You are not logged in.
I've finished a first draft of the physical design I didn't want to do:
https://ensmer.dev.java.net/physical.html
I figured I should put this in a new post since it won't go off topic quite as fast. :-D
I posted a more involved post here:
https://ensmer.dev.java.net/servlets/Pr … ageID=4427
I'd like the replies to go there, but I think you have to be a java.net member, which sucks, so I'll keep reading here too if you want to comment.
Thanks for tearing it apart; I know there's a lot of changes required so...
I'm going to start coding next week, but that doesn't mean the physical design has to be completely finished by then; I'll just work on the stuff I'm most comfortable with the design.
Dusty
Offline
just an update: I've 'officially moved the hompage to http://ensmer.dev.java.net/ I set up a discussion board to hold 'journal' style posts, as wdemoss suggested. I've also edited the physical design a little bit since I originally posted it.
Dusty
Offline
Hey folks,
Such a shame about the interest in Ensmer dying down, guess I'll have to try to revamp it. Kind of sucks that I don't have screen shots yet (image a black screen), but I do have some sourcecode available in the CVS repository at ensmer.dev.java.net. See here:
https://ensmer.dev.java.net/development.html
and here:
https://ensmer.dev.java.net/source/browse/ensmer/
for details on how to access it.
I'm documenting my progress in a special forum here:
https://ensmer.dev.java.net/servlets/Fo … orumID=992
today's work listed here:
https://ensmer.dev.java.net/servlets/Pr … ageID=4509
Let me know what you all think. I'm begining to believe I can actually pull this off in the time allocated...
Dusty
Offline
Such a shame about the interest in Ensmer dying down, guess I'll have to try to revamp it.
The interest is there but I don't think anyone will or be able to comment much on a document. You did a right decision to provide some source codes...."we" want to try the actual thing, see with our own eys how does it work, then we comment
Markku
Offline
not to mention you told us it was at the "school project for credit" phase...
Kinduv curtails the idea of providing input if you would get in trouble for recieving too much of it.
o_O
"Be conservative in what you send; be liberal in what you accept." -- Postel's Law
"tacos" -- Cactus' Law
"t̥͍͎̪̪͗a̴̻̩͈͚ͨc̠o̩̙͈ͫͅs͙͎̙͊ ͔͇̫̜t͎̳̀a̜̞̗ͩc̗͍͚o̲̯̿s̖̣̤̙͌ ̖̜̈ț̰̫͓ạ̪͖̳c̲͎͕̰̯̃̈o͉ͅs̪ͪ ̜̻̖̜͕" -- -̖͚̫̙̓-̺̠͇ͤ̃ ̜̪̜ͯZ͔̗̭̞ͪA̝͈̙͖̩L͉̠̺͓G̙̞̦͖O̳̗͍
Offline
as i code as well in java things i experiment with, i find your code highly interesting (did not do much with javax.media.j3d.* yet) ... especially because i want to see it in action
your project describes itself in a very positive tone ... go on and good luck! if you have some code that does something, let me/us know so that we can test and give feedback on it
evaluation and external meanings that are quoted in a work are no problem (at least i see scientifical work like this), as long as you describe, what you did (e.g. started a evaluation of your experiments in a forum of people interesting in your project - providing the thread, links and other details in your text) ... what i would advice is that you do evaluations only as additional info (do not rely on them in your work) and start them after milestones (e.g. after implementation of important parts)
take also care that evaluations cost time
for now i'm observing your work from time to time
info for other "observers" of this project:
a much better link for reading the code on the web is this:
https://ensmer.dev.java.net/source/brow … ki/ensmer/
The impossible missions are the only ones which succeed.
Offline
Would it be possible to have an ensmer-cvs PKGBUILD similar made as the fluxbox-cvs what does the cvs setup and update.
http://bbs.archlinux.org/viewtopic.php?t=6836
Markku
Offline
not to mention you told us it was at the "school project for credit" phase...
Kinduv curtails the idea of providing input if you would get in trouble for recieving too much of it.
o_O
I wasn't serious in that comment about lost interest, just wanted to add a jocular tone to the post. yikes... :-D
As for the school project portion of the work, my prof is really cool about it (he's done presentations on open source and stuff and uses OpenBSD at home). I think he'd even let me accept minor code contributions, but I personally didn't want to go that far. All comments are welcome. Since this is a bigger than average school project, I'll kind of be depending on feedback.
