You can subscribe to this list here.
2008 |
Jan
|
Feb
|
Mar
|
Apr
(4) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(17) |
Nov
(29) |
Dec
(18) |
---|---|---|---|---|---|---|---|---|---|---|---|---|
2009 |
Jan
(3) |
Feb
(2) |
Mar
(15) |
Apr
(7) |
May
(7) |
Jun
|
Jul
(2) |
Aug
(5) |
Sep
(7) |
Oct
(1) |
Nov
(3) |
Dec
|
2010 |
Jan
|
Feb
|
Mar
(3) |
Apr
|
May
(2) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(1) |
Dec
|
From: Kyle M. (qDot) <ky...@no...> - 2009-03-15 01:45:03
|
Hah, I said it'd take a week, and sure enough... Changelog: * Overhauled CMake system because I didn't really know what I was doing with it the first time around (and may still not this time) * Added logging capabilities using Apache's log4cxx library * Removed C code completely, still available in repository (boneyard directory) but no longer distributed in source tarballs * Rearranged src directory to only contain what used to be in cpp directory * Added pregen'd angular matrix LUT header since it currently never changes but takes a bit of time to generate on some systems * Added logging level options to FalconCLIBase * Fixed lots of small bugs in findfalcons_cpp, FalconCLIBase, falcon_test_cli * Fixed compiling in Visual Studio on windows (MinGW still works fine too) There will definitely be a beta 4, since I've found errors on the way through this release that I still need to fix, but it's nice to have some progress again at least. http://www.sourceforge.net/projects/libnifalcon/ |
From: Stephen S. <rad...@gm...> - 2009-03-09 14:27:00
|
Cool, I'll try to do some testing this week. I've been planning to write a CHAI 3D driver for libnifalcon (so I can integrate it into DIMPLE), so I'll be testing on all three operating systems. cheers, Steve On Mon, Mar 9, 2009 at 1:53 AM, Kyle Machulis (qDot) <ky...@no...> wrote: > Every time I plan a release day, without fail, the last thing I do is > think "Oh, wonder if things still compile on windows", and then the > release gets delayed a week. > > So yeah, once again, I seem to have completely hosed windows building. > At some point in February, I also broke libusb support and cmake > generation for any platform with case-sensitive file system. I really > gotta start testing on more than my macbook. :) > > Anyways, lots of fixes this weekend, as you can see in the commit logs > on master. Windows is well on its way to working again (watch the > beta3-win-fix branch if you're interested), both under mingw and visual > studio this time. > > Also, libusb-1.0 seems to be almost completely working on OS X, using > the newest dev commit of libusb. Firmware loading is still iffy at best, > but once firmware is loaded, it works great. Not a huge development > since libftdi already worked fine, but interesting nonetheless > (especially since I didn't really have to put any work into it :) ). > > > > ------------------------------------------------------------------------------ > Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA > -OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise > -Strategies to boost innovation and cut costs with open source participation > -Receive a $600 discount off the registration fee with the source code: SFAD > http://p.sf.net/sfu/XcvMzF8H > _______________________________________________ > Libnifalcon-devel mailing list > Lib...@li... > https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel > |
From: Kyle M. (qDot) <ky...@no...> - 2009-03-09 05:53:07
|
Every time I plan a release day, without fail, the last thing I do is think "Oh, wonder if things still compile on windows", and then the release gets delayed a week. So yeah, once again, I seem to have completely hosed windows building. At some point in February, I also broke libusb support and cmake generation for any platform with case-sensitive file system. I really gotta start testing on more than my macbook. :) Anyways, lots of fixes this weekend, as you can see in the commit logs on master. Windows is well on its way to working again (watch the beta3-win-fix branch if you're interested), both under mingw and visual studio this time. Also, libusb-1.0 seems to be almost completely working on OS X, using the newest dev commit of libusb. Firmware loading is still iffy at best, but once firmware is loaded, it works great. Not a huge development since libftdi already worked fine, but interesting nonetheless (especially since I didn't really have to put any work into it :) ). |
From: Jonas F. <jo...@fo...> - 2009-02-17 20:01:28
|
Good to see it alive. I have also a bit much of work overload (and btw, struggling with proprietary drivers of the Sensable Omni haptic device - one day we should free it to...). I did try once more with libnifalcon and H3D and it did not really work the whole way (H3D did not seem to find it as AnyDevice or NiFalconDevice) but I think that is more related to configuration/making of H3D, not HAPI or libnifalcon. Perhaps Daniel (cc:ed) has suggestion. HAPI worked with spring effect etc. I am looking forward to see what mechanical ppl can do to make the forces axis equal and that about center. By the way a bunch of game programmers and me are working on a haptic enabled snowball throwing game to be beta released (as part of a game design competition) in 2-3 months. Our plan is to have multi-OS support, and libnifalcon would fit perfect to that plan. I currently works with Windows/falcon api (through HAPI). Best regards Jonas 2009/2/14 Kyle Machulis (qDot) <ky...