bug#8405: Information needed for 'cp'

[Jim Meyering]

tags 8405 + moreinfo notabug
thanks

Eric Blake wrote:
> On 04/01/2011 02:42 AM, aakanksha gaur wrote:
>> Hi,
>>
>> I need to know what is the rationale behind this behavior ?
>>
>>> cp aa bb
>> aa does not exist and neither does bb. No warning is thrown on performing
>> this operation.
>>
>> However,
>> if I do
>>
>>> cp dd cc
>> where cc is a non empty file, then a warning is thrown about the source file
>> being empty| invalid.
>> This is the warning thrown.
>>
>> cp: cannot stat `dd': No such file or directory
>
> That's not a warning about being empty/invalid, but a warning about 'dd'
> not existing in the first place.
>
>>
>>
>> Can I please know the rationale behind this behavior?
>
> Most likely, you are mistaken, and 'aa' _did_ indeed exist.  cp _always_
> warns about a missing source file.
>
> But to know that for sure, it would help to see a complete trace of your
> shell session where you were testing, including ls output along the way,
> where you claim to reproduce the problem.
>
> It may also help your debugging to try 'type cp' to make sure there
> aren't any shell functions or aliases interfering with normal operation,
> as well as use 'set -vx' prior to cp to see what the shell really did
> run.  Also, 'cp -v' makes cp more verbose about what it did and tried.

By sending to address@hidden (rather than to address@hidden)
you have automatically created a bug report (http://debbugs.gnu.org/8405).
Since this seems to be due to a misunderstanding and not a bug,
I'm marking it as "notabug".  Of course, if you provide
details showing there is a legitimate bug, we'll adjust that.