Fluid beta 0.5 is out....

Developers
bob
2007-09-15
2013-06-05
1 2 > >> (Page 1 of 2)
  • bob

    bob - 2007-09-15
     
    • smappler

      smappler - 2007-09-15

      bit of a silly question but how do i install this? ive tried putting in the plugins folder and the button doesnt appear in the menu. (in 2.4). and if i do the same in 2.5 i get this error on start up
      "Uncaught exception in main method: java.lang.UnsupportedClassVersionError: Bad version number in .class file"

       
      • francois guillet

        Use a more recent Java Machine, 1.5 I guess.

        Francois.

         
    • smappler

      smappler - 2007-09-15

      hi ive tried updating everything. Java, jmf etc and still not working. im on mac os x 10.4.10. when in 2.4 the plugin shows in the SP manager but no information is shown apart from the file size. not sure if that helps.

      sam

       
      • francois guillet

        On MacOSX with Java 1.5 : Exception in thread "main" java.lang.UnsupportedClassVersionError: Bad version number in .class file

        So I suppose fluid plugin is compiled against 1.6. Java 1.6 is not available for MacOSX, not until Leopard is out, and I don't know if it will be available for Tiger.

        Francois.

         
    • Harald G

      Harald G - 2007-09-16

      Except that I couldn´t delete the baked version sometimes (?) it works pretty well with AoI 2.5 on Win XP.

      If deleting the already baked version fails it has helped me to change the particle size to convienvce  the plugin that I want it to recalculate the whole thing.

      Thankyou Deltor! I love it!

      I suggest to let the emitter ("inflow") object show their normal direction if possible.

      Harald

       
    • Nik Trevallyn-Jones

      Woo-Hoo!!  Surf's Up!!

      **attempt at humour**: Does it support themes? :o)

      Actually, as I'm sure Bob will remember, I've been hanging out for this for months, and would happily have been using it months ago even if there were no GUI, and changing parameters meant recompiling AOI :o)

      Thanks Bob!

      Cheers!
      Nik

       
    • bob

      bob - 2007-09-17

      Sorry about the class version thing. I forgot to change my eclipse build options at work. Also it included everything in my subversion project. So I will build a new one now. If that does not work i will use java1.5 to build the packages.

      Sorry I also should have said it needs java 1.5, which is the latest version for the mac. And to add to that I could only test on linux. So i'm happy that it works on XP (Try that in C).

      @ Harald G:
      Could you expand on that. Changing the particle size will not cause it to recalculate. But the back is from the time you start.
      So if i bake 5 sec. Then delete and bake again without setting the time back to 0. The new bake will be from 5 seconds onwards. Also check the logs for any exceptions.

      Finally I don't have internet at home, and I had a lazy Sunday. This week is insane at work. So you will just need to be patient for me to get back to you.

       
    • bob

      bob - 2007-09-17

      So the recompiled version with java 1.5 compliance is now avalible at the same place:
      http://www.cibiv.univie.ac.at/~greg/fluidDocs/Physics.jar

      And sometimes you need to delete the bake file to get it all "reseted". I don't know why.

      Oh and bear in mind i intended to have more documentation. Namely a explanation of each field and switch in the GUI.

       
    • Nik Trevallyn-Jones

      Hi Bob,

      I emailed you details of a modified Physics.jar I packaged with your help/mini-tute packaged inside.

      I used your SourceForge address. If you didn't get the email let me know, and I can post the details here. :o)

      Cheers!
      Nik

       
    • Nik Trevallyn-Jones

      Ok, I have already produced my first still renders, although not of the quality of some folks.

      However, I am a bit lost as to how to bake and/or render an animation. I'm not trying to "move" the actual containers, I just want the fluid to look different in successive frames (flow). I clicked the 'can move' boxes, just in case that helped, but it didn't seem to.

      So what is the bit that I'm missing?

      Cheers!
      Nik

       
    • bob

      bob - 2007-09-17

      Ok so you bake the simulation. Once that is done you now have a physics object that is *dependent* on time. So if you step forward or backward you will find the physics stuff changing. Use the set time option under the animate menu.  Finally when you finish a bake the scene time will be set to how long the bake went. So if you bake for 10 sec, and click ok, your on the frame for t=10sec. The next frame will have no simulation data, because that has not been baked. Remember that under 2.5 the preview animation does not work. I don't know why.

      When things don't work *check the logs* there are probably some stacktraces (there is sometimes even when things are ok. Error catching is not finished yet).

      I really wanted to do video tut so that all the little details I probably don't think of would be in there. Unfortunately the software for that just would not build on my systems. I mite try at home. Any ideas on what could work for this would be great.

      Once you guys get it going. let me know what you think of the general work flow and ideas for improvement. Don't too specific, general flow ideas is the level I'm at at this point. Ie the baking/set object properties options.

       
    • Harald G

      Harald G - 2007-09-17

      >>I really wanted to do video tut so that all the little details I probably don't think of would
      >>be in there. Unfortunately the software for that just would not build on my systems. I mite try
      >>at home. Any ideas on what could work for this would be great.

