratchet and Clank Future: Tools of Destruction.
| December 4, 2008 9:50 AM PST
Ricardo Bicalho | Does it work with Windows Vista 64 or is it 32 bit only? |
| December 4, 2008 4:51 PM PST
Orion Granatir (Intel)
|
Thanks for the comment DDd ^_^ I saw your post on AiGameDev.com, I'd like to hear what you end up doing with Smoke |
| December 4, 2008 4:52 PM PST
Orion Granatir (Intel)
|
Hey Ricardo, We built and tested Smoke using 32bit. You can try and build and run 64... let me know how it goes! |
| December 6, 2008 6:09 AM PST
James Munro |
Hey guys, I think this is an excellent project and something that deserves more attention in undergraduate games development based degrees. I'm a third year student at the University of Lincoln, UK doing a games based degree, but we don't really cover any basic engine fundamentals, let alone threading and paralell game design. As part of my dissertation I'm doing an in-depth analysis of various open-source game engines, particularly the Quake-series. I'm planning to do a chapter on the importance of paralell game engines in today/tomorrow's games industry and all of this information you have provided will prove very useful. I had a quick play at bringing some threaded functionality to the Q3 engine (though it is very, very basic). I have a short blog entry about it here: http://learninglab.lincoln.ac.uk/blogs/jmunro/2008/12/02/att.....quake-iii/ Thanks for the hard-work! |
| December 8, 2008 9:38 AM PST
Orion Granatir (Intel)
|
Hey James, I'm glad to hear about your interest in parallel game design. I'm working with Intel Software College (ISC: http://software.intel.com/en-us/college) to develop material about parallel game design using Smoke. Here is a link you might be interested in: http://software.intel.com/en-us/videos/gdc-session-threading..... ars-part-1 Anu has done some interesting working threading Quake 4. It sounds like a very interesting dissertation. Please keep me up-to-date and let me how your project turns out. |
| December 13, 2008 9:04 AM PST
CJ | Thank You!! I can't wait to check this out. I'm on an Intel Mac... I believe all the libraries you mention are cross-platform. Have you had anyone test out the engine on OSX yet? |
| December 13, 2008 9:27 AM PST
CJ | woops, I missed the mention of the win32/Dx dependency, though I know Ogre3d claims abstraction over Dx and OGL. What might prevent this from running on an OSX intel mahine? |
| December 15, 2008 2:17 PM PST
Orion Granatir (Intel)
|
Hey CJ, No one has tried in on OSX. I would be very interested to know if you can get it to work. The biggest dependencies on Windows are DirectX and .NET. DirectX is used by Ogre3D... however, you should be able to use OpenGL instead of DirectX (OpenGL is just a separate plug for Ogre). The scripting system uses .NET... but you can feel free to disable the scripting project (it's not used in the demo at the moment). In the \bin folder you will find Smoke.gdf... delete the reference to the scripting system (you might need to remove it from Demo.cdf too). Now here are the big challenges. There are a few systems that also depend on DX. These is SystemInput and SystemFire. SystemInput uses DirectX's input support. But that's easy enough to replace (or disable). SystemFire uses DirectX for some of the math calculations. Our math library should support all the needed functionality... we just didn’t have time to get all the DX out of there. Good luck and let me know how it goes! |
| February 4, 2009 7:54 AM PST
c0d1f1ed
| Is there any way to make the demo stop spinning like crazy? :) |
| February 5, 2009 10:40 AM PST
Orion Granatir (Intel)
| Yeah! Move the camera and it will stop moving automatically for 30 seconds. Also, search for CamBot in Demo.cdf and comment out that line to disable it completely. |
| February 6, 2009 6:56 PM PST
som-dylan
| Hello ! |
| February 11, 2009 8:16 AM PST
c0d1f1ed
| My apologies, the spinning was actually caused by having my joystick plugged in, which was causing control issues in other applications as well. It wasn't a critique on the smoothly moving camera. |
| February 24, 2009 10:25 AM PST
Raghava Darisi (Intel)
| cool demo |
| March 12, 2009 3:13 AM PDT
Marisa |
Does it exist any Linux version? Yours, Marisa |
| March 12, 2009 9:13 AM PDT
Orion Granatir (Intel)
|
Hey Marisa, You would have to port it to Linux. Read my comment above about porting the demo to OSX. It would be a very similar effort. Good luck and let me know if you give it a try. Thanks, Orion |
| March 12, 2009 8:35 PM PDT
Gastón C. Hillar
|
Hi Orion, I've tried this demo today, as I've discovered this community a few weeks ago... No comments, just awsome. :) Cheers, Gastón |
| March 13, 2009 9:05 AM PDT
Orion Granatir (Intel)
| Thanks Gastón ^_^ |