@no...> > So, after two months of near dead quiet due to work eating all my free > time, I'm hoping to get back to working on libnifalcon this weekend. > Here's a few things of note in and around the library: > > - libftdi now has a mailing list and repository access - > http://www.intra2net.com/en/developer/libftdi/ > - libusb 1.0 now has a patch to work on OS X. I've been playing with it > a little bit, but I run into the same problems I had when I tried > writing my own pure carbon drivers. I end up getting lockups in the low > level read function for no discernible reason. But, if you want to give > it a shot. http://www.cs.unm.edu/~hjelmn/libusb_1.0_darwin.patch<http://www.cs.unm.edu/%7Ehjelmn/libusb_1.0_darwin.patch> > - I'm talking to a few mechanical engineers about tightening up the > dynamics a little bit, mainly fixing the rigidity issues with x/y axis > forces. > - Gonna be doing a ton of cmake cleanup. Been doing a lot of work with > cmake build systems lately, learned a lot, especially about things I'm > doing completely wrong currently :) > > I'd like to get beta 3 out pretty quick just to let everyone know the > project is still alive. :) > > Kyle > > > ------------------------------------------------------------------------------ > Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, > CA > -OSBC tackles the biggest issue in open source: Open Sourcing the > Enterprise > -Strategies to boost innovation and cut costs with open source > participation > -Receive a $600 discount off the registration fee with the source code: > SFAD > http://p.sf.net/sfu/XcvMzF8H > _______________________________________________ > Libnifalcon-devel mailing list > Lib...@li... > https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel > |
From: Kyle M. (qDot) <ky...@no...> - 2009-02-14 18:43:53
|
So, after two months of near dead quiet due to work eating all my free time, I'm hoping to get back to working on libnifalcon this weekend. Here's a few things of note in and around the library: - libftdi now has a mailing list and repository access - http://www.intra2net.com/en/developer/libftdi/ - libusb 1.0 now has a patch to work on OS X. I've been playing with it a little bit, but I run into the same problems I had when I tried writing my own pure carbon drivers. I end up getting lockups in the low level read function for no discernible reason. But, if you want to give it a shot. http://www.cs.unm.edu/~hjelmn/libusb_1.0_darwin.patch - I'm talking to a few mechanical engineers about tightening up the dynamics a little bit, mainly fixing the rigidity issues with x/y axis forces. - Gonna be doing a ton of cmake cleanup. Been doing a lot of work with cmake build systems lately, learned a lot, especially about things I'm doing completely wrong currently :) I'd like to get beta 3 out pretty quick just to let everyone know the project is still alive. :) Kyle |
From: Kyle M. (qDot) <ky...@no...> - 2009-01-12 02:53:47
|
Jonas Forsslund wrote: > Happy continuation of new year! > > I did some hacking yesterday with HAPI using libnifalcon-1.0-beta2 and > kylies hapi object from 8th of December (+libusb1.0). I am happy to be > able run a hapi examples! So, just curious, how well does this bind over the H3D then? I ended up trying to build an H3D device by copying the one for the novint SDK, but never really had much luck getting anything interesting to happen... > * The units: when I tried I got the feeling that 3 (Newton?) in Z > direction is much stronger than 3 in X and Y? I'm starting to wonder if our model needs to reflect this specifically, since you've usually got similarly signed forces across all actuators when dealing with the z axis, but that's not the case with x or y. Just kinda guessing here though. > * I think it has been discuessed before, that z=0 should be in middle > of the z axis, and i would like that too, to better work out of box > with examples etc. Yeah, I just need to figure out exactly where that is. I'm still not quite understanding the center offset in the kinematics model in relation to the falcon workspace itself. I've been back in touch with Kevin, I'll poke him about it. > * It seems to start working only about every second time, those times > it has outputed message about device need to be homed. Hmm, ok, I think I remember seeing that bug too, I'll check it out again soon. Kyle |
From: Jonas F. <jo...@fo...> - 2009-01-11 18:11:09
|
Happy continuation of new year! I did some hacking yesterday with HAPI using libnifalcon-1.0-beta2 and kylies hapi object from 8th of December (+libusb1.0). I am happy to be able run a hapi examples! My comments so far: * The units: when I tried I got the feeling that 3 (Newton?) in Z direction is much stronger than 3 in X and Y? * I think it has been discuessed before, that z=0 should be in middle of the z axis, and i would like that too, to better work out of box with examples etc. * It seems to start working only about every second time, those times it has outputed message about device need to be homed. my junk package of it all is at (not intended as release at all, I just use it for my personal testing, but anyone welcome): https://code.launchpad.net/~jonas-forsslundsystems/+junk/hapifalcon<https://code.launchpad.net/%7Ejonas-forsslundsystems/+junk/hapifalcon> Best regards Jonas Forsslund 2009/1/1 Kyle Machulis (qDot) <ky...