      I´ll try to explain how I do those kind of stuff:

      1. For most tuts I use WINK - which is quite good - it works also with sound.
         Usually I did w/o sound but to get every menu you need to make a picture on every move and
         hit a key...

      2. For more fluid (hehe) tutorials (aka videos) I have used Camstudio
         http://camstudio.org/
         which has been updated recently!!!  Free and nice

      3. If you need it faster and more comfortable and could live with the fact that you use a
         commercial software - than Camtasia Studio is the way to go (with windows). Version 2
         (old but very good - the Techsmith video codec alone is worth the money) - is available on
         Chip - Foto & Video Magazine in Germany - and I guess in Austria also. It costs 4,20 Euros...
         (You´ll need issue 9/2007 - the lastest is 10/2007) - Maybe asked some
         picture-taking enthusiasts.

      After recording that way I use a video application - that is not neccessary with Camtasia Studio because everything is included there. A cheap alternative for CamStudio would be Virtualdub (just for hard-cuts).

      HTH
      Harald

       
      • bob

        bob - 2007-09-18

        I believe they are all windows. We have about 30 boxes here at work, and I have 2 at home. We don't have windows on any of them. Wink is nice but only produces flash and it simply would not capture open menus (My attempt is on the blog). I started to write a quick one as a plugin for AoI. Screen captures are fast but saving the jpg takes seconds, so i dropped that.

         
        • bob

          bob - 2007-09-18

          the msg:
          By: bob (delt0r) - 2007-09-18 07:09

          Was a reply to nik777 about making movie tuts. Sorry, i just don't quite get the SF forums.

           
    • Nik Trevallyn-Jones

      Hi Bob,

      Thanks for your clarification. It seems it works fairly much as I was expecting, so I shouldn't be too far off getting a good result. :o)

      (i) I followed your really helpful tute, and baked 10 seconds of flow.

      After the bake was complete, I opened the animation score.

      (ii) If I moved the (green) timepoint in the animation score, the displayed physics remained the same: always showing the state as it was at the end of the bake (t=10)

      ** There was actually a slight difference the first time I moved the green time point. As I dragged the line for the first time, I seemed to see the display change, out of the corner of my eye. Thereafter, moving the green time point had no visible effect.

      (iii) If I rendered a movie from t=0 to t=10, I was seeing multiple frames all showing how the bake looked at t=10.

      Having gone back to look, I see that there are *small* changes between the various time periods. For example, changing from t=10 to t=0 shows a small change. However, the state at t=0 does not seem to represent how the bake looked at the beginning. Moreover, changing from t=6 to t=7, for example, seems to have no effect at all.

      This may actually be all correct, and it's just that my expectation is incorrect. I do realise that fluid flow will create multiple states over time that look very much like other states. However, I had expected the beginning of the animation to look like the beginning of the bake - ie very little liquid, that then grows as more pours in.

      I'll mess with it further later today.

      Cheers!
      Nik

       
    • Kevin D.

      Kevin D. - 2007-09-18

      This isn't really a problem, just an observation.  Has anyone noticed that the Fluids icon, the polymesh icon, and the implicit object icon seem to swap places on the tool bar each time AOI is restarted?

      Take a look at this:
      http://www.mts.net/~zorksox/screen1.PNG

      This isn't a big deal to me, I'm just curious.

       
    • bob

      bob - 2007-09-18

      Dammit: so I stick to just writing everything i a single post.

      The previous was a reply to vidiot.

      nik777:
      there clearly is something going wrong with the bake save. Check the logs. Even better turn of the logging plugin and pipe the output. I bet you will see a lot of stack traces each bake frame. This could be for lots of things. I still haven't found all the bugs in the file code. But I find that if nothing works. Exit AoI, and remove the bake file, and then restart AoI. For some reason the file gets corrupted, i had hoped that it was now fixed.

      Well the only other thing would be the normal value you set to inflow. If its small not much will happen. But i don't think this will be the problem.

      Kevin D.:
      Yes i noticed it too. Its a 2.5 thing. Because it does not break anything i didn't bother to follow it up. its a "polish" thing. I will rewrite all the plugin code at some point. It was just to get the thing up and running.

       
    • Nik Trevallyn-Jones

      Hey Bob!

      > there clearly is something going wrong with the bake save.
      Uh, ok. I had assumed the problem was between seat and keyboard.

