Re: NSLog() / NSLogv() annoyance

[Adam Fedor]

On Friday, August 1, 2003, at 12:10 AM, Richard Frith-Macdonald wrote:

> > So NSLog should back-up the original? It's kind-of crazy, but 
> > probably the best solution, since there are a lot of places where 
> > gnustep-base uses c-library functions (we don't want to have to 
> > change all of them), and NSLog is probably the only place where the 
> > developer would expect errno to be the same before and after the >> call.
>
> Sorry to come in so late to this ... I've been away from home on 
> business ...
>
> In unix, the value of errno is conventionally (and I think perhaps 
> explicitly) undefined
> except for immediately after executing a system call which failed (or 
> after a library
> function which is explicitly documented to set errno).
>
> So tracking down exactly why something within NSLog() sets errno a 
> particular way
> seem of dubious benefit, since any code which relies on testing errno 
> after the call
> to NSLog() is by definition buggy  in that it is depending on 
> undefined behavior.
>
> Is there anything I've missed in this thread to give a reason why 
> NSLog() should behave
> differently from all other functions and methods?  Perhaps this is a 
> documented Apple
> behavior or something?

The only thing is that you might want to do something like call some 
library function and if there was an error, print it out, like:

file = fopen("/dev", "w");
if (file == NULL)
  NSLog(@"Got an error opening /dev, errno = %d", errno);

GNUstep will (more than likely) print out the wrong error number, since 
there are hidden library calls in NSLog, MacOSX will print out the 
correct value. I'm not sure if you should rely on this being correct 
though.