Re: [2/3] ntdll: don't treat DOS paths starting with / as Unix paths
Vincent Povirk <madewokherd+8cd9(a)gmail.com> writes:
This patchset should fix bug 15883.
Support for Unix paths is a useful feature, you can't just remove it. -- Alexandre Julliard julliard(a)winehq.org
On Mon, Apr 6, 2009 at 5:23 AM, Alexandre Julliard <julliard(a)winehq.org> wrote:
Support for Unix paths is a useful feature, you can't just remove it.
How can the bug be fixed without removing it? Vincent Povirk
Vincent Povirk <madewokherd+8cd9(a)gmail.com> writes:
On Mon, Apr 6, 2009 at 5:23 AM, Alexandre Julliard <julliard(a)winehq.org> wrote:
Support for Unix paths is a useful feature, you can't just remove it.
How can the bug be fixed without removing it?
By coming up with a better heuristic for detecting Unix paths. -- Alexandre Julliard julliard(a)winehq.org
On Mon, Apr 6, 2009 at 8:41 AM, Alexandre Julliard <julliard(a)winehq.org> wrote:
By coming up with a better heuristic for detecting Unix paths.
It would have to reject "/usr/lib/libSDL_ttf.a" and "/bin/sh", both of which are likely to be Unix paths to files that exist. Vincent Povirk
Am Montag, 6. April 2009 15:41:51 schrieb Alexandre Julliard:
Vincent Povirk <madewokherd+8cd9(a)gmail.com> writes:
On Mon, Apr 6, 2009 at 5:23 AM, Alexandre Julliard <julliard(a)winehq.org> wrote:
Support for Unix paths is a useful feature, you can't just remove it.
How can the bug be fixed without removing it?
By coming up with a better heuristic for detecting Unix paths. Suggestion: Check if the windows path leads to an existing file, if it does, use the windows style. Otherwise try unix. Or the other way around
2009/4/6 Stefan Dösinger <stefandoesinger(a)gmx.at>:
Am Montag, 6. April 2009 15:41:51 schrieb Alexandre Julliard:
Vincent Povirk <madewokherd+8cd9(a)gmail.com> writes:
On Mon, Apr 6, 2009 at 5:23 AM, Alexandre Julliard <julliard(a)winehq.org> wrote:
Support for Unix paths is a useful feature, you can't just remove it.
How can the bug be fixed without removing it?
By coming up with a better heuristic for detecting Unix paths. Suggestion: Check if the windows path leads to an existing file, if it does, use the windows style. Otherwise try unix. Or the other way around
Trying host-system path first wouldn't fix the bug in question :) I think the only way to do this would be to test if the path exists as a Windows path before assuming it's a host-system path.
On Mon, Apr 6, 2009 at 8:27 PM, Ben Klein <shacklein(a)gmail.com> wrote:
2009/4/6 Stefan Dösinger <stefandoesinger(a)gmx.at>:
Am Montag, 6. April 2009 15:41:51 schrieb Alexandre Julliard:
Vincent Povirk <madewokherd+8cd9(a)gmail.com> writes:
On Mon, Apr 6, 2009 at 5:23 AM, Alexandre Julliard <julliard(a)winehq.org> wrote:
Support for Unix paths is a useful feature, you can't just remove it.
How can the bug be fixed without removing it?
By coming up with a better heuristic for detecting Unix paths. Suggestion: Check if the windows path leads to an existing file, if it does, use the windows style. Otherwise try unix. Or the other way around
Trying host-system path first wouldn't fix the bug in question :)
I think the only way to do this would be to test if the path exists as a Windows path before assuming it's a host-system path.
Even if you check for a Windows path first, that will still break the python tests (they check for /bin/sh and expect it to not exist). Vincent Povirk
On Monday 06 April 2009 6:39:47 pm Vincent Povirk wrote:
Even if you check for a Windows path first, that will still break the python tests (they check for /bin/sh and expect it to not exist).
Isn't that the python test's fault, then? Why can't c:\bin\sh exist?
participants (5)
-
Alexandre Julliard -
Ben Klein -
Chris Robinson -
Stefan Dösinger -
Vincent Povirk