Georg Mischler wrote:
Peter Apian-Bennewitz wrote:
Hi folks,
having fixed a mior bug with the generator of the HEAD zip file (due to
the fact that zip doesn't erase previous content), I found the path
ray/test/test data/histo.dat
Are you sure you the blank in it is a good thing ? An underscore would
be as feasable, IMHO, and makes life of shell users so much easier. Even
if the original authors of the test scenes prefer a different parser.
I used a space there very intentionally. We need to make sure
that Radiance handles this case (among othes) correctly, which
is, after all, the sole purpose of the test directory tree.This is again in preparation for the Windows port. Other than on
unix, Windows users will uses space characters in file and
directory names without thinking about it. And even some system
directories are named that way (depending on the language).
Whatever we do, we just need to assume that there can be space
characters. Any program or script that fails to function
correctly under those circumstances is to be considered broken.Sorry for any extra work that might cause for your auxiliary
scripting, but I don't see a good way around it.-schorsch
You don't cause me worry, but others will be bothered.
Windows breaks on half-a-dozen characters which are perfectly handled by UNIX shells. Some backup programs (take HP's Omniback as an example) fail on some of these too. Shells treat spaces as default delimiters. It's all stupid, but it is a fact. That's why there's a somewhat friendly common sense not to use characters which cause problems on other systems (like space, double colon, ambersands, etc) volantarily. If you insist on a space in a directory name of a general test scenario and declare the whole directory, which serves others purposes, a test case for filename handling, that's fine, but a little pointless IMHO.
-Peter
···
--
pab-opto, Freiburg, Germany, http://www.pab-opto.de
[see web page to check digital email signature]