| March 26, 2009 4:15 PM PDT
Rodrigo Colares |
Orion, this is a great demo for developers and students of game development. I have a particular question not about the demo, what are the essential requirements to be a console game programmer as Playstation 3? |
| March 27, 2009 9:58 AM PDT
Juan Toro |
Hello Orion, We´re evaluating how easy would be to implement some generic 3D interaction techniques in Smoke, and we wonder if you could point to us some generic information before we start our development. We have as reference the interaction techniques mentioned in "3D User Interfaces. Theory and Practice", from Bowman, Kruijff, LaViola, and Poupyrev. Could you please answer the following questions with a number between 1 and 9 (1 for difficult, 9 for easy) and a short comment. 1. Integration of novel devices, i.e. Wiimote or phasespace trackers (http://phasespace.com/productsMain.html) 2. Other selection techniques (i.e. select by a ray, select by colision with a virtual hand) 3. Other travel techniques (i.e. move to a visible object, move by a predefined path, move to an object by name) 4. Show a map, a gamer´s position, a gamer´s known world, and a target. Thank you for your time. |
| March 30, 2009 11:19 AM PDT
Orion Granatir (Intel)
|
Hey Rodrigo, Thanks for the interest. Whether you are programming for the console or the PC, it's essential to be a solid C/C++ programmer. There is a lot of love hate for C/C++, but it's still the most popular language with just the right level of access to system resources (threads, memory, etc). Future versions of consoles and PC are going to require more and more threading. And threading expertise is lacking in the game industry (but it's getting better). I hope that helps. If you have future questions about non-Smoke related topics, please feel free to contact me directly so we can keep this comments/discussion on topic. Thanks, Orion |
| March 30, 2009 11:43 AM PDT
Orion Granatir (Intel)
| Hey Juan ^_^ Those are some good questions. Adding generic 3D interactions should be easy enough; there is a separate system (SystemInput) that is designed to process user interactions. The code in this system is very minimal and easy enough to replace/extend. Here is some feedback on your questions. Number 1 would be easy enough if you already had the code to read the input (e.g. reading input from the Wiimote). So about a 7. For number 2, this might be a little more challenging because of the need for ray cast. There is already code to support calls to the physics system from other systems... but it's not being used at the moment and not that straightforward (give this one a 3). Number 3 is interesting. There is already code for scripting (SystemScript) that was never actually used. You could extend this system to do most of these travel techniques... or just write a simple AI. I would give this one about a 7. Number 4 is easy... if you can figure out how to use the overlay code in SystemGraphicsOgre. Give that one a 8 if the overlay code looks easy or a 5 if not. These are all very interesting ideas! Please keep me up-to-date on your progress. Thanks, Orion |
| April 11, 2009 7:29 AM PDT
19900904
| I am very surprise to see the project. |
| April 12, 2009 2:06 PM PDT
DDd
|
Really good stuff. Thanks for the GDC09 videos, very nice/fun and illustrative. I have yet to do anything with Smoke because i don't have a proper CPU that can take advantage of the threading features, i am on a dual core CPU. However i have read the articles, seen the videos and looked at the code, they gave me a few ideas that perhaps i will implement in a entry for the Level Up 2009 competition ;) |
| April 12, 2009 11:54 PM PDT
rg |
hi anybody use this kind of thing for ultra hi-frequency stock market trading? might sound silly, i don't know...???? is the messaging faster that FIX 5.0 or FAST 1.2 protocols? Is the AI somehow more integrated or enhanced for statistical learning than compteing technologies?? how fast and reliable is the concurrency? vs. erlang level robustness? vs CUDA or Cilk++ speed? how many porcessors are needed before the distribution cost is outweighed by the speed advantage? More than 4, or in the hundreds? I am truly interersted and able to deploy if there is an advantage here - rg |
| April 13, 2009 9:30 AM PDT
Orion Granatir (Intel)
| Hey DDd, Thanks for the feedback. For those that don't know, you can see the GDC all day threading tutorial here: http://software.intel.com/en-us/articles/intel-at-gdc/. This includes some useful information about Smoke! Creating an entry for the Level Up 2009 contest would be fantastic. Please let me know how it goes :D |
| April 13, 2009 9:45 AM PDT
Orion Granatir (Intel)
|
