There is another update coming which should fix archives with variable LODs (kind a new thing in TerraVista .. well ... couple of years "new") that wasn't be around 2003 when the TileMapper was originally written
-Nick On Sun, Dec 19, 2010 at 10:47 PM, Trajce (Nick) Nikolov < [email protected]> wrote: > Hi Robert, > > here are some changes. Tested with archives 2.0, 2.1, 2.2 .. Can you give > them a try? > > Also I dont know what is the business with David's changes > > -Nick > > > On Sun, Dec 19, 2010 at 9:26 PM, Trajce (Nick) Nikolov < > [email protected]> wrote: > >> Hi Robert, >> >> I donty have handy the nellis archive. what I am seeing now with the >> archives from TV is solid work. I took in the code and the most suspitious >> place is the seam detection which I think is not quite right (the archives I >> work with are extended with new trpage nodes so this code is not int use). >> Could you place the Nellis db somewhere I can download it? I will have a >> look and will fix it >> >> Let me know >> -Nick >> >> >> On Sun, Dec 19, 2010 at 5:31 PM, Robert Osfield < >> [email protected]> wrote: >> >>> Hi Nick, >>> >>> Thanks for the clean up, the code looks a lot cleaner now :-) >>> >>> I've merged the changes and checked them in to svn/trunk. >>> >>> After merging the changes I tested osgviewer against the old Nellis >>> demo database archive.txp and found two issues, one of which is >>> clearly a regression. My test was: >>> >>> cd Nellis >>> osgviewer archive.txp >>> # zoom into the see the model >>> >>> First issue, a warning emited to the console a made be related to a >>> recent merge of a submission last week on achive creation/destruction >>> from David Fries (r11999). The warning on the console is: >>> >>> txp::ReaderWriterTXP::getArchive() error: archive id 1 not found: >>> "/archive.txp" >>> remove archive 1 size 0 result 1 >>> >>> The second issue is that cracks are appearing between tiles when >>> zooming in. This could be related to the above. I reverted your >>> clean up changes and to see if they were causing the regression and >>> confirmed that they aren't, so the changes I've just merged don't look >>> like they are related at all to the two above issues. >>> >>> Are you seeing any of these issues? >>> >>> Robert. >>> >>> >>> >>> >>> >>> >>> >>> On Fri, Dec 17, 2010 at 11:19 PM, Trajce (Nick) Nikolov >>> <[email protected]> wrote: >>> > Hi Robert, >>> > >>> > I went thru the TileMapper today very carefuly and I remove all the >>> > tileStack approach - it made everything simplier and I think more >>> proper. >>> > Cleaned the code as well. It was tested on huge archive with many lods >>> as >>> > well with variable lods. Fix attached. >>> > >>> > I swear, with this no more issues will be reported ! >>> > >>> > Updated code attached. Please CI into svn, I will call again for >>> testing >>> > (the community is silent though :-) .. is this so acient so noone is >>> using >>> > it anymore? :) >>> > >>> > Cheers, >>> > -Nick >>> > >>> > >>> > On Tue, Dec 14, 2010 at 11:20 PM, Trajce (Nick) Nikolov >>> > <[email protected]> wrote: >>> >> >>> >> Okay >>> >> -Nick >>> >> >>> >> >>> >> On Tue, Dec 14, 2010 at 11:10 PM, Robert Osfield >>> >> <[email protected]> wrote: >>> >>> >>> >>> Hi Nick, >>> >>> >>> >>> On Tue, Dec 14, 2010 at 6:43 PM, Trajce (Nick) Nikolov >>> >>> <[email protected]> wrote: >>> >>> > So, good time to call the community to do tests. If all ok, then I >>> >>> > clean it >>> >>> >>> >>> Could you do the call for testing and coordinate the thread as I've >>> >>> got a number of other issues to juggle with right now. >>> >>> >>> >>> Cheers, >>> >>> Robert. >>> >>> _______________________________________________ >>> >>> osg-submissions mailing list >>> >>> [email protected] >>> >>> >>> >>> >>> http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org >>> >> >>> > >>> > >>> > _______________________________________________ >>> > osg-submissions mailing list >>> > [email protected] >>> > >>> http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org >>> > >>> > >>> _______________________________________________ >>> osg-submissions mailing list >>> [email protected] >>> >>> http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org >>> >> >> >
_______________________________________________ osg-submissions mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
