When generating an image of Microscene1 I received the following error:
Facet vertexes from a line!(discarded)
Line number = 1216249
Filename = "HumveeWithSupports.gdb"
dirsig: CDproblem.nw:405: void CDProblem::createsolution(CDSolution&, bool): Assertion 'hitList.count() != 0' failed.
Did you use one of the .cfg files distributed with microscene or did you roll or own? If the latter, would it be possible to attach your cfg to this bug report?
Also, which version of DIRSIG4 are you using?
Created attachment 93 [details]
I am using version 4.
That is version 4.0.6. Also I have tried to load some of the cfg files that were provided for this scene but they cause cfg_edit to crash (ADB_FILENAME varible was never set!). I'm using verison 3.6.4 of cfg_edit, could this be the problem. When installing the distribution of 4, a version 4 of cfg_edit was not included.
Created attachment 94 [details]
PanelScene cfg file
I was able to get DIRSIG to run, using one of the provided cfg files after editing it based on previous posts. However, the image returned is all null values. The only difference between the post and what I did was change the material IDs to from 501 to 11 and 502 to 10, instead of adding these materials. I have attached the new cfg file.
Your sensor is floating way out in space; you won't be hitting any geometry out there. Try changing your position information to:
TARGET_LOCATION = 115, 90, 0
PLATFORM_LOCATION = 115, 90, 300
The assertion is triggering because DIRSIG isn't hitting the skydome (the geometry for the atmosphere). The next release of DIRSIG4 will include a more user-friendly error message.
Error message now displays in CVS version of DIRSIG4:
Are you still having problems after tweaking the platform/target or can I close this bug?
I ran the simulation again with the new sensor position and got the original error. Any thoughts?
Could you post your updated cfg? I'll see if I can reproduce the assertion again. Right now, things are working for me if I use the new platform/location.
Created attachment 95 [details]
cfg file with platform update
Created attachment 96 [details]
Error file associated with core dump
Created attachment 97 [details]
Output file associated with run
Have post the cfg, error, and output file.
Thanks for your help,
I ran make_adb again with the new sensor location and then ran dirsig, I no longer get the above error, however, the core is dumped after current platform location is written.
With your latest config file things are running fine on my development box (I had to fix the paths for my installation, including the MAPS_PATH). I'll give it a try on our Sun and Linux machines tonight.
Niek, I was able to get Microscene to run using UTM coordinates, removing the panels from the scene, and running it on our older sun box. I guess I'm unclear as to why this worked, the same cfg still does not work on our new linux cluster. If you have any questions let me know?
I'll hopefully be able to do a few runs on our linux boxes tomorrow. I'm a bit swamped right now, so it may take I while before I can work on this again.