Of course, discretion is required, if code's not acceptable, pseudocode's not either, but *ideas*... that should be ok, no?
As for the physical design doc, I wrote that for public benefit, its not *required* by the course. Probably it will be worth a few marks though. I think its out of date already, I keep changing my mind. :-D
Dusty
Offline
Would it be possible to have an ensmer-cvs PKGBUILD similar made as the fluxbox-cvs what does the cvs setup and update.
http://bbs.archlinux.org/viewtopic.php?t=6836
I'm planning on making a package for Ensmer, either with the first release (it will be in a week or two), or possibly for CVS as suggested. First I have to make packages for the dependencies. I think the only one I won't be including in the package is Java 3D though. Problem is, I'm using Java 1.5 betas until Java 5 is moved from testing, and my packages are packaged differently from Xentac's, so I won't be making the J3D package until I download that.
Dusty
Offline
Just a quick note (the long note is here: https://ensmer.dev.java.net/servlets/Pr … geID=4530) to say that I made some progress today (not as much as I would have liked. I'm getting distracted/lazy). There's an aweful lot of code there for what you get (an app that allows you to keyNavigator around a ColorCube. I could have done that in ONE METHOD, lol), but the framework is looking solidish.
I've made some decisions for the first milestone release. I should be making that sometime next week (the earlier the better!). Basically it will be what I have here with an InputManager and a really nice navigation system (I tested the navigation system in an earlier prototype. I have to rewrite it from scratch, but I really liked how it worked).
On another note, I wouldn't mind if some artistic folks would contribute some VRML (or X3D?) objects that I can use for demonstrating the interface. I don't know much about the Java 3D loaders yet (more research), so I can't say exactly what they can or can't contain. Probably it would be better if I figure it out first, eh? I'm personally studying wings3D (recommended by cdeserva and in punkrockguy318's TUR), as blender is just too complex for me to figure out.
Dusty
Offline
I've been silent for a while, but keeping busy. I've got to my first milestone (I make up milestones as I go) release in Ensmer. Basically, its just a Navigation module with a pretty lil colorcube in it.
Read about it at http://ensmer.dev.java.net/ Most of the pages have been updated, there's a download page, and documentation describing the very little you can do with Ensmer as it stands. Check out the Discussion Forums to read up on my journal of what I've been doing.
For those that are going to ask, yes I did make an Arch package for Ensmer. Its available (along with its java3d dependency and jedit 4.2 for anybody that might want it in the following repository:)
[dusty]
Server = http://www.archlinux.org/~dusty/packages
If you want to see the PKGBUILDS:
http://www.archlinux.org/~dusty/pkgbuilds/
Please note that the Ensmer package creates a world writable directory in /usr/share. You're probably better off downloading the file from the Ensmer homepage and running it from your home directory. I haven't gotten around to figuring out how to save user settings yet, so it all goes in the same directory as the jar file.
See the documentation on the home page for more information.
I've been preparing this release all day and now its suppertime, so I'm not in the mood for giving a lot of details... you can find it at the home page. :-D
Dusty
Offline
sweet - a release... i'll check it out later on at home...
Offline
I did a test with milestone1-2 as per instructed in documentation. I am using USB wheel mouse.
To see the cube, move your mouse backwards a bit. You will notice that moving the mouse forward or backwards moves the user view (ie: you) backwards and forwards in the scene.
Unstable and often falls back where it started.... it could also be the fault of my mouse. Do we need a good mouse using Ensmer?
Try moving the mouse left and then right. You will notice that it rotates your view to the left or right, on the spot, much as though you were turning your head or body in that direction.
The cube doesn't rotate nor stay at the spot but moves to left or right.
Now scroll the wheel on your mouse forward or back. It should move your view up and down.
This works fine..... moves up or down with rotation.
Markku
Offline
To see the cube, move your mouse backwards a bit. You will notice that moving the mouse forward or backwards moves the user view (ie: you) backwards and forwards in the scene.
