Hello Ken, --enable-64bit doesn't deal with GRASS 64bitness in direct way. What You should had looking for was compiling 32bit apps on 64bit OS.
Regarding float overflow - it makes sense. Can You attach a patch to trac ticket? I (or somebody else) will take a look at it and commit to SVN. Thanks for spotting this. Moving conversation to -dev. Maris. 2010/6/16 Ken Kwasnicki <ken.kwasni...@gmail.com>: > Yes, I my slackware can run 32bit apps. I tried recompiling grass without > the --enable-64bit option but still had the same nviz problem , although i > expect i'm still compiling to 64bit. > > However, I've looked at the togl_flythrough.c code and found the problem. > Or at least found a change the fixes this on my system. > > I had added some debug output to mouse_valuator just to show the thisTime > value, initially I just wanted to see if I would even see the output from a > printf statement when running nviz. What I noticed was that the "seconds" > value for thisTime was not incrementing, only the decimal part was changing, > which I guess is why the view would return to its original position after a > 1 second lapse. > > in the function "this_time" the time values are being cast to float which is > maxing out so i changed them to double which is the return value anyway. > so, i changed this line [320]: > return ((float)tv.tv_sec + ((float)tv.tv_usec / 1000000.0)); > to: > return ((double)tv.tv_sec + ((double)tv.tv_usec / 1000000.0)); > > and that appears to have fixed the problem. the same change may need to be > made for the MINGW32 version of the code, but I left that alone. > > i've made this change in my download of the 6.4.0RC6 version of the code but > for now I'll leave it up to you, or others to make this change as I'm > currently not set up to submit code. > > thanks! > ken > > > On 06/15/2010 12:53 PM, Maris Nartiss wrote: > > Hello, > at first - You can compile 32bit version on 64bit machine (if it's > possible on Slackware). > > Second - once I took a look at code. It was not good, still I didn't > managed to find the root cause. > Fly mode is TCL/C hybrid. > TCL part is here: > http://trac.osgeo.org/grass/browser/grass/branches/develbranch_6/visualization/nviz/scripts > (flythrough.tcl) > Some support is located in nviz/src/ (togl_flythrough.c) > Main rendering is done by OGSF lib/ogsf > > OGSF should be fine, problem is somewhere in nviz part. Probably in > C/TCL interaction. > > Good luck. > If You need some help with coding, feel free to as at grass-dev ML. We > are short on resources and thys any help is wecomed. > > Maris. > > > 2010/6/15, Ken Kwasnicki <ken.kwasni...@gmail.com>: > > > Hi Maris, > Thanks for the reply! Too bad. I guess I'll have to install on a 32bit > system. For what it's worth my system is Slackware 13.0 64bit and also > Nvidia driver, so nothing really new to add to the bug report other than the > variance in linux distribution. > > Any speculation where this bug might occur in the nviz code? Just in case > I'm feeling ambitious enough to take a crack at debugging it. > > Thanks, > Ken > > > On 06/15/2010 12:14 AM, Maris Nartiss wrote: > > > Known issue, currently no fix. > > http://trac.osgeo.org/grass/ticket/46 > > Maris. > > > 2010/6/14, Kwas <ken.kwasni...@gmail.com>: > > > > Helo, > I've used NVIZ in the past and not had this problem but after rebuilding > Grass/Nviz for a new system (Slackware 13 64bit) I now find when I try > to > "fly" to move the landscape it keeps resetting back to its originall > position after a few frames. ie: it will start to move in the > direction > of my mouse motion but then keeps jumping back so that I am unable to > zoom > in or position the view where I want. > > Has anybody seen this problem before or know what I should do to fix it? > Appreciate any assistance. > Thanks, > Ken > -- > View this message in context: > http://osgeo-org.1803224.n2.nabble.com/NVIZ-help-fly-keeps-resetting-jumping-back-to-center-tp5178852p5178852.html > Sent from the Grass - Users mailing list archive at Nabble.com. > _______________________________________________ > grass-user mailing list > grass-u...@lists.osgeo.org > http://lists.osgeo.org/mailman/listinfo/grass-user > > > > > > > _______________________________________________ grass-dev mailing list grass-dev@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/grass-dev