Hey rg, I can’t say a lot about Fix 5.0 or FAST 1.2; however, I believe those are both network protocols. Smoke is about sharing data among multiple threads running on the same processor. The Change Control Manager idea could be extended to work over a network, but this would introduce increased latency. The concurrency is fast and reliable (we did analysis for deadlocks, race conditions, etc). However, Smoke was designed for games, not mission critical applications. You can still apply the concepts to your field of expertise, but please do your own validation. The AI system is easy enough to rewrite for more traditional AI (e.g. for data stock analysis). This framework was designed for parallelism on full function and powerful cores (e.g. an x86 core) and parallelism scales well with the number of cores. Parallelism for something like CUDA would require major rewrites of code (especially in regards to data management and propagation). I hope this information helps. Please keep me informed about your work with Smoke :D Thanks, Orion |
| April 16, 2009 11:34 AM PDT
RSR | Где русские? Хехе. |
| May 1, 2009 3:05 AM PDT
Jon Gagnon | Thanks for the great demo, the performance in 1920x1080 on my system (Core i7, ATI4870) is fantastic. Interactivity works well with Xbox 360 controller...where is the functionality for this implemented? I dont see anything in Input.sdf - I'd like to try and map the water hose function to a button on the controller. |
| May 1, 2009 5:46 PM PDT
Orion Granatir (Intel)
|
Hey Jon, Check out Scene.cpp in SystemInput. This is the main file that sets up DirectX input to read the controller and keyboard. It should be easy enough to extend the functionality to make a key activate the water hose. Let me know how it goes! Thanks, Orion |
| May 19, 2009 9:13 AM PDT
Kingpin |
This seems related to http://www.projectoffset.org way to go keep it up for indies!!!! |
| May 30, 2009 10:23 AM PDT
bill | come on guys.put a moster in it.just houses bruning.give it some real action.put a evil vilen in it that brun stuff |
| June 1, 2009 7:06 AM PDT
weicheng |
I am curious that how you organize all the links between subject, observer, subjectsystem, observersystem? is there a tool or you just manually added them into scene files? I think this is the key issue for the framework going to production end. Thanks |
| June 1, 2009 9:18 AM PDT
Orion Granatir (Intel)
|
Hey Weicheng, We did all the links between objects manually :-p It was a pain and was obviously a hotspot for errors. One thing we always wanted but never had time to implement was a level editor, something that would output the XML file with all the links. It's still on our wish list... we just need to find time or someone to help. Thanks, Orion |
| June 25, 2009 7:17 AM PDT
currentloops
|
Hello Orion, I am investigating the use of VmWare to increase the quanitity of dedicated processes to simulation functions in my aerospace project. Have you tried running Smoke virtually? |
| June 25, 2009 9:31 AM PDT
Orion Granatir (Intel)
|
Hey CurrentLoops, No, I have not tried running Smoke virtually. That an interesting idea... however, I believe the biggest problem will be the speed of virtualized DirectX. I'd love to hear about the results if you give it a try... enter you results in the Thread Like Wildfire contest here: http://software.intel.com/en-us/contests/thread-like-wildfire/con tests.php and you'll probably win a $100 gift card :D |
| June 29, 2009 11:55 AM PDT
currentloops
| Ok, I anticipate receiving our Blade Server and vSphere ESX in the coming month, so will try out my "Eye Of The Predator" Visual Adrenaline entry with the instrumented Intel metrics to record the comparison in fps. The objective is to separate out each networked player as a virtual participant for regression testing and use HP Loadrunner and HP Diagnostics to monitor system performance. I will report the results on the forum then. Cheers |
| July 2, 2009 5:27 AM PDT
Koichi Senada
|
Has anyone managed to port Smoke to any other IDE than Visual Studio 2005? I am trying to compile the Smoke solution with Visual Studio 2008. Seems like these comments are more popular than Smoke related forum where I have posted my question at http://software.intel.com/en-us/forums/smoke-game-technology-demo/topic/66696/ |
| July 6, 2009 4:26 PM PDT
Orion Granatir (Intel)
|
Hey koichisenada, I'll post a longer reply on your forum post :) Short answer: We did manage to get it running on VS2008, but there were some limiting factors to shipping that version. Thanks, Orion |
| July 8, 2009 3:52 AM PDT
didito |