@no...> > Well, what was going to be two vacation weeks of hardcore coding to push > out libnifalcon v1 before the end of the year turned into two weeks of > hardly being on my computer at all, which was quite refreshing and > probably better for the project. :) > > I did spent a little bit of time here and there implementing log4cxx > into libnifalcon as a nice, compile-time switchable logging system. This > way I no longer need to manage my debug prints on release, and it's > almost stupidly powerful (You can now watch your falcon logs over telnet > or web services if you so please). I have a FalconLogger.h include that > is pretty much a straight copy of a really useful header file I found, > that allows for compile time redefinition of the macros, so I can > completely remove any calls to the library in release versions. > > That's about all that's happened with the library so far. I haven't > really heard any bug reports or anything, so I'm hoping things are > relatively bug-free. Either that, or everyone has taken the same break I > did. :) > > Happy New Year! > > Kyle > > > ------------------------------------------------------------------------------ > _______________________________________________ > Libnifalcon-devel mailing list > Lib...@li... > https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel > |
From: Kyle M. (qDot) <ky...@no...> - 2009-01-01 00:51:15
|
Well, what was going to be two vacation weeks of hardcore coding to push out libnifalcon v1 before the end of the year turned into two weeks of hardly being on my computer at all, which was quite refreshing and probably better for the project. :) I did spent a little bit of time here and there implementing log4cxx into libnifalcon as a nice, compile-time switchable logging system. This way I no longer need to manage my debug prints on release, and it's almost stupidly powerful (You can now watch your falcon logs over telnet or web services if you so please). I have a FalconLogger.h include that is pretty much a straight copy of a really useful header file I found, that allows for compile time redefinition of the macros, so I can completely remove any calls to the library in release versions. That's about all that's happened with the library so far. I haven't really heard any bug reports or anything, so I'm hoping things are relatively bug-free. Either that, or everyone has taken the same break I did. :) Happy New Year! Kyle |
From: Kyle M. (qDot) <ky...@no...> - 2008-12-19 06:45:23
|
For anyone not watching the project over on sourceforge, libusb-1.0 stable for linux was released on Dec 13th. http://sourceforge.net/project/showfiles.php?group_id=1674&package_id=277626 Now all we need is an OS X core for it and I'll be really happy. :) |
From: Liam K. <qua...@go...> - 2008-12-16 13:44:32
|
Sorry follow up on the HAPI stuff, but i had to priorities writing my msc thesis and had to put playing with libnifalcon on hold for a while. It'll all done now and hopefully i'll get time to play with the falcon over the holidays. I'll try to write a JNI for libnifalcon soon. Liam On Sun, Dec 14, 2008 at 10:16 PM, Kyle Machulis (qDot) < ky...@no...> wrote: > Just to let everyone know, I'm still alive, albeit barely. Crunch time > at the day job means I haven't even gotten to connect my falcon to > anything, much less work on development. I doubt there'll be anything > new released from my end until after next weekend. If you've sent me > email, don't worry, I've read it and I'll get to it at some point. :) > > Anyone had a chance to play with the HAPI stuff yet? > > Kyle > > > ------------------------------------------------------------------------------ > SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. > The future of the web can't happen without you. Join us at MIX09 to help > pave the way to the Next Web now. Learn more and register at > > http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ > _______________________________________________ > Libnifalcon-devel mailing list > Lib...@li... > https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel > |
From: Stephen S. <rad...@gm...> - 2008-12-16 04:09:31
|
Hi, Here's a video of my project from this semester. The last portion of this video shows it working with the Falcon. I used libnifalcon for this part. Thanks esp. to Kyle, and of course to everyone who has contributed! Being able to run this on my Linux-only laptop when away from the lab contributed significantly to this success.. :) http://www.youtube.com/watch?v=0FpszHuxQbc Steve |
From: Kyle M. (qDot) <ky...@no...> - 2008-12-14 22:43:46
|
Just to let everyone know, I'm still alive, albeit barely. Crunch time at the day job means I haven't even gotten to connect my falcon to anything, much less work on development. I doubt there'll be anything new released from my end until after next weekend. If you've sent me email, don't worry, I've read it and I'll get to it at some point. :) Anyone had a chance to play with the HAPI stuff yet? Kyle |
From: Stephen S. <rad...@gm...> - 2008-12-11 04:19:18
|
