Guy Harris
changed
bug 8609
| What |
Removed |
Added |
| Status |
RESOLVED
|
CONFIRMED
|
| Resolution |
FIXED
|
---
|
Comment # 9
on bug 8609
from Guy Harris
Well, r49025 backed out the [v]printf_stderr changes, *and* adds
console_create() calls on Win32 before printing the output of -D and -L, but
that doesn't work when the standard output isn't redirected - which is odd,
because it's not obvious, other than stdout vs. stderr, why the code path does
things differently from the old [v]fprintf_stderr code path.
Note: I did that on my XP virtual machine, and the bug was filed against W7, so
maybe the old [v]fprintf_stderr code path didn't work on XP and the new code
path works on W7. I'll check that.
It also doesn't make -D work *with* the standard output redirected, which may
be because console_create() redirects the standard output and error *even if
they're going to a file*, so it undoes whatever redirection the user did.
The latter problem may require some way of figuring out whether the standard
output and/or error are going to whatever location they go by default in a
windows-subsystem process, and, if one of them isn't, don't redirect it to the
newly-created console.
You are receiving this mail because:
- You are watching all bug changes.