Unstable and often falls back where it started.... it could also be the fault of my mouse. Do we need a good mouse using Ensmer?
I'd like it to work with any mouse. I know it would be more useable with a trackball, however. I'll test it with my parent's shitty Microsoft optical mouse and see what happens and get back to you.
Try moving the mouse left and then right. You will notice that it rotates your view to the left or right, on the spot, much as though you were turning your head or body in that direction.
The cube doesn't rotate nor stay at the spot but moves to left or right.
Its teh user view, not the cube that rotates. It may be hard to tell though, without the forward/back motions working properly.
It doesn't act unstable or fall back like the forward/back one does? that would be very odd, as the code is practically identical.
I'll try to test it with a regular mouse tomorrow, I'm in the middle of a mess with the next batch of code right now....
Dusty
Offline
Rasat:
I tested Ensmer with my parent's optical mouse. I can sort of duplicate your problem. This optical mouse has a habit of jumping across the screen every once in a while, so you lose track of the pointer. If that happens in Ensmer, it is interpretted as a huge navigational movement, which may explain what is happening to you.
I was having a difficult time controlling the mouse when I was running it directly on my desk. Its a wood grain desk, and the optical eye seemed to be interpreting the grain as massive mouse movement. I could hardly even control it in two dimensions.
However, if I placed a piece of plain white paper under the mouse, it reduced the problem drastically. It still hops about once in a while, but is more or less useable, at least for testing.
Do you think this is the same problem you are experiencing, or would you suspect it is something else? I'm guessing its related to the hardware, because Java's event processing code is pretty much uniform.... only I had to hack it a bit to get the direct navigation feature. If its not the same problem, can you describe the effects in more detail so I can try to track it down?
I could add code to Ensmer to ignore any movements over a certain threshold; I think this would reduce the problem, but not solve it (and may introduce other difficult bugs). Such a resolution should probably be done by the X Server, not in application code anyway. However, it would give you a chance to test Ensmer for the time being. What do you think?
Has anybody else tried the current version? Do you get similar problem,s or does it work as expected?
Thanks,
Dusty
Offline
Just a quick note, I was happily working on the object loading part of ensmer, testing as I went when suddenly the navigation module blew up and started acting exactly like you describe (this with my trackball). So apparently there's something in my code that needs to be deciphered.
I'll try to look into it tomorrow, but I might not get to it. First I have to get it back to the point where it will run.
Dusty
Offline
Just a quick note, I was happily working on the object loading part of ensmer, testing as I went when suddenly the navigation module blew up and started acting exactly like you describe (this with my trackball). So apparently there's something in my code that needs to be deciphered.
I'll try to look into it tomorrow, but I might not get to it. First I have to get it back to the point where it will run.
Dusty
ha ha... I love that.
the best thing I could say is this: you can never have enough debug output - output everything even if you know it will never be needed...
Offline
Just a note.
This is one of *those* bugs. It won't reproduce now. The good news is I got the object management stuff all working perfectly. That is, there are absolutely no *known* bugs in it becasue I haven't tested it.
I'm thinking of making a release with all the new code, but the actual end result looks identical to milestone 1. So I think I'll add some more exciting, interesting, visual artifacts first.
Dusty
Offline
I think I found and fixed the bug you were describing, rasat. I'm not sure, but the fix seems to have made the whole navigation system a bit edgier (ie: you can see the edges on each individual movement) but that might be my graphics card or config or something.
I think I'm almost ready for another release. I'm going to clean up some code, maybe write some more tests (damn tests), and will be packaging up the next milestone today or tomorrow. It doesn't act much different, but I've got it loading some objects (a table and chair rotated at funny angles) I made in wings3D.
Dusty
Offline
I didn't realize how painless it would be to make a new release; its all ready! Visit http://ensmer.dev.java.net to read up on the changes. Visually, there aren't many, but the changes to code are huge. things are going to get exciting soon.