Hi, Today I finally had some time to do a few measurements like I was talking about. It's a bit silly though, I haven't taken the time to update my sources lately, so this is unfortunately done on an older version, before the latest changes to the kinematics. But anyways it might still help. This is all a bit rough: I used a good electronic scale, but the numbers did tend to jump around quite a bit. In particular on each trial there was an initial high number, and then it would settle, and then it seemed to slowly fall. I didn't have the scale connected to a computer or anything, so I eyeballed the numbers and tried to write something that seemed about average. I tried to write down the number after the initial peak, but before it started to decline. Anyways, the methodology was as follows, (and this could probably be done better): I put the scale vertically up against the wall. I put the falcon in front of it, with the handle pushed up against the falcon as far as it would go. The handle was touching the scale. I stacked a bunch of heavy books behind the falcon so it would move backwards. I hit the 'zero' button on the scale. I wrote a little program that simply called setForce() with a desired value in the Z component, e.g.: setForce(0,0,10); I then called this program with various values of Z. The results were measured in grams. Here they are, the first column is Z, second column is grams: 10, 90 20, 350 30, 600 40, 880 50, 1100 60, 1320 70, 1420 80, 1420 90, 1490 100, 1350 Attached is a plot. I'm impressed, I really didn't expect the force to get all the way up to 1.4 kg! Not bad. :) As you can see, it's fairly linear up to a Z value of 60, which is a good thing. I hope that these results will have some proportional relationship with the new kinematics. Hopefully I get some time to do the tests again, though my schedule's pretty full until next week. cheers, Steve |
From: Kyle M. (qDot) <ky...@no...> - 2008-12-08 06:33:12
|
Here's my edited version of the HAPI object. The archive is laid out in the HAPI directory format, so if you unzip it in your HAPI checkout, it'll put things in the correct place (but will overwrite your build file, so be careful) It's built to work with libnif v1b2, so make sure you've updated. Note that I've also made the absolute minimum changes (and even then they may be wrong) to make the CMakeFile work, and only for OS X. You may need to change the comm library if you're on linux or windows. The workspace center seems like it might be a little far out on the z-axis too. I'm gonna play with this a little and see what I can come up with. I'll keep working on this though, definitely excited about it. :) |
From: Kyle M. (qDot) <ky...@no...> - 2008-12-08 05:48:57
|
Here's the changelog notes: --- * Release v1 beta 2 * Changed units in stamper kinematic core position readings from millimeters to meters * Changed units in stamper kinematic core force values to newtons (now matches HDAL) * Changed units in FalconGeometry to meters * Updated falcon_test_cli to reflect unit changes * Added many more comments, though still nothing in depth * Removed falcon_mouse demo (many people have difficulty building and it doesn't work right, moved to own repository, http://www.github.com/qdot/libnifalcon_mouse) * Fixed some const/reference passing arguments * Changed comm libraries to use pointers and prototypes to remove necessity for comm includes in lib headers * Fixed FTD2XX compatibility (comm driver hadn't been working in the past few releases) * Fixed compilation and compatibility for windows (now using mingw as test compiler) --- So yeah, if you're programming for the falcon right now, note that the units are different. By a orders of magnitude. Thank god we don't have a warranty, this is the first software I've produced where I can safely say going between version may cause your hardware to punch you. We're moving toward having less and less obvious hangups for v1, which is good. I'm really hoping the changes I made to the kinematics this week do the trick on that end. If so, we're down to commenting and small tweaks to the code, and beta 3 may just end up being v1. I'm dead set on having v1 out before the end of the year. HAPI source files to follow this pretty quick. I'll probably just post them here to the list. |
From: Kyle M. (qDot) <ky...@no...> - 2008-12-08 00:35:22
|
Ok, after finally reading up some and getting windows running (the falcon works perfectly on a windows VM on parallels on OS X, funny enough), I started comparing numbers between the HDAL kinematics output and libnifalcon. Turns out, I just needed to add an internal gain function, and it made things look almost exactly the same between the two APIs .This is all hand measured, but it certainly seems "good enough" for the moment. So, now there's a magic scalar in FalconKinematicStamper, but it does the job. I've tested the new code with the fixes I made for HAPI, and the HAPI demos feel great now. I'm gonna do a little more code cleanup on libnifalcon, merge unit-change to master, and call that beta 2. Hoping to have it out this evening, assuming Fallout 3 will stop taking larger chunks of time than I mean for it to. :) Kyle |
From: Jonas F. <jo...@fo...> - 2008-12-06 00:37:25
|
