Re: STatus of MPS branch

[Gerd Möllmann]

Gerd Möllmann <gerd.moellmann@gmail.com> writes:

> Michael Albinus <michael.albinus@gmx.de> writes:
>
>> Gerd Möllmann <gerd.moellmann@gmail.com> writes:
>>
>> Hi Gerd,
>>
>>>> Could you please instrument tramp-tests.el with the attached patch. Then
>>>> rerun the tests with
>>>>
>>>> # make -C test tramp-tests.log
>>>>
>>>> and show the log file.
>>>
>>> Danke, Michael. Please find attached
>>
>> Well, this is weird. "Hallo 1" shows the invocation of
>> '(insert-directory tmp-name1 nil)', which should just return the file
>> name. In my case, it is
>>
>> Hallo 1
>> /mock:gandalf:/tmp/tramp-test9LSbRf
>>
>>
>> But your log shows the long listing
>>
>> Hallo 1
>>   drwxr-xr-x 3 gerd staff 96 Apr 20 18:40 
>> /mock:Pro.fritz.box:/var/folders/1d/k_6t25f94sl83szqbf8gpkrh0000gn/T/tramp-testT6CzoZ
>>
>>
>> Hmm, we must write more traces. Could you, pls, apply the extended
>> patch?. Running the test could be restricted to the test case in
>> question via
>>
>> # make -C test tramp-tests.log SELECTOR=tramp-test17-insert-directory
>>
>> Let's see what tramp-tests.log tells us.
>
> Thanks. Before I do that, could this be caused by macOS' ls vs. gls?
>
>   % gls --version
>   ls (GNU coreutils) 9.5
>
> while ls is the macOS one (probably from FreeBSD, but the man page
> doesn't say so).

I found this in mit init.el

  (if (string-match "Pro" system-name)
      (setq insert-directory-program "/usr/local/bin/gls")
    (setq insert-directory-program "/opt/homebrew/bin/gls"))

meaning that the tests wouldn't do that, right?