On 2/15/07, Jörg Schaible <[EMAIL PROTECTED]> wrote:
Oliver Heger wrote:
> For the protocol: I think I know now what is going on:
>
> Some tests in FileUtilsCleanDirectoryTestCase try to delete files that
> have been set to read-only. This is expected to throw an exception. To
> set the read-only flag the method chmod() tries to execute the unix
> "chmod" command. If this fails (which should normally be the case on
> windows), the test is ignored.
>
> Now I am on windows, but I happen to have some unix commands in my path
> (these are windows ports of some typical unix commands), including
> chmod. So the execution of chmod is successful, but obviously the
> command does not have the desired effect: the files can be deleted, and
> no exception is thrown. This causes the tests to fail.
>
> I guess this is a rather unusual scenario.
Well, IMHO not *that* unusual. A lot of people use Cygwin (incl. myself),
MKS Toolkit, Microsoft's Posix Tools, MingW32 or software that installs
such things to run. On a Windows box c-io should not try to run chmod.
Agreed. Could someone add it to JIRA?
Hen
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