I'm looking forward to test new versions when they are available. About HAPI and H3D, if it works in HAPI it's as easy as put <NifalconDevice /> in an x3d file. HAPI can be considered a layer for H3D. So for implementation all we need to focus on is HAPI. 2008/12/6 Kyle Machulis (qDot) <ky...@no...> > Yeah, I have a feeling this either has to do with using mm as our units, or > the workspace center that's not origin. I'm gonna work on all of this stuff > this weekend, and will hopefully include the updated, working version of the > libnifalcon HAPI driver along with all the other changes I've made once it's > done (haven't had a lot of time to concentrate on it this week > unfortuantely). > > On a similar note, what does it take to turn an HAPI device into one that > will work with H3D? I haven't actually read all the H3D documentation yet, > so if the answer is in there, just tell me. :) > > > Jonas Forsslund wrote: > >> Hi, >> >> I (and Daniel of Sensegraphics) tried libnifalcon-1.0-beta-1.tar.gz with >> libusb-0.9.4.tar.bz2 and HAPI revision 440 (H3DUtil rev 112), and the >> falcon hapi class that has been circulating 10 days ago. >> >> We mostly got to the point when we run it we got error 402 (or similar im >> sorry >> did not remeber the code, or can provide any better report, I was in a >> hurry >> packing up for a train and have not had the chance to try again since). We >> got >> falcon to react and give positions but no correct forces. The example we >> compiled and >> run was HAPI/examples/SpringExample/SpringExample.cpp with spring position >> at 0.02, 0.02, 0.02 and force factor 200 (or similar, but these numbers >> works >> well in Windows with the proprietary falcon drivers) >> >> Its probably deprecated now but if anyone want to pick up where we left I >> put >> the HAPI revision on bazaar with falcon node made by Daniel. (need bzr) >> >> bzr branch lp:~jonas-forsslundsystems/+junk/hapifalcon >> >> /Jonas >> >> >> 2008/11/30 Kyle Machulis (qDot) <ky...@no... <mailto: >> ky...@no...>> >> >> >> Here's what I posted to the H3D Boards concerning the libnifalcon >> layer >> they have available: >> >> --- >> >> Ok, got the falcon class up and running using HAPIDemo, though it >> took a >> good bit of work. >> >> Our only kinematics implementation is a little /too/ literal at the >> moment, and it uses the FF origin to EE origin distance as the >> workspace >> center (which is something like [0,0,.15m] instead of [0,0,0]). >> Also, it >> uses millimeters instead of meters for distance units, which means the >> scaling is completely off. After fiddling with the numbers a bit, >> I got >> everything under control and feeling pretty good (outside of some >> of the >> extremities, which we've been having problems with anyways). >> >> So, this does bring up a lot of issues with my documentation (or lack >> thereof) and selection of units in the library, but that's why we're >> still in beta. :) I'll start getting some of this stuff fixed, and >> I'll >> keep using the HAPI examples while testing all of this. Any advice you >> might have would also be helpful, since I'm quite new to doing haptics >> work, and may be missing some obvious things. :) >> >> --- >> >> For anyone interested in following, the thread is at >> http://www.h3dapi.org/modules/newbb/viewtopic.php?topic_id=864&forum=3 >> < >> http://www.h3dapi.org/modules/newbb/viewtopic.php?topic_id=864&forum=3> >> >> So yeah. I've been meaning to fix that "Center isn't center" stuff >> forever, figured I'd do it the same way novint did (workspace >> transform >> function), but just never got around to it. As for units, changing >> everything to meters shouldn't be too difficult, though I'm not >> sure how >> many people have already tuned their software to use those units. >> >> >> ------------------------------------------------------------------------- >> This SF.Net email is sponsored by the Moblin Your Move Developer's >> challenge >> Build the coolest Linux based applications with Moblin SDK & win >> great prizes >> Grand prize is a trip for two to an Open Source event anywhere in >> the world >> http://moblin-contest.org/redirect.php?banner_id=100&url=/ >> <http://moblin-contest.org/redirect.php?banner_id=100&url=/> >> _______________________________________________ >> Libnifalcon-devel mailing list >> Lib...@li... >> <mailto:Lib...@li...> >> https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel >> >> >> ------------------------------------------------------------------------ >> >> ------------------------------------------------------------------------- >> This SF.Net email is sponsored by the Moblin Your Move Developer's >> challenge >> Build the coolest Linux based applications with Moblin SDK & win great >> prizes >> Grand prize is a trip for two to an Open Source event anywhere in the >> world >> http://moblin-contest.org/redirect.php?banner_id=100&url=/ >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Libnifalcon-devel mailing list >> Lib...@li... >> https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel >> >> > > |
From: Kyle M. (qDot) <ky...@no...> - 2008-12-05 23:50:42
|
