Hi Tim,
Yes, very unexpected. I don't know why dctimestep would care if the sky vector were all zeroes. Can you send me a private message with the input files that caused dctimestep to fault?
Thanks,
-Greg
···
From: Tim Perry <tim.v2.0@gmail.com>
Date: February 20, 2012 5:12:39 PM PSTI get a SEGFAULT from dctimestep if the sky vector is all zeros.
If I run dctimestep during the night it SEGFAULTs:
~/radiance/3phase$ gensky 9 21 04:00 +s -a 37.71 -o 122.21 -m 120 | genskyvec > sky.dat
~/radiance/3phase$ dctimestep /dev/shm/comp/back_glazing_mat_%03d.hdr ~/1122/dynamicRadianceTemplate/scripts/70_NoBlinds.xml win1.dmx sky.dat > pic.hdr
Segmentation fault
~/radiance/3phase$ cat sky.dat | sort | uniq
0 0 0
perry@chinstrap:~/radiance/3phase$
At a more normal time of day dctimestep works properly.
~/radiance/3phase$ gensky 9 21 08:00 +s -a 37.71 -o 122.21 -m 120 | genskyvec > sky.dat
~/radiance/3phase$ dctimestep /dev/shm/comp/back_glazing_mat_%03d.hdr ~/1122/dynamicRadianceTemplate/scripts/70_NoBlinds.xml win1.dmx sky.dat > pic.hdr
I was trying to generate a sequence of images to animate. I can substitute black frames, but I thought maybe dctimestep could be extended to handle this condition as well.
Tim