Re: fork() for microsoft windows (native but not Cygwin).

[Tatsuro MATSUOKA]

----- Original Message -----

> From: Mike Miller 
> To: Tatsuro MATSUOKA 
> Cc: "octave-maintainers
> Date: 2016/5/2, Mon 11:36
> Subject: Re: fork() for microsoft windows (native but not Cygwin).
> 
> On Mon, May 02, 2016 at 11:25:08 +0900, Tatsuro MATSUOKA wrote:
>>  Sorry for confusing writing.
>>  On octave-4.0.2 windows, from octave prompt
>> 
>>  >> fork()
>>  ans = -1
>> 
>>  fork returns -1 but not say unimplemented
>> 
>>  I think that return "-1" is strange but show 
> "unimplemented"
>>  by
>>  >> fork()
>> 
>>  Am I wrong?
> 
> Try using the two-return-value form,
> 
>   [pid, msg] = fork ()
> 
> The msg should say something like "fork: not supported on this 
> system".
> It's not raised as an error, it's an optional output argument, I 
> don't
> know the reason why, but there it is.
> 
> -- 
> mike


OK. I have completely misled.

>> [a,b]=fork()
a = -1
b = fork: not supported on this system

However,

>> help fork
'fork' is a built-in function from the file libinterp/corefcn/syscalls.cc

 -- Built-in Function: [PID, MSG] = fork ()
     Create a copy of the current process.

This is not a comprehensive behavior for users.
But this is not specific to fork but also to all other unrecognized functions.

I registered this not comprehensive behavior to the bug tracker as a feature 
request.
http://savannah.gnu.org/bugs/?47820


On this thread, topic of "this not comprehensive behavior" is stopped.

Tatsuro