Re: cygnus bugs

[Fergus Henderson <fjh at cs dot mu dot OZ dot AU>]

"$Bill Luebkert" <dbe@wgn.net> writes:

>Fergus Henderson wrote:
>> 
>> franl@world.std.omit-this.com (Francis Litterio) writes:
>> 
>> >[Fergus Henderson wrote:]
>> >
>> >> How can cygwin know which arguments are pathnames and which are
>> >> just ordinary strings that should not be transformed in this manner?
>> >
>> >Cygwin can know because the strings that are pathnames are passed to
>> >open(), [...]
>> 
>> No, in the case that Tim Newsham was referring to, "vi is not compiled
>> with cygwin", and so the strings that are pathnames are not passed
>> to cygwin's open().
>
>Maybe what he wants is for bash to transform the paths that 
>fall on mountpoints.  This seems fairly trivial.

You still didn't answer the question: how can cygwin know which
arguments are pathnames?  Finding a correct solution to this
problem is entirely non-trivial.  If you mean that cygwin should
treat everything that looks like it might be a pathname as a pathname,
then that is not a correct solution, and I think an incorrect
solution is going to cause more problems than it solves.
For example, consider the command `sed /usr/p /usr/p'.
In that command, the first argument is a sed command, meaning
"print all lines containing `usr'", while the second argument
is a file name.

--
Fergus Henderson <fjh@cs.mu.oz.au>   |  "I have always known that the pursuit
WWW: <http://www.cs.mu.oz.au/~fjh>   |  of excellence is a lethal habit"
PGP: finger fjh@128.250.37.3         |     -- the last words of T. S. Garp.
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".