hi intel smoke team, thanks for this great technology. i've wanted/tried to make something similar - threading different subsystems for "realtime" applications - for a while and smoke definitly comes in handy. i'll need some time to look through the code and documentation though. one question i have right now - you mentioned in one of your presentations that changing the physics subsystem was very easy. do you have any information on that? i try to build a crossplatform solution (macosx and windows, maybe linux later) and i would love to use havok but it seems that it does not work with macosx. so i'll try bullet for now. do you have any information/tutorial/documentation on how to generally implement a new subsystem or how to change the physics system. which one did you use for swapping? did you use (O)PAL? i hope you guys get more time from intel to maintain the code, develop and document it further, because this is an awesome and very valuable project! thank you very much and keep up the good work! didi |
| August 4, 2009 3:55 AM PDT
Koichi Senada
|
hi Intel Smoke team! I have posted another question to http://software.intel.com/en-us/forums/smoke-game-technology-demo/ This time it's related to Windows Messaging related functions like PeekMessage, which can be received correctly in SystemGraphics only and related to receiving those messages at SystemInput. Also, is that possible to talk to you by some instant messaging systems or email? |
| August 12, 2009 3:36 AM PDT
Koichi Senada
|
hi Intel Smoke team! We seem to have solved our issue related to Windows Messaging on our own already. Is there any work on Smoke lately? |
| August 21, 2009 6:20 AM PDT
Amit Makhija
| this can work on Pentium 4 or not?? |
| August 24, 2009 8:35 AM PDT
Koichi Senada
|
hi Amit Makhija! Sure thing, Intel Smoke Demo can run on Pentium 4 CPU. I am sure it can run on a single processor and single core CPU. Why did you ask? |
| August 27, 2009 10:29 AM PDT
Orion Granatir (Intel)
|
Hey Koichi, Thanks for posting on the forum about your issues related to Windows Messaging. Please keep up the posting on the forum. Even if there are not a lot of replied, we have a lot of people reading them :D I'll head over to the forums after reviewing these comments. Thanks, Orion |
| August 27, 2009 10:30 AM PDT
Orion Granatir (Intel)
|
Hey Amit Makhija, Koichi is right; it should run on a Pentium 4. However, we did not test it so I can't be 100% sure. The project is designed for modern multi-core processors. |
| September 9, 2009 8:14 PM PDT
QiangF |
Hello Orion i study the code, find that the render system spends much more time, can't it be optimized? |
| September 12, 2009 6:29 AM PDT
jsmith | How long is this demo supposed to run ? ,(Until you stop it?) , I get 24-26 on e5200 @ 3.96 ,Not bad?..... |
| September 12, 2009 8:00 PM PDT
jsmith | The first 2 passes on e5200 @ 3.96 give me 24fps after the 3rd pass , it averages 16fps ,....Anybody try this on older Pentium d 965 ee with hyper-threading ? Just curious on it's Hyper-threading capabilities... |
| September 21, 2009 2:43 AM PDT
zanders |
The demo video shows utilization for 8 cpus (reading all in 90%s)-- was that a quad processor and if so the calculation is misleading/inexact as the real utilization on a quad could only be a max sum of 400% (waiting for cache lines to fill shouldnt count as 'execution') If it was 2 quads then 'never mind' .... |
| October 19, 2009 12:20 PM PDT
Mike Yi (Intel)
|
A couple of the developers on the Smoke team are working on the Smoke project, focusing on performance optimization. We'll be releasing articles, videos and blogging about what we did and what tools we used to extract the most performance out of Smoke. Hopefully, game developers will find these series of articles helpful in reaching greater performance in their games. We have already presented our first phase of optimization work on Smoke at the Intel Developer Forum in September. Luckily, the presentation was recorded and is already live on Intel Software Network. Part 1 of the video can be found here: http://software.intel.com/en-us/videos/optimizing-a-video-ga.....rn-part-1/ which also has links to the rest of the presentation. The slides can be found here: http://www.intel.com/go/idfsessions The presentation reviewed the demo code design and then showed a complete performance study, with step-by-step use of Intel tools: - Benchmark and measure a baseline - Find common memory and data race bugs with Intel® Parallel Studio - Drill down into hot spots in the code, and highlight why they're hot with Intel Parallel Studio and Intel® VTuneTM Analyzer - Find concurrency problems with Intel® Thread Profiler - Show some speedups made in the code We are planning to release the updates we made to the Smoke code in the coming weeks, which includes a performance speed up and a port of the code to Visual Studio 2008. Both an executable and source code will be made available, as usual. Please let us know if you have any suggestions for making Smoke better! |
