(Sorry for top-posting - my newsreader made me do
it!)
Yes, that's a bug. Fixed in CVS.
Thanks!
The next version will change the behavior of
exists() somewhat. It should be less prone to
misclassification.
By the way, I almost missed this message because it
was posted as a reply rather than a new message. That caused my newsreader to
thread it way down in the middle of another discussion. The only reason I saw it
was because I was following that thread.
--Beman
(Using Boost 1.33 on VC7.1 and Win XP
Pro SP2)
I know there are talks
about possible changes to the filesystem library, but in the meantime, I have
found that ERROR_NOT_READY happens if there's no CD in the CD-ROM drive, but
fs::exists(path) returns "true" for files on that drive. My expectation
was it would return "false"....