Yeah, I have a feeling this either has to do with using mm as our units, or the workspace center that's not origin. I'm gonna work on all of this stuff this weekend, and will hopefully include the updated, working version of the libnifalcon HAPI driver along with all the other changes I've made once it's done (haven't had a lot of time to concentrate on it this week unfortuantely). On a similar note, what does it take to turn an HAPI device into one that will work with H3D? I haven't actually read all the H3D documentation yet, so if the answer is in there, just tell me. :) Jonas Forsslund wrote: > Hi, > > I (and Daniel of Sensegraphics) tried libnifalcon-1.0-beta-1.tar.gz with > libusb-0.9.4.tar.bz2 and HAPI revision 440 (H3DUtil rev 112), and the > falcon hapi class that has been circulating 10 days ago. > > We mostly got to the point when we run it we got error 402 (or similar > im sorry > did not remeber the code, or can provide any better report, I was in a > hurry > packing up for a train and have not had the chance to try again > since). We got > falcon to react and give positions but no correct forces. The example > we compiled and > run was HAPI/examples/SpringExample/SpringExample.cpp with spring > position > at 0.02, 0.02, 0.02 and force factor 200 (or similar, but these > numbers works > well in Windows with the proprietary falcon drivers) > > Its probably deprecated now but if anyone want to pick up where we > left I put > the HAPI revision on bazaar with falcon node made by Daniel. (need bzr) > > bzr branch lp:~jonas-forsslundsystems/+junk/hapifalcon > > /Jonas > > > 2008/11/30 Kyle Machulis (qDot) <ky...@no... > <mailto:ky...@no...>> > > Here's what I posted to the H3D Boards concerning the libnifalcon > layer > they have available: > > --- > > Ok, got the falcon class up and running using HAPIDemo, though it > took a > good bit of work. > > Our only kinematics implementation is a little /too/ literal at the > moment, and it uses the FF origin to EE origin distance as the > workspace > center (which is something like [0,0,.15m] instead of [0,0,0]). > Also, it > uses millimeters instead of meters for distance units, which means the > scaling is completely off. After fiddling with the numbers a bit, > I got > everything under control and feeling pretty good (outside of some > of the > extremities, which we've been having problems with anyways). > > So, this does bring up a lot of issues with my documentation (or lack > thereof) and selection of units in the library, but that's why we're > still in beta. :) I'll start getting some of this stuff fixed, and > I'll > keep using the HAPI examples while testing all of this. Any advice you > might have would also be helpful, since I'm quite new to doing haptics > work, and may be missing some obvious things. :) > > --- > > For anyone interested in following, the thread is at > http://www.h3dapi.org/modules/newbb/viewtopic.php?topic_id=864&forum=3 > <http://www.h3dapi.org/modules/newbb/viewtopic.php?topic_id=864&forum=3> > > So yeah. I've been meaning to fix that "Center isn't center" stuff > forever, figured I'd do it the same way novint did (workspace > transform > function), but just never got around to it. As for units, changing > everything to meters shouldn't be too difficult, though I'm not > sure how > many people have already tuned their software to use those units. > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's > challenge > Build the coolest Linux based applications with Moblin SDK & win > great prizes > Grand prize is a trip for two to an Open Source event anywhere in > the world > http://moblin-contest.org/redirect.php?banner_id=100&url=/ > <http://moblin-contest.org/redirect.php?banner_id=100&url=/> > _______________________________________________ > Libnifalcon-devel mailing list > Lib...@li... > <mailto:Lib...@li...> > https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel > > > ------------------------------------------------------------------------ > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's challenge > Build the coolest Linux based applications with Moblin SDK & win great prizes > Grand prize is a trip for two to an Open Source event anywhere in the world > http://moblin-contest.org/redirect.php?banner_id=100&url=/ > ------------------------------------------------------------------------ > > _______________________________________________ > Libnifalcon-devel mailing list > Lib...@li... > https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel > |
From: Kyle M. (qDot) <ky...@no...> - 2008-12-05 23:32:40
|
Charlie Matlack wrote: > Sorry for being out of the loop here, but does this mean that the > Falcon hardware is running feedback loops on the torque output from > each motor, or is an assumption being made about torque output being > proportional to motor current/voltage? Are there current sense > resistors on the motors? > Yeah, this is actually where I'm sort of running into problems. I'm not sure /what/ units are being sent to the falcon to apply forces at the moment, so I can't really answer this question properly. Time to get myself a force gauge. :) To do this in the non-mechanical way, I should be able to build a mapping between what novint's SDK puts out to the falcon on calls to their force setting functions by tapping the USB bus. However, I've had horrible luck with any windows machine a touch lately, so this has been much harder that it really should be. |
From: Charlie M. <ma...@gm...> - 2008-12-04 18:24:16
|
