Re: Re: Trying not to miss any output from a PROCESS

> > Both processes work in parallel. It may be that the first read_line
> > returns nothing if the process did not have a chance to run yet.
> 
> That's really strange behavior. Typical behavior for read on a pipe
> is to block until input is available, or an EOF occurs. While blocking
> I/O isn't always what's wanted, it certainly solves this problem.

I don't know. I'm not an OS wizard, maybe I guessed wrongly. The code is 
really simple, if someone could check or try...

> You mean the code to read data from a pipe is *different* than the
> code to read data from a disk file? Yuch. That means I can't write
> code that should work with either, which is something that command
> line utilities *really* need to be able to do.

Know what? You're perfectly right. Thank you for showing my mistake. I'll 
certainly have to change things.


Another topic, I'm still hoping for some help for the Windows ports of 
lib/exec and lib/net. I tried to code a bit, using only MSFT docs, but 
that's a nightmare. I cannot even test if it compiles. It'll stay that way 
if nobody helps. I'll consider it means nobody *cares*. Anyway, I don't.

Best regards,
--
Cyril Adrian -- ALPLOG
@ PSA Peugeot Citroën, Sochaux, France
tél. +33 (0)3 81 33 48 76
fax +33 (0)3 81 33 16 06