[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Problems with process*



On Apr  2, Matthew Flatt wrote:
> As far as I can tell, this requires that fdopen() returns NULL. Why
> would it do that? Maybe you could change the null-file-pointer error
> in port.c to include the value of errno; it might provide a useful
> hint.

It's not too bad so I can live with this now...


> In any case, the specific symptom should go away in 102 (starting
> with 102/10), which doesn't use fdopen() anymore. But I wonder
> whether it will be replaced by an error when reading/writing to the
> port.

So, what I'll do is keep an eye for problems once I start using the
next version (can't do it now due to work on the thing our students
use) and in case I get any problem with 102, I'll do some better
testing.

I wonder if debugging something like this is going to be easier or
harder with fd's instead of FILE*'s...  But I'll do it if I get to
that point and this problem doesn't go away.

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
         http://www.cs.cornell.edu/eli/meaning.html        Maze is Life!