Sorry for being out of the loop here, but does this mean that the Falcon hardware is running feedback loops on the torque output from each motor, or is an assumption being made about torque output being proportional to motor current/voltage? Are there current sense resistors on the motors? Charlie On Dec 3, 2008, at 7:22 PM, Kyle Machulis (qDot) wrote: > Newtons is the plan at the moment. > > Stephen Sinclair wrote: >> On Wed, Dec 3, 2008 at 4:16 PM, Jonas Forsslund >> <jo...@fo...> wrote: >> >>> Btw, >>> +1 for meters, I like to stick to SI units if possible >>> >> >> >> On that note, it'd be best if force could be specified in Newtons or >> grams. Not sure which is better actually. I think Newtons is more >> usual, whereas grams are for weight. >> >> Steve >> >> ------------------------------------------------------------------------- >> This SF.Net email is sponsored by the Moblin Your Move Developer's >> challenge >> Build the coolest Linux based applications with Moblin SDK & win >> great prizes >> Grand prize is a trip for two to an Open Source event anywhere in >> the world >> http://moblin-contest.org/redirect.php?banner_id=100&url=/ >> _______________________________________________ >> Libnifalcon-devel mailing list >> Lib...@li... >> https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel >> > > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's > challenge > Build the coolest Linux based applications with Moblin SDK & win > great prizes > Grand prize is a trip for two to an Open Source event anywhere in > the world > http://moblin-contest.org/redirect.php?banner_id=100&url=/ > _______________________________________________ > Libnifalcon-devel mailing list > Lib...@li... > https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel |
From: Kyle M. (qDot) <ky...@no...> - 2008-12-04 03:22:32
|
Newtons is the plan at the moment. Stephen Sinclair wrote: > On Wed, Dec 3, 2008 at 4:16 PM, Jonas Forsslund > <jo...@fo...> wrote: > >> Btw, >> +1 for meters, I like to stick to SI units if possible >> > > > On that note, it'd be best if force could be specified in Newtons or > grams. Not sure which is better actually. I think Newtons is more > usual, whereas grams are for weight. > > Steve > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's challenge > Build the coolest Linux based applications with Moblin SDK & win great prizes > Grand prize is a trip for two to an Open Source event anywhere in the world > http://moblin-contest.org/redirect.php?banner_id=100&url=/ > _______________________________________________ > Libnifalcon-devel mailing list > Lib...@li... > https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel > |
From: Stephen S. <rad...@gm...> - 2008-12-04 03:03:20
|
On Wed, Dec 3, 2008 at 4:16 PM, Jonas Forsslund <jo...@fo...> wrote: > Btw, > +1 for meters, I like to stick to SI units if possible On that note, it'd be best if force could be specified in Newtons or grams. Not sure which is better actually. I think Newtons is more usual, whereas grams are for weight. Steve |
From: Jonas F. <jo...@fo...> - 2008-12-03 21:17:03
|
Btw, +1 for meters, I like to stick to SI units if possible 2008/12/3 Jonas Forsslund <jo...@fo...> > Hi, > > I (and Daniel of Sensegraphics) tried libnifalcon-1.0-beta-1.tar.gz with > libusb-0.9.4.tar.bz2 and HAPI revision 440 (H3DUtil rev 112), and the > falcon hapi class that has been circulating 10 days ago. > > We mostly got to the point when we run it we got error 402 (or similar im > sorry > did not remeber the code, or can provide any better report, I was in a > hurry > packing up for a train and have not had the chance to try again since). We > got > falcon to react and give positions but no correct forces. The example we > compiled and > run was HAPI/examples/SpringExample/SpringExample.cpp with spring position > at 0.02, 0.02, 0.02 and force factor 200 (or similar, but these numbers > works > well in Windows with the proprietary falcon drivers) > > Its probably deprecated now but if anyone want to pick up where we left I > put > the HAPI revision on bazaar with falcon node made by Daniel. (need bzr) > > bzr branch lp:~jonas-forsslundsystems/+junk/hapifalcon > > /Jonas > > > 2008/11/30 Kyle Machulis (qDot) <ky...@no...> > > Here's what I posted to the H3D Boards concerning the libnifalcon layer >> they have available: >> >> --- >> >> Ok, got the falcon class up and running using HAPIDemo, though it took a >> good bit of work. >> >> Our only kinematics implementation is a little /too/ literal at the >> moment, and it uses the FF origin to EE origin distance as the workspace >> center (which is something like [0,0,.15m] instead of [0,0,0]). Also, it >> uses millimeters instead of meters for distance units, which means the >> scaling is completely off. After fiddling with the numbers a bit, I got >> everything under control and feeling pretty good (outside of some of the >> extremities, which we've been having problems with anyways). >> >> So, this does bring up a lot of issues with my documentation (or lack >> thereof) and selection of units in the library, but that's why we're >> still in beta. :) I'll start getting some of this stuff fixed, and I'll >> keep using the HAPI examples while testing all of this. Any advice you >> might have would also be helpful, since I'm quite new to doing haptics >> work, and may be missing some obvious things. :) >> >> --- >> >> For anyone interested in following, the thread is at >> http://www.h3dapi.org/modules/newbb/viewtopic.php?topic_id=864&forum=3 >> >> So yeah. I've been meaning to fix that "Center isn't center" stuff >> forever, figured I'd do it the same way novint did (workspace transform >> function), but just never got around to it. As for units, changing >> everything to meters shouldn't be too difficult, though I'm not sure how >> many people have already tuned their software to use those units. >> >> ------------------------------------------------------------------------- >> This SF.Net email is sponsored by the Moblin Your Move Developer's >> challenge >> Build the coolest Linux based applications with Moblin SDK & win great >> prizes >> Grand prize is a trip for two to an Open Source event anywhere in the >> world >> http://moblin-contest.org/redirect.php?banner_id=100&url=/ >> _______________________________________________ >> Libnifalcon-devel mailing list >> Lib...@li... >> https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel >> > > |