| November 11, 2009 7:49 AM PST
Abhay
|
Pure Awesomeness!! Thanks for the Smoke team for providing a great demo. It was really 'smokin' ! Orion, i have a question. I would like to create a similar game scene for my college project and perform a bencmarking analysis on different Intel processors . Could you briefly please explain me as to how i can make use of the Smoke source code/improvise to create a custom made video. |
| November 11, 2009 11:52 AM PST
Orion Granatir (Intel)
|
Hey Abhay, I assume you want to change the content of the demo. If you look at demo.cdf, this file is the description of the scene. It's a bunch of XML that tells the app where to place objects, how they interact with objects, and other relationships. You can see, most of the content is mesh files to be loaded by Ogre. The artist that worked on the project created content in 3DS Max and used a plug-in to export content in a format for Ogre. Visit the Ogre3D.org for more info on generating content for Ogre. |
| November 13, 2009 2:10 PM PST
Paul Lindberg (Intel)
|
For more info on changing demo content, turning different systems on and off, moving existing content around, etc., check out the "Configuring Smoke" doc, in docsConfiguration.rtf. It shows the different kinds of config files you can edit. You can also look at a little different content in the slightly-newer release of Smoke from GDC this year (especially the Zombie). :-) Check out the source zip from the Game Threading Tutorial at http://software.intel.com/en-us/articles/intel-at-gdc/. The Zombie content is also in the upcoming R1.2 Smoke release (although we haven't dropped him into the default scene - hm, may need to think about that). |
| November 16, 2009 10:55 PM PST
Abhay
| Thanks Orion & Paul. I am currently working on creating a customized rendering scene with the basic objective of running it better on Intel hardware. May not be as marvelous smoke but still :) ...taking my first steps into the world of game development :) |
| November 18, 2009 3:43 AM PST
Koichi Senada
|
How came there's no hint on "Game Developer Conference 2009 Threading Tutorial" existence around here? http://software.intel.com/en-us/articles/gdc-2009-threading-tutorial/ Any details more on that tutorial? It takes time to navigate through all the branches of this website to find anything related to Smoke. Organize with tags, labels, categories, please? youtube.com knows how to do that. Thanks! |
| November 18, 2009 3:53 AM PST
Koichi Senada
|
On the other hand, there are tags introduced for other modules, but the "Articles". Those modules are: - Knowledgebase - Forums - Videos - Blogs - Contests "Smoke" tag for them is located here: http://software.intel.com/en-us/tags/3648 Is there a chance to have tags associated with articles, as well? By the way, now I know, GDC related article provides a link to dowloading the "Intel Game Threading Tutorial" files: http://software.intel.com/en-us/articles/intel-at-gdc/ |
| November 18, 2009 4:17 AM PST
Abhay
| I agree with Koichi . The forum needs to be organized in such a way that browsing can be easier. |
| November 25, 2009 7:45 AM PST
kunanaya
|
Good Good Good Excellent Excellent Excellent WHAT ? This link tools and video , Intel software network is key ! to be master in IT. |
| November 30, 2009 12:23 PM PST
Dave Willis |
sorry sorry sorry At 69 and a new comer to windows, with a inborn trait of questioning everything, it is driving me nuts. I have just downloaded Audacity and as good as it is I am told, it too no doubt will drive me nuts too. So, sorry to say and with no disrespect, only when I have become more PC mature will I be able to even contemplate smoke-game-technology-demo. but, thanks for the opportunity anyway and good luck to all who have entered.. Cheers and goodwill to all Dave Willis |
| December 17, 2009 3:32 PM PST
Paul Lindberg (Intel)
|
Smoke R1.2 is now out. More details at the top of this page and on the download page, but it now builds with Visual Studio 2008, runs faster, and has some bugs fixed. Excellent comments about ways the forum could be organized, indexed, and searched, thanks Koichi and Abhay! We're looking at it now. |
| December 17, 2009 3:35 PM PST
Paul Lindberg (Intel)
| I hope you found the info from this year's GDC tutorial useful, Koichi. We're looking at possibly doing it again in 2010. What would be the best kinds of material we could share in that tutorial? How could we best share it, with those in the room, and with everybody else in the game development community? |
| December 18, 2009 5:59 AM PST
Koichi Senada
|