      Here is an excerpt from the bottom of the LogPlugin preview:
      5:57:24 PM EST [AWT-EventQueue-0] in SPHFile.readPrevHeader() (SPHFile.java:112)
         PRev FilePos:1002218    size:60523251    frame:265

      5:57:24 PM EST [AWT-EventQueue-0] in SPHFile.readPrevHeader() (SPHFile.java:120)
         Prev ++ FilePos:751511    size:60523251    frame:264

      5:57:24 PM EST [AWT-EventQueue-0] in SPHFile.readPrevHeader() (SPHFile.java:112)
         PRev FilePos:751511    size:60523251    frame:264

      5:57:24 PM EST [AWT-EventQueue-0] in SPHFile.readPrevHeader() (SPHFile.java:120)
         Prev ++ FilePos:501020    size:60523251    frame:263

      5:57:24 PM EST [AWT-EventQueue-0] in SPHFile.readPrevHeader() (SPHFile.java:112)
         PRev FilePos:501020    size:60523251    frame:263

      5:57:24 PM EST [AWT-EventQueue-0] in SPHFile.readPrevHeader() (SPHFile.java:120)
         Prev ++ FilePos:250723    size:60523251    frame:262

      5:57:24 PM EST [AWT-EventQueue-0] in SPHFile.readPrevHeader() (SPHFile.java:112)
         PRev FilePos:250723    size:60523251    frame:262

      5:57:24 PM EST [AWT-EventQueue-0] in SPHFile.readPrevHeader() (SPHFile.java:120)
         Prev ++ FilePos:24    size:60523251    frame:261

      5:57:24 PM EST [AWT-EventQueue-0] in SPHFile.readRecord() (SPHFile.java:155)
         RecordRead FilePos:24    size:60523251    frame:261

      5:57:24 PM EST [AWT-EventQueue-0] in SPHSolver.initHash() (SPHSolver.java:975)
         Grid:143    231

      5:57:25 PM EST [AWT-EventQueue-0] in ImplicitFluid.getRenderingMesh() (ImplicitFluid.java:378)
         GettingRenderingMesh:

      5:57:34 PM EST [AWT-EventQueue-0] in ImplicitFluid.getRenderingMesh() (ImplicitFluid.java:378)
         GettingRenderingMesh:

      ... I note that the frame numbers are going backwards...

      Cheers!
      Nik

       
    • Nik Trevallyn-Jones

      Thought I'd put these in a separate post...

      > Even better turn of the logging plugin and pipe the output.
      Ages ago I got tired of having to click the LogPlugin preview button during debugging, and added a feature to pipe to the output to stdout as well (classic unix "tee" function). However, a subsequent re-org of the code broke part of that. An update of the LogPlugin will be released shortly (within a week, I hope) which will have that bug fixed, and some new features added as well.

      > I started to write a quick one as a plugin for AoI.

      Would you be prepared to send me the code of what you have (please, oh please)? I would like to try connecting the streaming output pipeline from AdvancedRendering to your capture code, to see if we can make something workable between us. A tutorial capture plugin for AOI would be awesome!

      Cheers!
      Nik

       
    • bob

      bob - 2007-09-18

      The output looks fine. It should be working. What OS do you use? Have you tried dragging the views around to ensure they refresh? what is the partical count?

      About the only thing i can think of is to delete the bake file and restart. Otherwise what are the preview tab options. you may have clicked one off. Also the "box" that is the fluid object is the only thing to contain fluid. So fluid will get removed outside its bounds. Furthermore the thing does not like getting moved around after its been baked, and then rebaked. But none of these things would do what you have described.

      Could you tell me the configuration (I just realized that a config to txt would be useful), or even send the *.aoi file?--mm send how.. Well provide a link to it.  

      As for the tut creation code. It wasn't tested at all. I might test it later today when work gets me down.

       
    • bob

      bob - 2007-09-18

      Oh i forgot. Moving the green "time" line can be really really slow. Its trys to load a *lot* of data for each frame. It tends to freek out a bit.  Tyy using ctrl-[ ctrl-] instead or use the getToTime  setTime  in the animation menu(I forget the exact option).

      By now you should see why i say its very much a beta.

       
    • Julian MacDonald

      I had the same problems as Nik on *some* occasions, i.e. as you move the timeline, nothing dynamic happens.  This seemed to be resolved by manually deleting the bake file and re-baking. 

      By the way, I assume the bake file is referred to in the scene and that there must therefore be a separate bake file for each fluid object ?  If this is the case, it's presumably important to not to move the bake file ?

      Great plugin, by the way - thanks for your hard work !

      Julian

       
    • bob

      bob - 2007-09-18

      the bake tab has the file name there, and you can change it. ie point to a different "bake" if you like. The bake files are probably even platform inderpendant, since java should take care of little endian stuff (Macs should be the same anyway)

       
    • Nik Trevallyn-Jones

      Hi Bob,

      I hadn't tried dragging the views around.

      I'm on linux (Fedora Core 6), running Sun JDK 1.6.

      Moving the green timeline is very fast, so it sounds like my bake file is probably wrong. Whether that's my fault or something else, I can't tell. :o)

      I've put the file here:

      http://nik.homelinux.net/files/Fluid-demo.aoi

      I did render a complete 10 sec movie. Watching the rendered images, those that I saw all looked identical. And the resulting "movie" file looks too small to actually be a movie.

      As for the tut code, I'm happy to look at whatever you have.

      Cheers!
      Nik

       
1 2 > >> (Page 1 of 2)

Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

JavaScript is required for this form.





No, thanks