From: Jonas F. <jo...@fo...> - 2008-12-03 21:15:16
|
Hi, I (and Daniel of Sensegraphics) tried libnifalcon-1.0-beta-1.tar.gz with libusb-0.9.4.tar.bz2 and HAPI revision 440 (H3DUtil rev 112), and the falcon hapi class that has been circulating 10 days ago. We mostly got to the point when we run it we got error 402 (or similar im sorry did not remeber the code, or can provide any better report, I was in a hurry packing up for a train and have not had the chance to try again since). We got falcon to react and give positions but no correct forces. The example we compiled and run was HAPI/examples/SpringExample/SpringExample.cpp with spring position at 0.02, 0.02, 0.02 and force factor 200 (or similar, but these numbers works well in Windows with the proprietary falcon drivers) Its probably deprecated now but if anyone want to pick up where we left I put the HAPI revision on bazaar with falcon node made by Daniel. (need bzr) bzr branch lp:~jonas-forsslundsystems/+junk/hapifalcon /Jonas 2008/11/30 Kyle Machulis (qDot) <ky...@no...> > Here's what I posted to the H3D Boards concerning the libnifalcon layer > they have available: > > --- > > Ok, got the falcon class up and running using HAPIDemo, though it took a > good bit of work. > > Our only kinematics implementation is a little /too/ literal at the > moment, and it uses the FF origin to EE origin distance as the workspace > center (which is something like [0,0,.15m] instead of [0,0,0]). Also, it > uses millimeters instead of meters for distance units, which means the > scaling is completely off. After fiddling with the numbers a bit, I got > everything under control and feeling pretty good (outside of some of the > extremities, which we've been having problems with anyways). > > So, this does bring up a lot of issues with my documentation (or lack > thereof) and selection of units in the library, but that's why we're > still in beta. :) I'll start getting some of this stuff fixed, and I'll > keep using the HAPI examples while testing all of this. Any advice you > might have would also be helpful, since I'm quite new to doing haptics > work, and may be missing some obvious things. :) > > --- > > For anyone interested in following, the thread is at > http://www.h3dapi.org/modules/newbb/viewtopic.php?topic_id=864&forum=3 > > So yeah. I've been meaning to fix that "Center isn't center" stuff > forever, figured I'd do it the same way novint did (workspace transform > function), but just never got around to it. As for units, changing > everything to meters shouldn't be too difficult, though I'm not sure how > many people have already tuned their software to use those units. > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's > challenge > Build the coolest Linux based applications with Moblin SDK & win great > prizes > Grand prize is a trip for two to an Open Source event anywhere in the world > http://moblin-contest.org/redirect.php?banner_id=100&url=/ > _______________________________________________ > Libnifalcon-devel mailing list > Lib...@li... > https://lists.sourceforge.net/lists/listinfo/libnifalcon-devel > |
From: Stephen S. <rad...@gm...> - 2008-12-03 16:47:03
|
Hm, we should have a scale around the lab somewhere. If I can find one, I'll try measuring the force output with various force settings. Might be a little difficult to figure out how to set that up exactly, but I'll see if I can do it. Steve On Tue, Dec 2, 2008 at 11:07 PM, Kyle Machulis (qDot) <ky...@no...> wrote: > Stephen Sinclair wrote: >> >> I think fixing the centering is probably a good idea. As for units, >> meters is usually a good choice. From my perspective it doesn't >> really matter all that much, as long as it stays consistent. So the >> thing to do is make a decision on it and change it _now_, since the >> library isn't too widely used yet. It's not something you're going to >> want to deal with in the future. >> > > Ok, meters it is, then. This changeover is happening in the unit-change > branch of the git repository. So far, I've changed the units and the > cube_test example in falcon_test_cli, though now the question of forces > comes up. I'm honestly not sure what units we'd been sending the kinematics > setForces function so far, as I'd never really done any tests with it > outside of simple proportional loops. Now is probably a good time for me to > sit down, document, and actually understand whats happening in those > classes. I'll probably also go ahead and implement a pregenerated header > based lookup table to save us on the kinematics initialization time (on > platforms where -ffast-math doesn't make that trivial). > > Kyle > |