I haven't updated the Arch package in my repo because I kind of bjorked pacman (totally bjorked it, actually) and I don't feel like going to the trouble of fixing it to make a package. Here's a PKGBUILD though; I'm not certain it works.
pkgname=ensmer
pkgver=milestone2
pkgrel=1
pkgdesc="Ensmer is an experiment in 3D interfaces"
makedepends=(j2sdk apache-ant)
depends=('j2re>=1.5.0' java3d)
pkgurl="https://ensmer.dev.java.net/"
source=(https://ensmer.dev.java.net/files/documents/2312/8263/Ensmer-milestone-2-src.jar)
md5sums=('cee16563eb749fcd9f490f6f39c0b13b')
build() {
cd $startdir/src/
mkdir -p $startdir/pkg/usr/bin
mkdir -p $startdir/pkg/usr/share/ensmer
jar xf Ensmer-milestone-1-src.jar
cd Ensmer-milestone-1-src
ant dist
cp -r dist/* $startdir/pkg/usr/share/ensmer
echo -e "#/bin/shncd /usr/share/ensmernjava -jar Ensmer.jar" > $startdir/pkg/usr/bin/ensmer
chmod a+x $startdir/pkg/usr/bin/ensmer
chmod a+w $startdir/pkg/usr/share/ensmer
}
One note:
If you used the package for both milestone 1 and 2, you had better delete /usr/share/ensmer/PrevalenceBase before running milestone 2 or I have no idea what will happen. :-P I still strongly recommend downloading the jar from http://ensmer.dev.java.net/ and installing it according to the instructions in the documentation.
Visually, the new version contains a table and a chair rotated at an impossible angle that I created in a few seconds in wings 3d. Navigation is the same as before, except I believe I fixed the bug rasat mentioned.
Now, what are you waiting for, go find some bugs!!!!! :-D
Dusty
Offline
Ok....try this one
#Pkgbuild updated by MrGreen
#
pkgname=ensmer
pkgver=milestone2
pkgrel=1
pkgdesc="Ensmer is an experiment in 3D interfaces"
makedepends=(j2sdk apache-ant)
depends=('j2re>=1.5.0' java3d)
pkgurl="https://ensmer.dev.java.net/"
source=(https://ensmer.dev.java.net/files/documents/2312/8506/Ensmer-milestone-2-src.jar)
md5sums=('eb653acf3b3e59f7de1dbf1675c7581c')
build() {
cd $startdir/src/
mkdir -p $startdir/pkg/usr/bin
mkdir -p $startdir/pkg/usr/share/ensmer
jar xf Ensmer-milestone-2-src.jar
cd Ensmer-milestone-2-src
ant dist
cp -r dist/* $startdir/pkg/usr/share/ensmer
echo -e "#/bin/shncd /usr/share/ensmernjava -jar Ensmer.jar" > $startdir/pkg/usr/bin/ensmer
chmod a+x $startdir/pkg/usr/bin/ensmer
chmod a+w $startdir/pkg/usr/share/ensmer
}
Please let me know if there are any problems with this pkgbuild....
Later
Mr Green
Offline
Please let me know if there are any problems with this pkgbuild....
Getting the following error:
[root@localhost ~]# ensmer
Unable to access jarfile Ensmer.jar
[root@localhost ~]#
There is no files in the generated package except usr/bin/ensmer:
usr/
usr/bin/
usr/bin/ensmer
usr/share/
usr/share/ensmer/
Markku
Offline
works fine here
Mr Green
Offline
Mr Green you posted at the same time as I did edit. Most likely you didin't noticed me saying the PKGBUILD package is empty.
I downloaded from the website. Dusty FYI, the mouse movement is smooth and works well. What I cannot get is to rotate the table when moving left or right. The table angle, the view remains same. It creates a curve movement when seeing the same angle of the table.
Markku
Offline
I downloaded from the website.
This is probably the better idea anyway.
Dusty FYI, the mouse movement is smooth and works well.
woot! That's the very first user-submitted bug I've ever fixed my life!!!!!! weeeeeeeheeeeee!
What I cannot get is to rotate the table when moving left or right. The table angle, the view remains same. It creates a curve movement when seeing the same angle of the table.
I'm not certain I understand. When you move the mouse left/right, it should rotate the user view, not the table. It would be like rotating your head from left to right. Is this what's happening? The objects are not moveable yet.... I'm working on that part now.
Dusty
Offline