Paul, thank you for the commitment! I am downloading the Smoke Rev 1.2 Source. That's a very nice gift for the Christmas! Also, thank you for asking about the possible ways of further discussions. I think I will share the common opinion, that everybody wants to do about Smoke is to extend and adapt it. Thus, could you guys please let us know your vision on possible ways of extending and adapting it? For example, how to replace and extend current systems for graphics and input, and probably others. In our case, we are curious to know where to put the window message peeking and dispatching loop, where to create the window and DirectX entities so, that input system is also able to read keyboad, mouse, joystick, and especially multitouch data from the loop, yet to keep it robust. That's a serious case, indeed, programmers often note that it's hard to separate input and graphics under Windows environment, while it's relatively easy to separate audio, physics, artifical intelligence, logics, scripts, etc from them. The trouble is that windows, graphics, input all are tied to the window messaging loop and thread which creates the window. Some more nuts'n'bolts about this spot is very appreciated. Thanks! |
| December 18, 2009 6:13 AM PST
Koichi Senada
|
By the way, if anyone is intrested to see what we have managed to produce so far using Smoke and why is multitouch so important to our team, please check out our posted videos: http://www.youtube.com/user/vigroup Most of the videos posted there are related to the Smoke based engine. More to come soon! Those who can speak russian can visit our blogs to read about our development activities: http://multivi.ru/blog/ http://blog.vigroup.ru/ You all can also use http://translate.google.com/ to read our blogs translated from russian, of course. Thanks for the Smoke! |
| December 18, 2009 3:54 PM PST
Paul Lindberg (Intel)
|
Koichi, very cool videos! Which ones are actually running Smoke code? Good idea to have something written on ways to extend Smoke, but you have everything we've written so far on the subject. :-( Let me think about the best ways to answer this and get back to you. I haven't seriously thought about decoupling the input processing from the rest of the main loop in Smoke. What's your biggest concern - quick response time to input, minimal dropped input, easy access to your multi touch messages, or ...? Or all? :-) |
| December 19, 2009 6:53 AM PST
Koichi Senada
|
Paul, applications that are running Smoke code are those that display 3D graphics with reflections and refractions: http://www.youtube.com/watch?v=dzckzmqdDMg http://www.youtube.com/watch?v=Iw5VxBN_vM0 http://www.youtube.com/watch?v=dY5XpOlISb0 http://www.youtube.com/watch?v=BhVlWbEG_Yo http://www.youtube.com/watch?v=ksrNus1uvQE Though most of the code is placed inside the rewritten SystemGraphics, becouse it's not as easy to separate multitouch related handling as you have done it with keyboard and mouse handling. SystemInput is only slightly involved, as it serves for keyboard and mouse input as secondary and debugging input handling. SystemAudio is also completely rewritten to use OpenAL instead of FMOD. SystemPhysics is yet to be completely rewritten to use Nvidia PhysX. SystemGeometry is to be extended to support hierarchy, though meanwhile it is supported by SystemGraphics only as an additional property named "parent". I have searched trying to find an alternative way to receive multitouch data under Windows, but it appears that window messaging loop is the only one available. Thus the most important question is still the same, how would you recommend to realize the multitouch handling and other window messaging related code? Would it be a good idea to place it somewhere between the synchronization frames, so that very next frame every system has the most current data about the input? Would it be a good idea to keep some part of input handling code there inside the SystemGraphics? By the way, we keep talking to russian speaking representatives of Intel in Moscow, and they recommend to keep discussions about the Smoke here, in this part of your website, so we just do that. Any advice and recommendation from your team is highly appreciated, as it speeds up and improves our development process. |
| December 20, 2009 4:59 AM PST
Koichi Senada
|
Paul, by the way I have also found yet another way to organize and categorize the Smoke related resources on this website based on the architecture that already exists. This time it's all about the articles only, but how surprised I am having found that some articles are grouped! What I mean is a collection of articles subsets: http://software.intel.com/en-us/articles/tools/ Maybe the Smoke related resourced can fit in? |

English | 中文 | Русский | Français
Mike Yi (Intel)
| ||
Orion Granatir (Intel)
| ||
Paul Lindberg (Intel)
|
DDd
210
Status Points:
0